combobox

.github/workflows/docker-build-push-cloud-web-container-on-tag.yml

@@ -65,7 +65,6 @@ jobs:
             NEXT_PUBLIC_POSTHOG_KEY=${{ secrets.POSTHOG_KEY }}
             NEXT_PUBLIC_POSTHOG_HOST=${{ secrets.POSTHOG_HOST }}
             NEXT_PUBLIC_SENTRY_DSN=${{ secrets.SENTRY_DSN }}
-            NEXT_PUBLIC_GTM_ENABLED=true
           # needed due to weird interactions with the builds for different platforms
           no-cache: true
           labels: ${{ steps.meta.outputs.labels }}

.github/workflows/pr-Integration-tests.yml

@@ -13,10 +13,7 @@ on:
 env:
   OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
   SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
-  CONFLUENCE_TEST_SPACE_URL: ${{ secrets.CONFLUENCE_TEST_SPACE_URL }}
-  CONFLUENCE_USER_NAME: ${{ secrets.CONFLUENCE_USER_NAME }}
-  CONFLUENCE_ACCESS_TOKEN: ${{ secrets.CONFLUENCE_ACCESS_TOKEN }}
-  
+
 jobs:
   integration-tests:
     # See https://runs-on.com/runners/linux/
@@ -198,9 +195,6 @@ jobs:
             -e API_SERVER_HOST=api_server \
             -e OPENAI_API_KEY=${OPENAI_API_KEY} \
             -e SLACK_BOT_TOKEN=${SLACK_BOT_TOKEN} \
-            -e CONFLUENCE_TEST_SPACE_URL=${CONFLUENCE_TEST_SPACE_URL} \
-            -e CONFLUENCE_USER_NAME=${CONFLUENCE_USER_NAME} \
-            -e CONFLUENCE_ACCESS_TOKEN=${CONFLUENCE_ACCESS_TOKEN} \
             -e TEST_WEB_HOSTNAME=test-runner \
             danswer/danswer-integration:test \
             /app/tests/integration/tests \

.github/workflows/pr-chromatic-tests.yml

@@ -1,225 +0,0 @@
-name: Run Chromatic Tests
-concurrency:
-  group: Run-Chromatic-Tests-${{ github.workflow }}-${{ github.head_ref || github.event.workflow_run.head_branch || github.run_id }}
-  cancel-in-progress: true
-
-on: push
-
-env:
-  OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
-  SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
-
-jobs:
-  playwright-tests:
-    name: Playwright Tests
-
-    # See https://runs-on.com/runners/linux/
-    runs-on: [runs-on,runner=8cpu-linux-x64,ram=16,"run-id=${{ github.run_id }}"]
-    steps:
-      - name: Checkout code
-        uses: actions/checkout@v4
-        with:
-          fetch-depth: 0
-          
-      - name: Set up Python
-        uses: actions/setup-python@v5
-        with:
-          python-version: '3.11'
-          cache: 'pip'
-          cache-dependency-path: |
-            backend/requirements/default.txt
-            backend/requirements/dev.txt
-            backend/requirements/model_server.txt
-      - run: |
-          python -m pip install --upgrade pip
-          pip install --retries 5 --timeout 30 -r backend/requirements/default.txt
-          pip install --retries 5 --timeout 30 -r backend/requirements/dev.txt
-          pip install --retries 5 --timeout 30 -r backend/requirements/model_server.txt
-        
-      - name: Setup node
-        uses: actions/setup-node@v4
-        with:
-          node-version: 22
-
-      - name: Install node dependencies
-        working-directory: ./web
-        run: npm ci
-
-      - name: Install playwright browsers
-        working-directory: ./web
-        run: npx playwright install --with-deps
-        
-      - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v3
-
-      - name: Login to Docker Hub
-        uses: docker/login-action@v3
-        with:
-          username: ${{ secrets.DOCKER_USERNAME }}
-          password: ${{ secrets.DOCKER_TOKEN }}
-
-      # tag every docker image with "test" so that we can spin up the correct set
-      # of images during testing
-      
-      # we use the runs-on cache for docker builds
-      # in conjunction with runs-on runners, it has better speed and unlimited caching
-      # https://runs-on.com/caching/s3-cache-for-github-actions/
-      # https://runs-on.com/caching/docker/
-      # https://github.com/moby/buildkit#s3-cache-experimental
-      
-      # images are built and run locally for testing purposes. Not pushed.
-
-      - name: Build Web Docker image
-        uses: ./.github/actions/custom-build-and-push
-        with:
-          context: ./web
-          file: ./web/Dockerfile
-          platforms: linux/amd64
-          tags: danswer/danswer-web-server:test
-          push: false
-          load: true
-          cache-from: type=s3,prefix=cache/${{ github.repository }}/integration-tests/web-server/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }}
-          cache-to: type=s3,prefix=cache/${{ github.repository }}/integration-tests/web-server/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }},mode=max
-
-      - name: Build Backend Docker image
-        uses: ./.github/actions/custom-build-and-push
-        with:
-          context: ./backend
-          file: ./backend/Dockerfile
-          platforms: linux/amd64
-          tags: danswer/danswer-backend:test
-          push: false
-          load: true
-          cache-from: type=s3,prefix=cache/${{ github.repository }}/integration-tests/backend/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }}
-          cache-to: type=s3,prefix=cache/${{ github.repository }}/integration-tests/backend/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }},mode=max
-
-      - name: Build Model Server Docker image
-        uses: ./.github/actions/custom-build-and-push
-        with:
-          context: ./backend
-          file: ./backend/Dockerfile.model_server
-          platforms: linux/amd64
-          tags: danswer/danswer-model-server:test
-          push: false
-          load: true
-          cache-from: type=s3,prefix=cache/${{ github.repository }}/integration-tests/model-server/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }}
-          cache-to: type=s3,prefix=cache/${{ github.repository }}/integration-tests/model-server/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }},mode=max
-
-      - name: Start Docker containers
-        run: |
-          cd deployment/docker_compose
-          ENABLE_PAID_ENTERPRISE_EDITION_FEATURES=true \
-          AUTH_TYPE=basic \
-          REQUIRE_EMAIL_VERIFICATION=false \
-          DISABLE_TELEMETRY=true \
-          IMAGE_TAG=test \
-          docker compose -f docker-compose.dev.yml -p danswer-stack up -d
-        id: start_docker
-
-      - name: Wait for service to be ready
-        run: |
-          echo "Starting wait-for-service script..."
-          
-          docker logs -f danswer-stack-api_server-1 &
-
-          start_time=$(date +%s)
-          timeout=300  # 5 minutes in seconds
-          
-          while true; do
-            current_time=$(date +%s)
-            elapsed_time=$((current_time - start_time))
-            
-            if [ $elapsed_time -ge $timeout ]; then
-              echo "Timeout reached. Service did not become ready in 5 minutes."
-              exit 1
-            fi
-            
-            # Use curl with error handling to ignore specific exit code 56
-            response=$(curl -s -o /dev/null -w "%{http_code}" http://localhost:8080/health || echo "curl_error")
-            
-            if [ "$response" = "200" ]; then
-              echo "Service is ready!"
-              break
-            elif [ "$response" = "curl_error" ]; then
-              echo "Curl encountered an error, possibly exit code 56. Continuing to retry..."
-            else
-              echo "Service not ready yet (HTTP status $response). Retrying in 5 seconds..."
-            fi
-            
-            sleep 5
-          done
-          echo "Finished waiting for service."
-
-      - name: Run pytest playwright test init
-        working-directory: ./backend
-        env: 
-          PYTEST_IGNORE_SKIP: true
-        run: pytest -s tests/integration/tests/playwright/test_playwright.py
-
-      - name: Run Playwright tests
-        working-directory: ./web
-        run: npx playwright test
-
-      - uses: actions/upload-artifact@v4
-        if: always()
-        with:
-          # Chromatic automatically defaults to the test-results directory.
-          # Replace with the path to your custom directory and adjust the CHROMATIC_ARCHIVE_LOCATION environment variable accordingly.
-          name: test-results
-          path: ./web/test-results
-          retention-days: 30
-                    
-      # save before stopping the containers so the logs can be captured
-      - name: Save Docker logs
-        if: success() || failure()
-        run: |
-          cd deployment/docker_compose
-          docker compose -f docker-compose.dev.yml -p danswer-stack logs > docker-compose.log
-          mv docker-compose.log ${{ github.workspace }}/docker-compose.log
-      
-      - name: Upload logs
-        if: success() || failure()
-        uses: actions/upload-artifact@v4
-        with:
-          name: docker-logs
-          path: ${{ github.workspace }}/docker-compose.log
-
-      - name: Stop Docker containers
-        run: |
-          cd deployment/docker_compose
-          docker compose -f docker-compose.dev.yml -p danswer-stack down -v
-
-  chromatic-tests:
-    name: Chromatic Tests
-    
-    needs: playwright-tests
-    runs-on: [runs-on,runner=8cpu-linux-x64,ram=16,"run-id=${{ github.run_id }}"]
-    steps:
-      - name: Checkout code
-        uses: actions/checkout@v4
-        with:
-          fetch-depth: 0
-          
-      - name: Setup node
-        uses: actions/setup-node@v4
-        with:
-          node-version: 22
-          
-      - name: Install node dependencies
-        working-directory: ./web
-        run: npm ci
-        
-      - name: Download Playwright test results
-        uses: actions/download-artifact@v4
-        with:
-          name: test-results
-          path: ./web/test-results
-          
-      - name: Run Chromatic
-        uses: chromaui/action@latest
-        with:
-          playwright: true
-          projectToken: ${{ secrets.CHROMATIC_PROJECT_TOKEN }}
-          workingDir: ./web
-        env: 
-          CHROMATIC_ARCHIVE_LOCATION: ./test-results

.github/workflows/pr-python-connector-tests.yml

@@ -20,7 +20,6 @@ env:
   JIRA_API_TOKEN: ${{ secrets.JIRA_API_TOKEN }}
   # Google
   GOOGLE_DRIVE_SERVICE_ACCOUNT_JSON_STR: ${{ secrets.GOOGLE_DRIVE_SERVICE_ACCOUNT_JSON_STR }}
-  GOOGLE_DRIVE_OAUTH_CREDENTIALS_JSON_STR_TEST_USER_1: ${{ secrets.GOOGLE_DRIVE_OAUTH_CREDENTIALS_JSON_STR_TEST_USER_1 }}
   GOOGLE_DRIVE_OAUTH_CREDENTIALS_JSON_STR: ${{ secrets.GOOGLE_DRIVE_OAUTH_CREDENTIALS_JSON_STR }}
   GOOGLE_GMAIL_SERVICE_ACCOUNT_JSON_STR: ${{ secrets.GOOGLE_GMAIL_SERVICE_ACCOUNT_JSON_STR }}
   GOOGLE_GMAIL_OAUTH_CREDENTIALS_JSON_STR: ${{ secrets.GOOGLE_GMAIL_OAUTH_CREDENTIALS_JSON_STR }}

.gitignore

@@ -7,4 +7,3 @@
 .vscode/
 *.sw?
 /backend/tests/regression/answer_quality/search_test_config.yaml
-/web/test-results/

README.md

@@ -12,7 +12,7 @@
 <a href="https://docs.danswer.dev/" target="_blank">
     <img src="https://img.shields.io/badge/docs-view-blue" alt="Documentation">
 </a>
-<a href="https://join.slack.com/t/danswer/shared_invite/zt-2twesxdr6-5iQitKZQpgq~hYIZ~dv3KA" target="_blank">
+<a href="https://join.slack.com/t/danswer/shared_invite/zt-2lcmqw703-071hBuZBfNEOGUsLa5PXvQ" target="_blank">
     <img src="https://img.shields.io/badge/slack-join-blue.svg?logo=slack" alt="Slack">
 </a>
 <a href="https://discord.gg/TDJ59cGV2X" target="_blank">
@@ -135,7 +135,7 @@ Looking to contribute? Please check out the [Contribution Guide](CONTRIBUTING.md
 
 ## ✨Contributors
 
-<a href="https://github.com/danswer-ai/danswer/graphs/contributors">
+<a href="https://github.com/aryn-ai/sycamore/graphs/contributors">
   <img alt="contributors" src="https://contrib.rocks/image?repo=danswer-ai/danswer"/>
 </a>
 

backend/alembic/versions/177de57c21c9_display_custom_llm_models.py

@@ -1,59 +0,0 @@
-"""display custom llm models
-
-Revision ID: 177de57c21c9
-Revises: 4ee1287bd26a
-Create Date: 2024-11-21 11:49:04.488677
-
-"""
-from alembic import op
-import sqlalchemy as sa
-from sqlalchemy.dialects import postgresql
-from sqlalchemy import and_
-
-revision = "177de57c21c9"
-down_revision = "4ee1287bd26a"
-branch_labels = None
-depends_on = None
-depends_on = None
-
-
-def upgrade() -> None:
-    conn = op.get_bind()
-    llm_provider = sa.table(
-        "llm_provider",
-        sa.column("id", sa.Integer),
-        sa.column("provider", sa.String),
-        sa.column("model_names", postgresql.ARRAY(sa.String)),
-        sa.column("display_model_names", postgresql.ARRAY(sa.String)),
-    )
-
-    excluded_providers = ["openai", "bedrock", "anthropic", "azure"]
-
-    providers_to_update = sa.select(
-        llm_provider.c.id,
-        llm_provider.c.model_names,
-        llm_provider.c.display_model_names,
-    ).where(
-        and_(
-            ~llm_provider.c.provider.in_(excluded_providers),
-            llm_provider.c.model_names.isnot(None),
-        )
-    )
-
-    results = conn.execute(providers_to_update).fetchall()
-
-    for provider_id, model_names, display_model_names in results:
-        if display_model_names is None:
-            display_model_names = []
-
-        combined_model_names = list(set(display_model_names + model_names))
-        update_stmt = (
-            llm_provider.update()
-            .where(llm_provider.c.id == provider_id)
-            .values(display_model_names=combined_model_names)
-        )
-        conn.execute(update_stmt)
-
-
-def downgrade() -> None:
-    pass

backend/alembic/versions/47e5bef3a1d7_add_persona_categories.py

@@ -1,45 +0,0 @@
-"""add persona categories
-
-Revision ID: 47e5bef3a1d7
-Revises: dfbe9e93d3c7
-Create Date: 2024-11-05 18:55:02.221064
-
-"""
-from alembic import op
-import sqlalchemy as sa
-
-
-# revision identifiers, used by Alembic.
-revision = "47e5bef3a1d7"
-down_revision = "dfbe9e93d3c7"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    # Create the persona_category table
-    op.create_table(
-        "persona_category",
-        sa.Column("id", sa.Integer(), nullable=False),
-        sa.Column("name", sa.String(), nullable=False),
-        sa.Column("description", sa.String(), nullable=True),
-        sa.PrimaryKeyConstraint("id"),
-        sa.UniqueConstraint("name"),
-    )
-
-    # Add category_id to persona table
-    op.add_column("persona", sa.Column("category_id", sa.Integer(), nullable=True))
-    op.create_foreign_key(
-        "fk_persona_category",
-        "persona",
-        "persona_category",
-        ["category_id"],
-        ["id"],
-        ondelete="SET NULL",
-    )
-
-
-def downgrade() -> None:
-    op.drop_constraint("fk_persona_category", "persona", type_="foreignkey")
-    op.drop_column("persona", "category_id")
-    op.drop_table("persona_category")

backend/alembic/versions/4ee1287bd26a_add_multiple_slack_bot_support.py

@@ -1,280 +0,0 @@
-"""add_multiple_slack_bot_support
-
-Revision ID: 4ee1287bd26a
-Revises: 47e5bef3a1d7
-Create Date: 2024-11-06 13:15:53.302644
-
-"""
-import logging
-from typing import cast
-from alembic import op
-import sqlalchemy as sa
-from sqlalchemy.orm import Session
-from danswer.key_value_store.factory import get_kv_store
-from danswer.db.models import SlackBot
-from sqlalchemy.dialects import postgresql
-
-# revision identifiers, used by Alembic.
-revision = "4ee1287bd26a"
-down_revision = "47e5bef3a1d7"
-branch_labels: None = None
-depends_on: None = None
-
-# Configure logging
-logger = logging.getLogger("alembic.runtime.migration")
-logger.setLevel(logging.INFO)
-
-
-def upgrade() -> None:
-    logger.info(f"{revision}: create_table: slack_bot")
-    # Create new slack_bot table
-    op.create_table(
-        "slack_bot",
-        sa.Column("id", sa.Integer(), nullable=False),
-        sa.Column("name", sa.String(), nullable=False),
-        sa.Column("enabled", sa.Boolean(), nullable=False, server_default="true"),
-        sa.Column("bot_token", sa.LargeBinary(), nullable=False),
-        sa.Column("app_token", sa.LargeBinary(), nullable=False),
-        sa.PrimaryKeyConstraint("id"),
-        sa.UniqueConstraint("bot_token"),
-        sa.UniqueConstraint("app_token"),
-    )
-
-    # # Create new slack_channel_config table
-    op.create_table(
-        "slack_channel_config",
-        sa.Column("id", sa.Integer(), nullable=False),
-        sa.Column("slack_bot_id", sa.Integer(), nullable=True),
-        sa.Column("persona_id", sa.Integer(), nullable=True),
-        sa.Column("channel_config", postgresql.JSONB(), nullable=False),
-        sa.Column("response_type", sa.String(), nullable=False),
-        sa.Column(
-            "enable_auto_filters", sa.Boolean(), nullable=False, server_default="false"
-        ),
-        sa.ForeignKeyConstraint(
-            ["slack_bot_id"],
-            ["slack_bot.id"],
-        ),
-        sa.ForeignKeyConstraint(
-            ["persona_id"],
-            ["persona.id"],
-        ),
-        sa.PrimaryKeyConstraint("id"),
-    )
-
-    # Handle existing Slack bot tokens first
-    logger.info(f"{revision}: Checking for existing Slack bot.")
-    bot_token = None
-    app_token = None
-    first_row_id = None
-
-    try:
-        tokens = cast(dict, get_kv_store().load("slack_bot_tokens_config_key"))
-    except Exception:
-        logger.warning("No existing Slack bot tokens found.")
-        tokens = {}
-
-    bot_token = tokens.get("bot_token")
-    app_token = tokens.get("app_token")
-
-    if bot_token and app_token:
-        logger.info(f"{revision}: Found bot and app tokens.")
-
-        session = Session(bind=op.get_bind())
-        new_slack_bot = SlackBot(
-            name="Slack Bot (Migrated)",
-            enabled=True,
-            bot_token=bot_token,
-            app_token=app_token,
-        )
-        session.add(new_slack_bot)
-        session.commit()
-        first_row_id = new_slack_bot.id
-
-    # Create a default bot if none exists
-    # This is in case there are no slack tokens but there are channels configured
-    op.execute(
-        sa.text(
-            """
-            INSERT INTO slack_bot (name, enabled, bot_token, app_token)
-            SELECT 'Default Bot', true, '', ''
-            WHERE NOT EXISTS (SELECT 1 FROM slack_bot)
-            RETURNING id;
-            """
-        )
-    )
-
-    # Get the bot ID to use (either from existing migration or newly created)
-    bot_id_query = sa.text(
-        """
-        SELECT COALESCE(
-            :first_row_id,
-            (SELECT id FROM slack_bot ORDER BY id ASC LIMIT 1)
-        ) as bot_id;
-        """
-    )
-    result = op.get_bind().execute(bot_id_query, {"first_row_id": first_row_id})
-    bot_id = result.scalar()
-
-    # CTE (Common Table Expression) that transforms the old slack_bot_config table data
-    # This splits up the channel_names into their own rows
-    channel_names_cte = """
-        WITH channel_names AS (
-            SELECT
-                sbc.id as config_id,
-                sbc.persona_id,
-                sbc.response_type,
-                sbc.enable_auto_filters,
-                jsonb_array_elements_text(sbc.channel_config->'channel_names') as channel_name,
-                sbc.channel_config->>'respond_tag_only' as respond_tag_only,
-                sbc.channel_config->>'respond_to_bots' as respond_to_bots,
-                sbc.channel_config->'respond_member_group_list' as respond_member_group_list,
-                sbc.channel_config->'answer_filters' as answer_filters,
-                sbc.channel_config->'follow_up_tags' as follow_up_tags
-            FROM slack_bot_config sbc
-        )
-    """
-
-    # Insert the channel names into the new slack_channel_config table
-    insert_statement = """
-        INSERT INTO slack_channel_config (
-            slack_bot_id,
-            persona_id,
-            channel_config,
-            response_type,
-            enable_auto_filters
-        )
-        SELECT
-            :bot_id,
-            channel_name.persona_id,
-            jsonb_build_object(
-                'channel_name', channel_name.channel_name,
-                'respond_tag_only',
-                COALESCE((channel_name.respond_tag_only)::boolean, false),
-                'respond_to_bots',
-                COALESCE((channel_name.respond_to_bots)::boolean, false),
-                'respond_member_group_list',
-                COALESCE(channel_name.respond_member_group_list, '[]'::jsonb),
-                'answer_filters',
-                COALESCE(channel_name.answer_filters, '[]'::jsonb),
-                'follow_up_tags',
-                COALESCE(channel_name.follow_up_tags, '[]'::jsonb)
-            ),
-            channel_name.response_type,
-            channel_name.enable_auto_filters
-        FROM channel_names channel_name;
-    """
-
-    op.execute(sa.text(channel_names_cte + insert_statement).bindparams(bot_id=bot_id))
-
-    # Clean up old tokens if they existed
-    try:
-        if bot_token and app_token:
-            logger.info(f"{revision}: Removing old bot and app tokens.")
-            get_kv_store().delete("slack_bot_tokens_config_key")
-    except Exception:
-        logger.warning("tried to delete tokens in dynamic config but failed")
-    # Rename the table
-    op.rename_table(
-        "slack_bot_config__standard_answer_category",
-        "slack_channel_config__standard_answer_category",
-    )
-
-    # Rename the column
-    op.alter_column(
-        "slack_channel_config__standard_answer_category",
-        "slack_bot_config_id",
-        new_column_name="slack_channel_config_id",
-    )
-
-    # Drop the table with CASCADE to handle dependent objects
-    op.execute("DROP TABLE slack_bot_config CASCADE")
-
-    logger.info(f"{revision}: Migration complete.")
-
-
-def downgrade() -> None:
-    # Recreate the old slack_bot_config table
-    op.create_table(
-        "slack_bot_config",
-        sa.Column("id", sa.Integer(), nullable=False),
-        sa.Column("persona_id", sa.Integer(), nullable=True),
-        sa.Column("channel_config", postgresql.JSONB(), nullable=False),
-        sa.Column("response_type", sa.String(), nullable=False),
-        sa.Column("enable_auto_filters", sa.Boolean(), nullable=False),
-        sa.ForeignKeyConstraint(
-            ["persona_id"],
-            ["persona.id"],
-        ),
-        sa.PrimaryKeyConstraint("id"),
-    )
-
-    # Migrate data back to the old format
-    # Group by persona_id to combine channel names back into arrays
-    op.execute(
-        sa.text(
-            """
-            INSERT INTO slack_bot_config (
-                persona_id,
-                channel_config,
-                response_type,
-                enable_auto_filters
-            )
-            SELECT DISTINCT ON (persona_id)
-                persona_id,
-                jsonb_build_object(
-                    'channel_names', (
-                        SELECT jsonb_agg(c.channel_config->>'channel_name')
-                        FROM slack_channel_config c
-                        WHERE c.persona_id = scc.persona_id
-                    ),
-                    'respond_tag_only', (channel_config->>'respond_tag_only')::boolean,
-                    'respond_to_bots', (channel_config->>'respond_to_bots')::boolean,
-                    'respond_member_group_list', channel_config->'respond_member_group_list',
-                    'answer_filters', channel_config->'answer_filters',
-                    'follow_up_tags', channel_config->'follow_up_tags'
-                ),
-                response_type,
-                enable_auto_filters
-            FROM slack_channel_config scc
-            WHERE persona_id IS NOT NULL;
-            """
-        )
-    )
-
-    # Rename the table back
-    op.rename_table(
-        "slack_channel_config__standard_answer_category",
-        "slack_bot_config__standard_answer_category",
-    )
-
-    # Rename the column back
-    op.alter_column(
-        "slack_bot_config__standard_answer_category",
-        "slack_channel_config_id",
-        new_column_name="slack_bot_config_id",
-    )
-
-    # Try to save the first bot's tokens back to KV store
-    try:
-        first_bot = (
-            op.get_bind()
-            .execute(
-                sa.text(
-                    "SELECT bot_token, app_token FROM slack_bot ORDER BY id LIMIT 1"
-                )
-            )
-            .first()
-        )
-        if first_bot and first_bot.bot_token and first_bot.app_token:
-            tokens = {
-                "bot_token": first_bot.bot_token,
-                "app_token": first_bot.app_token,
-            }
-            get_kv_store().store("slack_bot_tokens_config_key", tokens)
-    except Exception:
-        logger.warning("Failed to save tokens back to KV store")
-
-    # Drop the new tables in reverse order
-    op.drop_table("slack_channel_config")
-    op.drop_table("slack_bot")

backend/alembic/versions/949b4a92a401_remove_rt.py

@@ -7,7 +7,6 @@ Create Date: 2024-10-26 13:06:06.937969
 """
 from alembic import op
 from sqlalchemy.orm import Session
-from sqlalchemy import text
 
 # Import your models and constants
 from danswer.db.models import (
@@ -16,6 +15,7 @@ from danswer.db.models import (
     Credential,
     IndexAttempt,
 )
+from danswer.configs.constants import DocumentSource
 
 
 # revision identifiers, used by Alembic.
@@ -30,11 +30,13 @@ def upgrade() -> None:
     bind = op.get_bind()
     session = Session(bind=bind)
 
-    # Get connectors using raw SQL
-    result = bind.execute(
-        text("SELECT id FROM connector WHERE source = 'requesttracker'")
+    connectors_to_delete = (
+        session.query(Connector)
+        .filter(Connector.source == DocumentSource.REQUESTTRACKER)
+        .all()
     )
-    connector_ids = [row[0] for row in result]
+
+    connector_ids = [connector.id for connector in connectors_to_delete]
 
     if connector_ids:
         cc_pairs_to_delete = (

backend/alembic/versions/9cf5c00f72fe_add_creator_to_cc_pair.py

@@ -1,7 +1,7 @@
 """add creator to cc pair
 
 Revision ID: 9cf5c00f72fe
-Revises: 26b931506ecb
+Revises: c0fd6e4da83a
 Create Date: 2024-11-12 15:16:42.682902
 
 """

backend/alembic/versions/dfbe9e93d3c7_extended_role_for_non_web.py

@@ -1,42 +0,0 @@
-"""extended_role_for_non_web
-
-Revision ID: dfbe9e93d3c7
-Revises: 9cf5c00f72fe
-Create Date: 2024-11-16 07:54:18.727906
-
-"""
-from alembic import op
-import sqlalchemy as sa
-
-# revision identifiers, used by Alembic.
-revision = "dfbe9e93d3c7"
-down_revision = "9cf5c00f72fe"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    op.execute(
-        """
-        UPDATE "user"
-        SET role = 'EXT_PERM_USER'
-        WHERE has_web_login = false
-    """
-    )
-    op.drop_column("user", "has_web_login")
-
-
-def downgrade() -> None:
-    op.add_column(
-        "user",
-        sa.Column("has_web_login", sa.Boolean(), nullable=False, server_default="true"),
-    )
-
-    op.execute(
-        """
-        UPDATE "user"
-        SET has_web_login = false,
-            role = 'BASIC'
-        WHERE role IN ('SLACK_USER', 'EXT_PERM_USER')
-    """
-    )

backend/danswer/auth/invited_users.py

@@ -2,8 +2,8 @@ from typing import cast
 
 from danswer.configs.constants import KV_USER_STORE_KEY
 from danswer.key_value_store.factory import get_kv_store
+from danswer.key_value_store.interface import JSON_ro
 from danswer.key_value_store.interface import KvKeyNotFoundError
-from danswer.utils.special_types import JSON_ro
 
 
 def get_invited_users() -> list[str]:

backend/danswer/auth/schemas.py

@@ -13,9 +13,6 @@ class UserRole(str, Enum):
         groups they are curators of
     - Global Curator can perform admin actions
         for all groups they are a member of
-    - Limited can access a limited set of basic api endpoints
-    - Slack are users that have used danswer via slack but dont have a web login
-    - External permissioned users that have been picked up during the external permissions sync process but don't have a web login
     """
 
     LIMITED = "limited"
@@ -23,14 +20,6 @@ class UserRole(str, Enum):
     ADMIN = "admin"
     CURATOR = "curator"
     GLOBAL_CURATOR = "global_curator"
-    SLACK_USER = "slack_user"
-    EXT_PERM_USER = "ext_perm_user"
-
-    def is_web_login(self) -> bool:
-        return self not in [
-            UserRole.SLACK_USER,
-            UserRole.EXT_PERM_USER,
-        ]
 
 
 class UserStatus(str, Enum):
@@ -45,8 +34,10 @@ class UserRead(schemas.BaseUser[uuid.UUID]):
 
 class UserCreate(schemas.BaseUserCreate):
     role: UserRole = UserRole.BASIC
+    has_web_login: bool | None = True
     tenant_id: str | None = None
 
 
 class UserUpdate(schemas.BaseUserUpdate):
     role: UserRole
+    has_web_login: bool | None = True

backend/danswer/auth/users.py

@@ -49,6 +49,7 @@ from httpx_oauth.oauth2 import BaseOAuth2
 from httpx_oauth.oauth2 import OAuth2Token
 from pydantic import BaseModel
 from sqlalchemy import text
+from sqlalchemy.orm import attributes
 from sqlalchemy.orm import Session
 
 from danswer.auth.api_key import get_hashed_api_key_from_request
@@ -221,8 +222,6 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
     reset_password_token_secret = USER_AUTH_SECRET
     verification_token_secret = USER_AUTH_SECRET
 
-    user_db: SQLAlchemyUserDatabase[User, uuid.UUID]
-
     async def create(
         self,
         user_create: schemas.UC | UserCreate,
@@ -248,9 +247,7 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
             verify_email_is_invited(user_create.email)
             verify_email_domain(user_create.email)
             if MULTI_TENANT:
-                tenant_user_db = SQLAlchemyUserAdminDB[User, uuid.UUID](
-                    db_session, User, OAuthAccount
-                )
+                tenant_user_db = SQLAlchemyUserAdminDB(db_session, User, OAuthAccount)
                 self.user_db = tenant_user_db
                 self.database = tenant_user_db
 
@@ -269,9 +266,14 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
             except exceptions.UserAlreadyExists:
                 user = await self.get_by_email(user_create.email)
                 # Handle case where user has used product outside of web and is now creating an account through web
-                if not user.role.is_web_login() and user_create.role.is_web_login():
+                if (
+                    not user.has_web_login
+                    and hasattr(user_create, "has_web_login")
+                    and user_create.has_web_login
+                ):
                     user_update = UserUpdate(
                         password=user_create.password,
+                        has_web_login=True,
                         role=user_create.role,
                         is_verified=user_create.is_verified,
                     )
@@ -285,7 +287,7 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
             return user
 
     async def oauth_callback(
-        self,
+        self: "BaseUserManager[models.UOAP, models.ID]",
         oauth_name: str,
         access_token: str,
         account_id: str,
@@ -296,7 +298,7 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
         *,
         associate_by_email: bool = False,
         is_verified_by_default: bool = False,
-    ) -> User:
+    ) -> models.UOAP:
         referral_source = None
         if request:
             referral_source = getattr(request.state, "referral_source", None)
@@ -322,11 +324,9 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
             verify_email_domain(account_email)
 
             if MULTI_TENANT:
-                tenant_user_db = SQLAlchemyUserAdminDB[User, uuid.UUID](
-                    db_session, User, OAuthAccount
-                )
+                tenant_user_db = SQLAlchemyUserAdminDB(db_session, User, OAuthAccount)
                 self.user_db = tenant_user_db
-                self.database = tenant_user_db
+                self.database = tenant_user_db  # type: ignore
 
             oauth_account_dict = {
                 "oauth_name": oauth_name,
@@ -378,11 +378,7 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
                         and existing_oauth_account.oauth_name == oauth_name
                     ):
                         user = await self.user_db.update_oauth_account(
-                            user,
-                            # NOTE: OAuthAccount DOES implement the OAuthAccountProtocol
-                            # but the type checker doesn't know that :(
-                            existing_oauth_account,  # type: ignore
-                            oauth_account_dict,
+                            user, existing_oauth_account, oauth_account_dict
                         )
 
             # NOTE: Most IdPs have very short expiry times, and we don't want to force the user to
@@ -395,15 +391,16 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
                 )
 
             # Handle case where user has used product outside of web and is now creating an account through web
-            if not user.role.is_web_login():
+            if not user.has_web_login:  # type: ignore
                 await self.user_db.update(
                     user,
                     {
                         "is_verified": is_verified_by_default,
-                        "role": UserRole.BASIC,
+                        "has_web_login": True,
                     },
                 )
                 user.is_verified = is_verified_by_default
+                user.has_web_login = True  # type: ignore
 
             # this is needed if an organization goes from `TRACK_EXTERNAL_IDP_EXPIRY=true` to `false`
             # otherwise, the oidc expiry will always be old, and the user will never be able to login
@@ -478,7 +475,9 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
                 self.password_helper.hash(credentials.password)
                 return None
 
-            if not user.role.is_web_login():
+            has_web_login = attributes.get_attribute(user, "has_web_login")
+
+            if not has_web_login:
                 raise BasicAuthenticationError(
                     detail="NO_WEB_LOGIN_AND_HAS_NO_PASSWORD",
                 )

backend/danswer/background/celery/apps/primary.py

@@ -1,6 +1,5 @@
 import multiprocessing
 from typing import Any
-from typing import cast
 
 from celery import bootsteps  # type: ignore
 from celery import Celery
@@ -15,16 +14,10 @@ from celery.signals import worker_shutdown
 import danswer.background.celery.apps.app_base as app_base
 from danswer.background.celery.apps.app_base import task_logger
 from danswer.background.celery.celery_utils import celery_is_worker_primary
-from danswer.background.celery.tasks.indexing.tasks import (
-    get_unfenced_index_attempt_ids,
-)
 from danswer.configs.constants import CELERY_PRIMARY_WORKER_LOCK_TIMEOUT
 from danswer.configs.constants import DanswerRedisLocks
 from danswer.configs.constants import POSTGRES_CELERY_WORKER_PRIMARY_APP_NAME
-from danswer.db.engine import get_session_with_default_tenant
 from danswer.db.engine import SqlEngine
-from danswer.db.index_attempt import get_index_attempt
-from danswer.db.index_attempt import mark_attempt_failed
 from danswer.redis.redis_connector_credential_pair import RedisConnectorCredentialPair
 from danswer.redis.redis_connector_delete import RedisConnectorDelete
 from danswer.redis.redis_connector_doc_perm_sync import RedisConnectorPermissionSync
@@ -98,15 +91,6 @@ def on_worker_init(sender: Any, **kwargs: Any) -> None:
     # by the primary worker. This is unnecessary in the multi tenant scenario
     r = get_redis_client(tenant_id=None)
 
-    # Log the role and slave count - being connected to a slave or slave count > 0 could be problematic
-    info: dict[str, Any] = cast(dict, r.info("replication"))
-    role: str = cast(str, info.get("role"))
-    connected_slaves: int = info.get("connected_slaves", 0)
-
-    logger.info(
-        f"Redis INFO REPLICATION: role={role} connected_slaves={connected_slaves}"
-    )
-
     # For the moment, we're assuming that we are the only primary worker
     # that should be running.
     # TODO: maybe check for or clean up another zombie primary worker if we detect it
@@ -156,23 +140,6 @@ def on_worker_init(sender: Any, **kwargs: Any) -> None:
 
     RedisConnectorExternalGroupSync.reset_all(r)
 
-    # mark orphaned index attempts as failed
-    with get_session_with_default_tenant() as db_session:
-        unfenced_attempt_ids = get_unfenced_index_attempt_ids(db_session, r)
-        for attempt_id in unfenced_attempt_ids:
-            attempt = get_index_attempt(db_session, attempt_id)
-            if not attempt:
-                continue
-
-            failure_reason = (
-                f"Orphaned index attempt found on startup: "
-                f"index_attempt={attempt.id} "
-                f"cc_pair={attempt.connector_credential_pair_id} "
-                f"search_settings={attempt.search_settings_id}"
-            )
-            logger.warning(failure_reason)
-            mark_attempt_failed(attempt.id, db_session, failure_reason)
-
 
 @worker_ready.connect
 def on_worker_ready(sender: Any, **kwargs: Any) -> None:

backend/danswer/background/celery/celery_utils.py

@@ -4,6 +4,7 @@ from typing import Any
 
 from sqlalchemy.orm import Session
 
+from danswer.background.indexing.run_indexing import RunIndexingCallbackInterface
 from danswer.configs.app_configs import MAX_PRUNING_DOCUMENT_RETRIEVAL_PER_MINUTE
 from danswer.connectors.cross_connector_utils.rate_limit_wrapper import (
     rate_limit_builder,
@@ -16,7 +17,6 @@ from danswer.connectors.models import Document
 from danswer.db.connector_credential_pair import get_connector_credential_pair
 from danswer.db.enums import TaskStatus
 from danswer.db.models import TaskQueueState
-from danswer.indexing.indexing_heartbeat import IndexingHeartbeatInterface
 from danswer.redis.redis_connector import RedisConnector
 from danswer.server.documents.models import DeletionAttemptSnapshot
 from danswer.utils.logger import setup_logger
@@ -78,7 +78,7 @@ def document_batch_to_ids(
 
 def extract_ids_from_runnable_connector(
     runnable_connector: BaseConnector,
-    callback: IndexingHeartbeatInterface | None = None,
+    callback: RunIndexingCallbackInterface | None = None,
 ) -> set[str]:
     """
     If the SlimConnector hasnt been implemented for the given connector, just pull
@@ -111,15 +111,10 @@ def extract_ids_from_runnable_connector(
     for doc_batch in doc_batch_generator:
         if callback:
             if callback.should_stop():
-                raise RuntimeError(
-                    "extract_ids_from_runnable_connector: Stop signal detected"
-                )
-
+                raise RuntimeError("Stop signal received")
+            callback.progress(len(doc_batch))
         all_connector_doc_ids.update(doc_batch_processing_func(doc_batch))
 
-        if callback:
-            callback.progress("extract_ids_from_runnable_connector", len(doc_batch))
-
     return all_connector_doc_ids
 
 

backend/danswer/background/celery/tasks/connector_deletion/tasks.py

@@ -1,12 +1,12 @@
 from datetime import datetime
 from datetime import timezone
 
+import redis
 from celery import Celery
 from celery import shared_task
 from celery import Task
 from celery.exceptions import SoftTimeLimitExceeded
 from redis import Redis
-from redis.lock import Lock as RedisLock
 from sqlalchemy.orm import Session
 
 from danswer.background.celery.apps.app_base import task_logger
@@ -19,7 +19,7 @@ from danswer.db.engine import get_session_with_tenant
 from danswer.db.enums import ConnectorCredentialPairStatus
 from danswer.db.search_settings import get_all_search_settings
 from danswer.redis.redis_connector import RedisConnector
-from danswer.redis.redis_connector_delete import RedisConnectorDeletePayload
+from danswer.redis.redis_connector_delete import RedisConnectorDeletionFenceData
 from danswer.redis.redis_pool import get_redis_client
 
 
@@ -87,7 +87,7 @@ def try_generate_document_cc_pair_cleanup_tasks(
     cc_pair_id: int,
     db_session: Session,
     r: Redis,
-    lock_beat: RedisLock,
+    lock_beat: redis.lock.Lock,
     tenant_id: str | None,
 ) -> int | None:
     """Returns an int if syncing is needed. The int represents the number of sync tasks generated.
@@ -118,7 +118,7 @@ def try_generate_document_cc_pair_cleanup_tasks(
         return None
 
     # set a basic fence to start
-    fence_payload = RedisConnectorDeletePayload(
+    fence_payload = RedisConnectorDeletionFenceData(
         num_tasks=None,
         submitted=datetime.now(timezone.utc),
     )

backend/danswer/background/celery/tasks/doc_permission_syncing/tasks.py

@@ -24,7 +24,7 @@ from danswer.db.engine import get_session_with_tenant
 from danswer.db.enums import AccessType
 from danswer.db.enums import ConnectorCredentialPairStatus
 from danswer.db.models import ConnectorCredentialPair
-from danswer.db.users import batch_add_ext_perm_user_if_not_exists
+from danswer.db.users import batch_add_non_web_user_if_not_exists
 from danswer.redis.redis_connector import RedisConnector
 from danswer.redis.redis_connector_doc_perm_sync import (
     RedisConnectorPermissionSyncData,
@@ -301,7 +301,7 @@ def update_external_document_permissions_task(
     try:
         with get_session_with_tenant(tenant_id) as db_session:
             # Then we build the update requests to update vespa
-            batch_add_ext_perm_user_if_not_exists(
+            batch_add_non_web_user_if_not_exists(
                 db_session=db_session,
                 emails=list(external_access.external_user_emails),
             )

backend/danswer/background/celery/tasks/external_group_syncing/tasks.py

@@ -29,7 +29,7 @@ from danswer.utils.logger import setup_logger
 from ee.danswer.db.connector_credential_pair import get_all_auto_sync_cc_pairs
 from ee.danswer.db.external_perm import ExternalUserGroup
 from ee.danswer.db.external_perm import replace_user__ext_group_for_cc_pair
-from ee.danswer.external_permissions.sync_params import EXTERNAL_GROUP_SYNC_PERIODS
+from ee.danswer.external_permissions.sync_params import EXTERNAL_GROUP_SYNC_PERIOD
 from ee.danswer.external_permissions.sync_params import GROUP_PERMISSIONS_FUNC_MAP
 
 logger = setup_logger()
@@ -66,9 +66,9 @@ def _is_external_group_sync_due(cc_pair: ConnectorCredentialPair) -> bool:
     if last_ext_group_sync is None:
         return True
 
-    source_sync_period = EXTERNAL_GROUP_SYNC_PERIODS.get(cc_pair.connector.source)
+    source_sync_period = EXTERNAL_GROUP_SYNC_PERIOD
 
-    # If EXTERNAL_GROUP_SYNC_PERIODS is None, we always run the sync.
+    # If EXTERNAL_GROUP_SYNC_PERIOD is None, we always run the sync.
     if not source_sync_period:
         return True
 

backend/danswer/background/celery/tasks/indexing/tasks.py

@@ -10,13 +10,12 @@ from celery import shared_task
 from celery import Task
 from celery.exceptions import SoftTimeLimitExceeded
 from redis import Redis
-from redis.exceptions import LockError
-from redis.lock import Lock as RedisLock
 from sqlalchemy.orm import Session
 
 from danswer.background.celery.apps.app_base import task_logger
 from danswer.background.indexing.job_client import SimpleJobClient
 from danswer.background.indexing.run_indexing import run_indexing_entrypoint
+from danswer.background.indexing.run_indexing import RunIndexingCallbackInterface
 from danswer.configs.app_configs import DISABLE_INDEX_UPDATE_ON_SWAP
 from danswer.configs.constants import CELERY_INDEXING_LOCK_TIMEOUT
 from danswer.configs.constants import CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT
@@ -33,8 +32,6 @@ from danswer.db.enums import ConnectorCredentialPairStatus
 from danswer.db.enums import IndexingStatus
 from danswer.db.enums import IndexModelStatus
 from danswer.db.index_attempt import create_index_attempt
-from danswer.db.index_attempt import delete_index_attempt
-from danswer.db.index_attempt import get_all_index_attempts_by_status
 from danswer.db.index_attempt import get_index_attempt
 from danswer.db.index_attempt import get_last_attempt_for_cc_pair
 from danswer.db.index_attempt import mark_attempt_failed
@@ -44,12 +41,10 @@ from danswer.db.models import SearchSettings
 from danswer.db.search_settings import get_current_search_settings
 from danswer.db.search_settings import get_secondary_search_settings
 from danswer.db.swap_index import check_index_swap
-from danswer.indexing.indexing_heartbeat import IndexingHeartbeatInterface
 from danswer.natural_language_processing.search_nlp_models import EmbeddingModel
 from danswer.natural_language_processing.search_nlp_models import warm_up_bi_encoder
 from danswer.redis.redis_connector import RedisConnector
-from danswer.redis.redis_connector_index import RedisConnectorIndex
-from danswer.redis.redis_connector_index import RedisConnectorIndexPayload
+from danswer.redis.redis_connector_index import RedisConnectorIndexingFenceData
 from danswer.redis.redis_pool import get_redis_client
 from danswer.utils.logger import setup_logger
 from danswer.utils.variable_functionality import global_version
@@ -61,97 +56,30 @@ from shared_configs.configs import SENTRY_DSN
 logger = setup_logger()
 
 
-class IndexingCallback(IndexingHeartbeatInterface):
+class RunIndexingCallback(RunIndexingCallbackInterface):
     def __init__(
         self,
         stop_key: str,
         generator_progress_key: str,
-        redis_lock: RedisLock,
+        redis_lock: redis.lock.Lock,
         redis_client: Redis,
     ):
         super().__init__()
-        self.redis_lock: RedisLock = redis_lock
+        self.redis_lock: redis.lock.Lock = redis_lock
         self.stop_key: str = stop_key
         self.generator_progress_key: str = generator_progress_key
         self.redis_client = redis_client
-        self.started: datetime = datetime.now(timezone.utc)
-        self.redis_lock.reacquire()
-
-        self.last_tag: str = ""
-        self.last_lock_reacquire: datetime = datetime.now(timezone.utc)
 
     def should_stop(self) -> bool:
         if self.redis_client.exists(self.stop_key):
             return True
         return False
 
-    def progress(self, tag: str, amount: int) -> None:
-        try:
-            self.redis_lock.reacquire()
-            self.last_tag = tag
-            self.last_lock_reacquire = datetime.now(timezone.utc)
-        except LockError:
-            logger.exception(
-                f"IndexingCallback - lock.reacquire exceptioned. "
-                f"lock_timeout={self.redis_lock.timeout} "
-                f"start={self.started} "
-                f"last_tag={self.last_tag} "
-                f"last_reacquired={self.last_lock_reacquire} "
-                f"now={datetime.now(timezone.utc)}"
-            )
-            raise
-
+    def progress(self, amount: int) -> None:
+        self.redis_lock.reacquire()
         self.redis_client.incrby(self.generator_progress_key, amount)
 
 
-def get_unfenced_index_attempt_ids(db_session: Session, r: redis.Redis) -> list[int]:
-    """Gets a list of unfenced index attempts. Should not be possible, so we'd typically
-    want to clean them up.
-
-    Unfenced = attempt not in terminal state and fence does not exist.
-    """
-    unfenced_attempts: list[int] = []
-
-    # inner/outer/inner double check pattern to avoid race conditions when checking for
-    # bad state
-    # inner = index_attempt in non terminal state
-    # outer = r.fence_key down
-
-    # check the db for index attempts in a non terminal state
-    attempts: list[IndexAttempt] = []
-    attempts.extend(
-        get_all_index_attempts_by_status(IndexingStatus.NOT_STARTED, db_session)
-    )
-    attempts.extend(
-        get_all_index_attempts_by_status(IndexingStatus.IN_PROGRESS, db_session)
-    )
-
-    for attempt in attempts:
-        fence_key = RedisConnectorIndex.fence_key_with_ids(
-            attempt.connector_credential_pair_id, attempt.search_settings_id
-        )
-
-        # if the fence is down / doesn't exist, possible error but not confirmed
-        if r.exists(fence_key):
-            continue
-
-        # Between the time the attempts are first looked up and the time we see the fence down,
-        # the attempt may have completed and taken down the fence normally.
-
-        # We need to double check that the index attempt is still in a non terminal state
-        # and matches the original state, which confirms we are really in a bad state.
-        attempt_2 = get_index_attempt(db_session, attempt.id)
-        if not attempt_2:
-            continue
-
-        if attempt.status != attempt_2.status:
-            continue
-
-        unfenced_attempts.append(attempt.id)
-
-    return unfenced_attempts
-
-
 @shared_task(
     name="check_for_indexing",
     soft_time_limit=300,
@@ -162,7 +90,7 @@ def check_for_indexing(self: Task, *, tenant_id: str | None) -> int | None:
 
     r = get_redis_client(tenant_id=tenant_id)
 
-    lock_beat: RedisLock = r.lock(
+    lock_beat = r.lock(
         DanswerRedisLocks.CHECK_INDEXING_BEAT_LOCK,
         timeout=CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT,
     )
@@ -172,7 +100,6 @@ def check_for_indexing(self: Task, *, tenant_id: str | None) -> int | None:
         if not lock_beat.acquire(blocking=False):
             return None
 
-        # check for search settings swap
         with get_session_with_tenant(tenant_id=tenant_id) as db_session:
             old_search_settings = check_index_swap(db_session=db_session)
             current_search_settings = get_current_search_settings(db_session)
@@ -191,18 +118,13 @@ def check_for_indexing(self: Task, *, tenant_id: str | None) -> int | None:
                         embedding_model=embedding_model,
                     )
 
-        # gather cc_pair_ids
         cc_pair_ids: list[int] = []
         with get_session_with_tenant(tenant_id) as db_session:
-            lock_beat.reacquire()
             cc_pairs = fetch_connector_credential_pairs(db_session)
             for cc_pair_entry in cc_pairs:
                 cc_pair_ids.append(cc_pair_entry.id)
 
-        # kick off index attempts
         for cc_pair_id in cc_pair_ids:
-            lock_beat.reacquire()
-
             redis_connector = RedisConnector(tenant_id, cc_pair_id)
             with get_session_with_tenant(tenant_id) as db_session:
                 # Get the primary search settings
@@ -253,35 +175,11 @@ def check_for_indexing(self: Task, *, tenant_id: str | None) -> int | None:
                     )
                     if attempt_id:
                         task_logger.info(
-                            f"Connector indexing queued: "
-                            f"index_attempt={attempt_id} "
+                            f"Indexing queued: index_attempt={attempt_id} "
                             f"cc_pair={cc_pair.id} "
                             f"search_settings={search_settings_instance.id} "
                         )
                         tasks_created += 1
-
-        # Fail any index attempts in the DB that don't have fences
-        # This shouldn't ever happen!
-        with get_session_with_tenant(tenant_id) as db_session:
-            unfenced_attempt_ids = get_unfenced_index_attempt_ids(db_session, r)
-            for attempt_id in unfenced_attempt_ids:
-                lock_beat.reacquire()
-
-                attempt = get_index_attempt(db_session, attempt_id)
-                if not attempt:
-                    continue
-
-                failure_reason = (
-                    f"Unfenced index attempt found in DB: "
-                    f"index_attempt={attempt.id} "
-                    f"cc_pair={attempt.connector_credential_pair_id} "
-                    f"search_settings={attempt.search_settings_id}"
-                )
-                task_logger.error(failure_reason)
-                mark_attempt_failed(
-                    attempt.id, db_session, failure_reason=failure_reason
-                )
-
     except SoftTimeLimitExceeded:
         task_logger.info(
             "Soft time limit exceeded, task is being terminated gracefully."
@@ -291,11 +189,6 @@ def check_for_indexing(self: Task, *, tenant_id: str | None) -> int | None:
     finally:
         if lock_beat.owned():
             lock_beat.release()
-        else:
-            task_logger.error(
-                "check_for_indexing - Lock not owned on completion: "
-                f"tenant={tenant_id}"
-            )
 
     return tasks_created
 
@@ -400,11 +293,10 @@ def try_creating_indexing_task(
     """
 
     LOCK_TIMEOUT = 30
-    index_attempt_id: int | None = None
 
     # we need to serialize any attempt to trigger indexing since it can be triggered
     # either via celery beat or manually (API call)
-    lock: RedisLock = r.lock(
+    lock = r.lock(
         DANSWER_REDIS_FUNCTION_LOCK_PREFIX + "try_creating_indexing_task",
         timeout=LOCK_TIMEOUT,
     )
@@ -433,7 +325,7 @@ def try_creating_indexing_task(
         redis_connector_index.generator_clear()
 
         # set a basic fence to start
-        payload = RedisConnectorIndexPayload(
+        payload = RedisConnectorIndexingFenceData(
             index_attempt_id=None,
             started=None,
             submitted=datetime.now(timezone.utc),
@@ -455,8 +347,6 @@ def try_creating_indexing_task(
 
         custom_task_id = redis_connector_index.generate_generator_task_id()
 
-        # when the task is sent, we have yet to finish setting up the fence
-        # therefore, the task must contain code that blocks until the fence is ready
         result = celery_app.send_task(
             "connector_indexing_proxy_task",
             kwargs=dict(
@@ -476,17 +366,15 @@ def try_creating_indexing_task(
         payload.index_attempt_id = index_attempt_id
         payload.celery_task_id = result.id
         redis_connector_index.set_fence(payload)
+
     except Exception:
+        redis_connector_index.set_fence(payload)
         task_logger.exception(
-            f"try_creating_indexing_task - Unexpected exception: "
+            f"Unexpected exception: "
             f"tenant={tenant_id} "
             f"cc_pair={cc_pair.id} "
             f"search_settings={search_settings.id}"
         )
-
-        if index_attempt_id is not None:
-            delete_index_attempt(db_session, index_attempt_id)
-        redis_connector_index.set_fence(None)
         return None
     finally:
         if lock.owned():
@@ -504,7 +392,7 @@ def connector_indexing_proxy_task(
 ) -> None:
     """celery tasks are forked, but forking is unstable.  This proxies work to a spawned task."""
     task_logger.info(
-        f"Indexing watchdog - starting: attempt={index_attempt_id} "
+        f"Indexing proxy - starting: attempt={index_attempt_id} "
         f"tenant={tenant_id} "
         f"cc_pair={cc_pair_id} "
         f"search_settings={search_settings_id}"
@@ -512,7 +400,7 @@ def connector_indexing_proxy_task(
     client = SimpleJobClient()
 
     job = client.submit(
-        connector_indexing_task_wrapper,
+        connector_indexing_task,
         index_attempt_id,
         cc_pair_id,
         search_settings_id,
@@ -523,7 +411,7 @@ def connector_indexing_proxy_task(
 
     if not job:
         task_logger.info(
-            f"Indexing watchdog - spawn failed: attempt={index_attempt_id} "
+            f"Indexing proxy - spawn failed: attempt={index_attempt_id} "
             f"tenant={tenant_id} "
             f"cc_pair={cc_pair_id} "
             f"search_settings={search_settings_id}"
@@ -531,7 +419,7 @@ def connector_indexing_proxy_task(
         return
 
     task_logger.info(
-        f"Indexing watchdog - spawn succeeded: attempt={index_attempt_id} "
+        f"Indexing proxy - spawn succeeded: attempt={index_attempt_id} "
         f"tenant={tenant_id} "
         f"cc_pair={cc_pair_id} "
         f"search_settings={search_settings_id}"
@@ -555,7 +443,7 @@ def connector_indexing_proxy_task(
 
         if job.status == "error":
             task_logger.error(
-                f"Indexing watchdog - spawned task exceptioned: "
+                f"Indexing proxy - spawned task exceptioned: "
                 f"attempt={index_attempt_id} "
                 f"tenant={tenant_id} "
                 f"cc_pair={cc_pair_id} "
@@ -567,7 +455,7 @@ def connector_indexing_proxy_task(
         break
 
     task_logger.info(
-        f"Indexing watchdog - finished: attempt={index_attempt_id} "
+        f"Indexing proxy - finished: attempt={index_attempt_id} "
         f"tenant={tenant_id} "
         f"cc_pair={cc_pair_id} "
         f"search_settings={search_settings_id}"
@@ -575,38 +463,6 @@ def connector_indexing_proxy_task(
     return
 
 
-def connector_indexing_task_wrapper(
-    index_attempt_id: int,
-    cc_pair_id: int,
-    search_settings_id: int,
-    tenant_id: str | None,
-    is_ee: bool,
-) -> int | None:
-    """Just wraps connector_indexing_task so we can log any exceptions before
-    re-raising it."""
-    result: int | None = None
-
-    try:
-        result = connector_indexing_task(
-            index_attempt_id,
-            cc_pair_id,
-            search_settings_id,
-            tenant_id,
-            is_ee,
-        )
-    except:
-        logger.exception(
-            f"connector_indexing_task exceptioned: "
-            f"tenant={tenant_id} "
-            f"index_attempt={index_attempt_id} "
-            f"cc_pair={cc_pair_id} "
-            f"search_settings={search_settings_id}"
-        )
-        raise
-
-    return result
-
-
 def connector_indexing_task(
     index_attempt_id: int,
     cc_pair_id: int,
@@ -643,8 +499,7 @@ def connector_indexing_task(
         logger.debug("Sentry DSN not provided, skipping Sentry initialization")
 
     logger.info(
-        f"Indexing spawned task starting: "
-        f"attempt={index_attempt_id} "
+        f"Indexing spawned task starting: attempt={index_attempt_id} "
         f"tenant={tenant_id} "
         f"cc_pair={cc_pair_id} "
         f"search_settings={search_settings_id}"
@@ -661,7 +516,6 @@ def connector_indexing_task(
     if redis_connector.delete.fenced:
         raise RuntimeError(
             f"Indexing will not start because connector deletion is in progress: "
-            f"attempt={index_attempt_id} "
             f"cc_pair={cc_pair_id} "
             f"fence={redis_connector.delete.fence_key}"
         )
@@ -669,18 +523,18 @@ def connector_indexing_task(
     if redis_connector.stop.fenced:
         raise RuntimeError(
             f"Indexing will not start because a connector stop signal was detected: "
-            f"attempt={index_attempt_id} "
             f"cc_pair={cc_pair_id} "
             f"fence={redis_connector.stop.fence_key}"
         )
 
     while True:
-        if not redis_connector_index.fenced:  # The fence must exist
+        # wait for the fence to come up
+        if not redis_connector_index.fenced:
             raise ValueError(
                 f"connector_indexing_task - fence not found: fence={redis_connector_index.fence_key}"
             )
 
-        payload = redis_connector_index.payload  # The payload must exist
+        payload = redis_connector_index.payload
         if not payload:
             raise ValueError("connector_indexing_task: payload invalid or not found")
 
@@ -703,7 +557,7 @@ def connector_indexing_task(
         )
         break
 
-    lock: RedisLock = r.lock(
+    lock = r.lock(
         redis_connector_index.generator_lock_key,
         timeout=CELERY_INDEXING_LOCK_TIMEOUT,
     )
@@ -712,7 +566,7 @@ def connector_indexing_task(
     if not acquired:
         logger.warning(
             f"Indexing task already running, exiting...: "
-            f"index_attempt={index_attempt_id} cc_pair={cc_pair_id} search_settings={search_settings_id}"
+            f"cc_pair={cc_pair_id} search_settings={search_settings_id}"
         )
         return None
 
@@ -747,7 +601,7 @@ def connector_indexing_task(
                 )
 
         # define a callback class
-        callback = IndexingCallback(
+        callback = RunIndexingCallback(
             redis_connector.stop.fence_key,
             redis_connector_index.generator_progress_key,
             lock,

backend/danswer/background/celery/tasks/pruning/tasks.py

@@ -12,7 +12,7 @@ from sqlalchemy.orm import Session
 
 from danswer.background.celery.apps.app_base import task_logger
 from danswer.background.celery.celery_utils import extract_ids_from_runnable_connector
-from danswer.background.celery.tasks.indexing.tasks import IndexingCallback
+from danswer.background.celery.tasks.indexing.tasks import RunIndexingCallback
 from danswer.configs.app_configs import ALLOW_SIMULTANEOUS_PRUNING
 from danswer.configs.app_configs import JOB_TIMEOUT
 from danswer.configs.constants import CELERY_PRUNING_LOCK_TIMEOUT
@@ -39,14 +39,7 @@ logger = setup_logger()
 
 
 def _is_pruning_due(cc_pair: ConnectorCredentialPair) -> bool:
-    """Returns boolean indicating if pruning is due.
-
-    Next pruning time is calculated as a delta from the last successful prune, or the
-    last successful indexing if pruning has never succeeded.
-
-    TODO(rkuo): consider whether we should allow pruning to be immediately rescheduled
-    if pruning fails (which is what it does now). A backoff could be reasonable.
-    """
+    """Returns boolean indicating if pruning is due."""
 
     # skip pruning if no prune frequency is set
     # pruning can still be forced via the API which will run a pruning task directly
@@ -232,8 +225,6 @@ def connector_pruning_generator_task(
     pruning_ctx_dict["request_id"] = self.request.id
     pruning_ctx.set(pruning_ctx_dict)
 
-    task_logger.info(f"Pruning generator starting: cc_pair={cc_pair_id}")
-
     redis_connector = RedisConnector(tenant_id, cc_pair_id)
 
     r = get_redis_client(tenant_id=tenant_id)
@@ -264,11 +255,6 @@ def connector_pruning_generator_task(
                 )
                 return
 
-            task_logger.info(
-                f"Pruning generator running connector: "
-                f"cc_pair={cc_pair_id} "
-                f"connector_source={cc_pair.connector.source}"
-            )
             runnable_connector = instantiate_connector(
                 db_session,
                 cc_pair.connector.source,
@@ -277,13 +263,12 @@ def connector_pruning_generator_task(
                 cc_pair.credential,
             )
 
-            callback = IndexingCallback(
+            callback = RunIndexingCallback(
                 redis_connector.stop.fence_key,
                 redis_connector.prune.generator_progress_key,
                 lock,
                 r,
             )
-
             # a list of docs in the source
             all_connector_doc_ids: set[str] = extract_ids_from_runnable_connector(
                 runnable_connector, callback
@@ -305,8 +290,8 @@ def connector_pruning_generator_task(
             task_logger.info(
                 f"Pruning set collected: "
                 f"cc_pair={cc_pair_id} "
-                f"connector_source={cc_pair.connector.source} "
-                f"docs_to_remove={len(doc_ids_to_remove)}"
+                f"docs_to_remove={len(doc_ids_to_remove)} "
+                f"doc_source={cc_pair.connector.source}"
             )
 
             task_logger.info(
@@ -329,10 +314,10 @@ def connector_pruning_generator_task(
             f"Failed to run pruning: cc_pair={cc_pair_id} connector={connector_id}"
         )
 
-        redis_connector.prune.reset()
+        redis_connector.prune.generator_clear()
+        redis_connector.prune.taskset_clear()
+        redis_connector.prune.set_fence(False)
         raise e
     finally:
         if lock.owned():
             lock.release()
-
-        task_logger.info(f"Pruning generator finished: cc_pair={cc_pair_id}")

backend/danswer/background/celery/tasks/shared/tasks.py

@@ -177,17 +177,7 @@ def document_by_cc_pair_cleanup_task(
                 f"Max celery task retries reached. Marking doc as dirty for reconciliation: "
                 f"tenant={tenant_id} doc={document_id}"
             )
-            with get_session_with_tenant(tenant_id) as db_session:
-                # delete the cc pair relationship now and let reconciliation clean it up
-                # in vespa
-                delete_document_by_connector_credential_pair__no_commit(
-                    db_session=db_session,
-                    document_id=document_id,
-                    connector_credential_pair_identifier=ConnectorCredentialPairIdentifier(
-                        connector_id=connector_id,
-                        credential_id=credential_id,
-                    ),
-                )
+            with get_session_with_tenant(tenant_id):
                 mark_document_as_modified(document_id, db_session)
         return False
 

backend/danswer/background/celery/tasks/vespa/tasks.py

@@ -5,6 +5,7 @@ from http import HTTPStatus
 from typing import cast
 
 import httpx
+import redis
 from celery import Celery
 from celery import shared_task
 from celery import Task
@@ -12,7 +13,6 @@ from celery.exceptions import SoftTimeLimitExceeded
 from celery.result import AsyncResult
 from celery.states import READY_STATES
 from redis import Redis
-from redis.lock import Lock as RedisLock
 from sqlalchemy.orm import Session
 from tenacity import RetryError
 
@@ -46,10 +46,13 @@ from danswer.db.document_set import fetch_document_sets_for_document
 from danswer.db.document_set import get_document_set_by_id
 from danswer.db.document_set import mark_document_set_as_synced
 from danswer.db.engine import get_session_with_tenant
+from danswer.db.enums import IndexingStatus
 from danswer.db.index_attempt import delete_index_attempts
+from danswer.db.index_attempt import get_all_index_attempts_by_status
 from danswer.db.index_attempt import get_index_attempt
 from danswer.db.index_attempt import mark_attempt_failed
 from danswer.db.models import DocumentSet
+from danswer.db.models import IndexAttempt
 from danswer.document_index.document_index_utils import get_both_index_names
 from danswer.document_index.factory import get_default_document_index
 from danswer.document_index.interfaces import VespaDocumentFields
@@ -164,7 +167,7 @@ def try_generate_stale_document_sync_tasks(
     celery_app: Celery,
     db_session: Session,
     r: Redis,
-    lock_beat: RedisLock,
+    lock_beat: redis.lock.Lock,
     tenant_id: str | None,
 ) -> int | None:
     # the fence is up, do nothing
@@ -182,12 +185,7 @@ def try_generate_stale_document_sync_tasks(
         f"Stale documents found (at least {stale_doc_count}). Generating sync tasks by cc pair."
     )
 
-    task_logger.info(
-        "RedisConnector.generate_tasks starting by cc_pair. "
-        "Documents spanning multiple cc_pairs will only be synced once."
-    )
-
-    docs_to_skip: set[str] = set()
+    task_logger.info("RedisConnector.generate_tasks starting by cc_pair.")
 
     # rkuo: we could technically sync all stale docs in one big pass.
     # but I feel it's more understandable to group the docs by cc_pair
@@ -195,21 +193,22 @@ def try_generate_stale_document_sync_tasks(
     cc_pairs = get_connector_credential_pairs(db_session)
     for cc_pair in cc_pairs:
         rc = RedisConnectorCredentialPair(tenant_id, cc_pair.id)
-        rc.set_skip_docs(docs_to_skip)
-        result = rc.generate_tasks(celery_app, db_session, r, lock_beat, tenant_id)
+        tasks_generated = rc.generate_tasks(
+            celery_app, db_session, r, lock_beat, tenant_id
+        )
 
-        if result is None:
+        if tasks_generated is None:
             continue
 
-        if result[1] == 0:
+        if tasks_generated == 0:
             continue
 
         task_logger.info(
             f"RedisConnector.generate_tasks finished for single cc_pair. "
-            f"cc_pair={cc_pair.id} tasks_generated={result[0]} tasks_possible={result[1]}"
+            f"cc_pair_id={cc_pair.id} tasks_generated={tasks_generated}"
         )
 
-        total_tasks_generated += result[0]
+        total_tasks_generated += tasks_generated
 
     task_logger.info(
         f"RedisConnector.generate_tasks finished for all cc_pairs. total_tasks_generated={total_tasks_generated}"
@@ -224,7 +223,7 @@ def try_generate_document_set_sync_tasks(
     document_set_id: int,
     db_session: Session,
     r: Redis,
-    lock_beat: RedisLock,
+    lock_beat: redis.lock.Lock,
     tenant_id: str | None,
 ) -> int | None:
     lock_beat.reacquire()
@@ -252,11 +251,12 @@ def try_generate_document_set_sync_tasks(
     )
 
     # Add all documents that need to be updated into the queue
-    result = rds.generate_tasks(celery_app, db_session, r, lock_beat, tenant_id)
-    if result is None:
+    tasks_generated = rds.generate_tasks(
+        celery_app, db_session, r, lock_beat, tenant_id
+    )
+    if tasks_generated is None:
         return None
 
-    tasks_generated = result[0]
     # Currently we are allowing the sync to proceed with 0 tasks.
     # It's possible for sets/groups to be generated initially with no entries
     # and they still need to be marked as up to date.
@@ -265,7 +265,7 @@ def try_generate_document_set_sync_tasks(
 
     task_logger.info(
         f"RedisDocumentSet.generate_tasks finished. "
-        f"document_set={document_set.id} tasks_generated={tasks_generated}"
+        f"document_set_id={document_set.id} tasks_generated={tasks_generated}"
     )
 
     # set this only after all tasks have been added
@@ -278,7 +278,7 @@ def try_generate_user_group_sync_tasks(
     usergroup_id: int,
     db_session: Session,
     r: Redis,
-    lock_beat: RedisLock,
+    lock_beat: redis.lock.Lock,
     tenant_id: str | None,
 ) -> int | None:
     lock_beat.reacquire()
@@ -307,11 +307,12 @@ def try_generate_user_group_sync_tasks(
     task_logger.info(
         f"RedisUserGroup.generate_tasks starting. usergroup_id={usergroup.id}"
     )
-    result = rug.generate_tasks(celery_app, db_session, r, lock_beat, tenant_id)
-    if result is None:
+    tasks_generated = rug.generate_tasks(
+        celery_app, db_session, r, lock_beat, tenant_id
+    )
+    if tasks_generated is None:
         return None
 
-    tasks_generated = result[0]
     # Currently we are allowing the sync to proceed with 0 tasks.
     # It's possible for sets/groups to be generated initially with no entries
     # and they still need to be marked as up to date.
@@ -320,7 +321,7 @@ def try_generate_user_group_sync_tasks(
 
     task_logger.info(
         f"RedisUserGroup.generate_tasks finished. "
-        f"usergroup={usergroup.id} tasks_generated={tasks_generated}"
+        f"usergroup_id={usergroup.id} tasks_generated={tasks_generated}"
     )
 
     # set this only after all tasks have been added
@@ -440,22 +441,11 @@ def monitor_connector_deletion_taskset(
                 db_session, cc_pair.connector_id, cc_pair.credential_id
             )
             if len(doc_ids) > 0:
-                # NOTE(rkuo): if this happens, documents somehow got added while
-                # deletion was in progress. Likely a bug gating off pruning and indexing
-                # work before deletion starts.
+                # if this happens, documents somehow got added while deletion was in progress. Likely a bug
+                # gating off pruning and indexing work before deletion starts
                 task_logger.warning(
-                    "Connector deletion - documents still found after taskset completion. "
-                    "Clearing the current deletion attempt and allowing deletion to restart: "
-                    f"cc_pair={cc_pair_id} "
-                    f"docs_deleted={fence_data.num_tasks} "
-                    f"docs_remaining={len(doc_ids)}"
-                )
-
-                # We don't want to waive off why we get into this state, but resetting
-                # our attempt and letting the deletion restart is a good way to recover
-                redis_connector.delete.reset()
-                raise RuntimeError(
-                    "Connector deletion - documents still found after taskset completion"
+                    f"Connector deletion - documents still found after taskset completion: "
+                    f"cc_pair={cc_pair_id} num={len(doc_ids)}"
                 )
 
             # clean up the rest of the related Postgres entities
@@ -519,7 +509,8 @@ def monitor_connector_deletion_taskset(
         f"docs_deleted={fence_data.num_tasks}"
     )
 
-    redis_connector.delete.reset()
+    redis_connector.delete.taskset_clear()
+    redis_connector.delete.set_fence(None)
 
 
 def monitor_ccpair_pruning_taskset(
@@ -635,8 +626,8 @@ def monitor_ccpair_indexing_taskset(
     progress = redis_connector_index.get_progress()
     if progress is not None:
         task_logger.info(
-            f"Connector indexing progress: cc_pair={cc_pair_id} "
-            f"search_settings={search_settings_id} "
+            f"Connector indexing progress: cc_pair_id={cc_pair_id} "
+            f"search_settings_id={search_settings_id} "
             f"progress={progress} "
             f"elapsed_submitted={elapsed_submitted.total_seconds():.2f}"
         )
@@ -645,53 +636,39 @@ def monitor_ccpair_indexing_taskset(
         # the task is still setting up
         return
 
+    # Read result state BEFORE generator_complete_key to avoid a race condition
     # never use any blocking methods on the result from inside a task!
     result: AsyncResult = AsyncResult(payload.celery_task_id)
+    result_state = result.state
 
-    # inner/outer/inner double check pattern to avoid race conditions when checking for
-    # bad state
-
-    # inner = get_completion / generator_complete not signaled
-    # outer = result.state in READY state
     status_int = redis_connector_index.get_completion()
-    if status_int is None:  # inner signal not set ... possible error
-        result_state = result.state
-        if (
-            result_state in READY_STATES
-        ):  # outer signal in terminal state ... possible error
-            # Now double check!
-            if redis_connector_index.get_completion() is None:
-                # inner signal still not set (and cannot change when outer result_state is READY)
-                # Task is finished but generator complete isn't set.
-                # We have a problem! Worker may have crashed.
+    if status_int is None:
+        if result_state in READY_STATES:
+            # IF the task state is READY, THEN generator_complete should be set
+            # if it isn't, then the worker crashed
+            task_logger.info(
+                f"Connector indexing aborted: "
+                f"cc_pair_id={cc_pair_id} "
+                f"search_settings_id={search_settings_id} "
+                f"elapsed_submitted={elapsed_submitted.total_seconds():.2f}"
+            )
 
-                msg = (
-                    f"Connector indexing aborted or exceptioned: "
-                    f"attempt={payload.index_attempt_id} "
-                    f"celery_task={payload.celery_task_id} "
-                    f"result_state={result_state} "
-                    f"cc_pair={cc_pair_id} "
-                    f"search_settings={search_settings_id} "
-                    f"elapsed_submitted={elapsed_submitted.total_seconds():.2f}"
+            index_attempt = get_index_attempt(db_session, payload.index_attempt_id)
+            if index_attempt:
+                mark_attempt_failed(
+                    index_attempt_id=payload.index_attempt_id,
+                    db_session=db_session,
+                    failure_reason="Connector indexing aborted or exceptioned.",
                 )
-                task_logger.warning(msg)
 
-                index_attempt = get_index_attempt(db_session, payload.index_attempt_id)
-                if index_attempt:
-                    mark_attempt_failed(
-                        index_attempt_id=payload.index_attempt_id,
-                        db_session=db_session,
-                        failure_reason=msg,
-                    )
-
-                redis_connector_index.reset()
+            redis_connector_index.reset()
         return
 
     status_enum = HTTPStatus(status_int)
 
     task_logger.info(
-        f"Connector indexing finished: cc_pair={cc_pair_id} "
-        f"search_settings={search_settings_id} "
+        f"Connector indexing finished: cc_pair_id={cc_pair_id} "
+        f"search_settings_id={search_settings_id} "
         f"status={status_enum.name} "
         f"elapsed_submitted={elapsed_submitted.total_seconds():.2f}"
     )
@@ -712,7 +689,7 @@ def monitor_vespa_sync(self: Task, tenant_id: str | None) -> bool:
     """
     r = get_redis_client(tenant_id=tenant_id)
 
-    lock_beat: RedisLock = r.lock(
+    lock_beat: redis.lock.Lock = r.lock(
         DanswerRedisLocks.MONITOR_VESPA_SYNC_BEAT_LOCK,
         timeout=CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT,
     )
@@ -750,6 +727,34 @@ def monitor_vespa_sync(self: Task, tenant_id: str | None) -> bool:
             f"permissions_sync={n_permissions_sync} "
         )
 
+        # do some cleanup before clearing fences
+        # check the db for any outstanding index attempts
+        with get_session_with_tenant(tenant_id) as db_session:
+            attempts: list[IndexAttempt] = []
+            attempts.extend(
+                get_all_index_attempts_by_status(IndexingStatus.NOT_STARTED, db_session)
+            )
+            attempts.extend(
+                get_all_index_attempts_by_status(IndexingStatus.IN_PROGRESS, db_session)
+            )
+
+            for attempt in attempts:
+                # if attempts exist in the db but we don't detect them in redis, mark them as failed
+                fence_key = RedisConnectorIndex.fence_key_with_ids(
+                    attempt.connector_credential_pair_id, attempt.search_settings_id
+                )
+                if not r.exists(fence_key):
+                    failure_reason = (
+                        f"Unknown index attempt. Might be left over from a process restart: "
+                        f"index_attempt={attempt.id} "
+                        f"cc_pair={attempt.connector_credential_pair_id} "
+                        f"search_settings={attempt.search_settings_id}"
+                    )
+                    task_logger.warning(failure_reason)
+                    mark_attempt_failed(
+                        attempt.id, db_session, failure_reason=failure_reason
+                    )
+
         lock_beat.reacquire()
         if r.exists(RedisConnectorCredentialPair.get_fence_key()):
             monitor_connector_taskset(r)

backend/danswer/background/indexing/run_indexing.py

@@ -1,5 +1,7 @@
 import time
 import traceback
+from abc import ABC
+from abc import abstractmethod
 from datetime import datetime
 from datetime import timedelta
 from datetime import timezone
@@ -29,7 +31,7 @@ from danswer.db.models import IndexingStatus
 from danswer.db.models import IndexModelStatus
 from danswer.document_index.factory import get_default_document_index
 from danswer.indexing.embedder import DefaultIndexingEmbedder
-from danswer.indexing.indexing_heartbeat import IndexingHeartbeatInterface
+from danswer.indexing.indexing_heartbeat import IndexingHeartbeat
 from danswer.indexing.indexing_pipeline import build_indexing_pipeline
 from danswer.utils.logger import setup_logger
 from danswer.utils.logger import TaskAttemptSingleton
@@ -40,6 +42,19 @@ logger = setup_logger()
 INDEXING_TRACER_NUM_PRINT_ENTRIES = 5
 
 
+class RunIndexingCallbackInterface(ABC):
+    """Defines a callback interface to be passed to
+    to run_indexing_entrypoint."""
+
+    @abstractmethod
+    def should_stop(self) -> bool:
+        """Signal to stop the looping function in flight."""
+
+    @abstractmethod
+    def progress(self, amount: int) -> None:
+        """Send progress updates to the caller."""
+
+
 def _get_connector_runner(
     db_session: Session,
     attempt: IndexAttempt,
@@ -91,7 +106,7 @@ def _run_indexing(
     db_session: Session,
     index_attempt: IndexAttempt,
     tenant_id: str | None,
-    callback: IndexingHeartbeatInterface | None = None,
+    callback: RunIndexingCallbackInterface | None = None,
 ) -> None:
     """
     1. Get documents which are either new or updated from specified application
@@ -123,7 +138,13 @@ def _run_indexing(
 
     embedding_model = DefaultIndexingEmbedder.from_db_search_settings(
         search_settings=search_settings,
-        callback=callback,
+        heartbeat=IndexingHeartbeat(
+            index_attempt_id=index_attempt.id,
+            db_session=db_session,
+            # let the world know we're still making progress after
+            # every 10 batches
+            freq=10,
+        ),
     )
 
     indexing_pipeline = build_indexing_pipeline(
@@ -136,7 +157,6 @@ def _run_indexing(
         ),
         db_session=db_session,
         tenant_id=tenant_id,
-        callback=callback,
     )
 
     db_cc_pair = index_attempt.connector_credential_pair
@@ -208,9 +228,7 @@ def _run_indexing(
                 # contents still need to be initially pulled.
                 if callback:
                     if callback.should_stop():
-                        raise RuntimeError(
-                            "_run_indexing: Connector stop signal detected"
-                        )
+                        raise RuntimeError("Connector stop signal detected")
 
                 # TODO: should we move this into the above callback instead?
                 db_session.refresh(db_cc_pair)
@@ -271,7 +289,7 @@ def _run_indexing(
                 db_session.commit()
 
                 if callback:
-                    callback.progress("_run_indexing", len(doc_batch))
+                    callback.progress(len(doc_batch))
 
                 # This new value is updated every batch, so UI can refresh per batch update
                 update_docs_indexed(
@@ -401,7 +419,7 @@ def run_indexing_entrypoint(
     tenant_id: str | None,
     connector_credential_pair_id: int,
     is_ee: bool = False,
-    callback: IndexingHeartbeatInterface | None = None,
+    callback: RunIndexingCallbackInterface | None = None,
 ) -> None:
     try:
         if is_ee:
@@ -415,13 +433,11 @@ def run_indexing_entrypoint(
         with get_session_with_tenant(tenant_id) as db_session:
             attempt = transition_attempt_to_in_progress(index_attempt_id, db_session)
 
-            tenant_str = ""
-            if tenant_id is not None:
-                tenant_str = f" for tenant {tenant_id}"
-
             logger.info(
-                f"Indexing starting{tenant_str}: "
-                f"connector='{attempt.connector_credential_pair.connector.name}' "
+                f"Indexing starting for tenant {tenant_id}: "
+                if tenant_id is not None
+                else ""
+                + f"connector='{attempt.connector_credential_pair.connector.name}' "
                 f"config='{attempt.connector_credential_pair.connector.connector_specific_config}' "
                 f"credentials='{attempt.connector_credential_pair.connector_id}'"
             )
@@ -429,8 +445,10 @@ def run_indexing_entrypoint(
             _run_indexing(db_session, attempt, tenant_id, callback)
 
             logger.info(
-                f"Indexing finished{tenant_str}: "
-                f"connector='{attempt.connector_credential_pair.connector.name}' "
+                f"Indexing finished for tenant {tenant_id}: "
+                if tenant_id is not None
+                else ""
+                + f"connector='{attempt.connector_credential_pair.connector.name}' "
                 f"config='{attempt.connector_credential_pair.connector.connector_specific_config}' "
                 f"credentials='{attempt.connector_credential_pair.connector_id}'"
             )

backend/danswer/chat/process_message.py

@@ -112,7 +112,6 @@ from danswer.tools.tool_implementations.search.search_tool import (
 )
 from danswer.tools.tool_runner import ToolCallFinalResult
 from danswer.utils.logger import setup_logger
-from danswer.utils.long_term_log import LongTermLogger
 from danswer.utils.timing import log_generator_function_time
 
 logger = setup_logger()
@@ -317,11 +316,6 @@ def stream_chat_message_objects(
         retrieval_options = new_msg_req.retrieval_options
         alternate_assistant_id = new_msg_req.alternate_assistant_id
 
-        # permanent "log" store, used primarily for debugging
-        long_term_logger = LongTermLogger(
-            metadata={"user_id": str(user_id), "chat_session_id": str(chat_session_id)}
-        )
-
         # use alternate persona if alternative assistant id is passed in
         if alternate_assistant_id is not None:
             persona = get_persona_by_id(
@@ -347,7 +341,6 @@ def stream_chat_message_objects(
                 persona=persona,
                 llm_override=new_msg_req.llm_override or chat_session.llm_override,
                 additional_headers=litellm_additional_headers,
-                long_term_logger=long_term_logger,
             )
         except GenAIDisabledException:
             raise RuntimeError("LLM is disabled. Can't use chat flow without LLM.")

backend/danswer/configs/constants.py

@@ -60,6 +60,7 @@ KV_GMAIL_CRED_KEY = "gmail_app_credential"
 KV_GMAIL_SERVICE_ACCOUNT_KEY = "gmail_service_account_key"
 KV_GOOGLE_DRIVE_CRED_KEY = "google_drive_app_credential"
 KV_GOOGLE_DRIVE_SERVICE_ACCOUNT_KEY = "google_drive_service_account_key"
+KV_SLACK_BOT_TOKENS_CONFIG_KEY = "slack_bot_tokens_config_key"
 KV_GEN_AI_KEY_CHECK_TIME = "genai_api_key_last_check_time"
 KV_SETTINGS_KEY = "danswer_settings"
 KV_CUSTOMER_UUID_KEY = "customer_uuid"
@@ -73,7 +74,7 @@ CELERY_PRIMARY_WORKER_LOCK_TIMEOUT = 120
 
 # needs to be long enough to cover the maximum time it takes to download an object
 # if we can get callbacks as object bytes download, we could lower this a lot.
-CELERY_INDEXING_LOCK_TIMEOUT = 3 * 60 * 60  # 60 min
+CELERY_INDEXING_LOCK_TIMEOUT = 60 * 60  # 60 min
 
 # needs to be long enough to cover the maximum time it takes to download an object
 # if we can get callbacks as object bytes download, we could lower this a lot.

backend/danswer/connectors/blob/connector.py

@@ -5,9 +5,9 @@ from io import BytesIO
 from typing import Any
 from typing import Optional
 
-import boto3  # type: ignore
-from botocore.client import Config  # type: ignore
-from mypy_boto3_s3 import S3Client  # type: ignore
+import boto3
+from botocore.client import Config
+from mypy_boto3_s3 import S3Client
 
 from danswer.configs.app_configs import INDEX_BATCH_SIZE
 from danswer.configs.constants import BlobType

backend/danswer/connectors/confluence/connector.py

@@ -7,9 +7,9 @@ from danswer.configs.app_configs import CONFLUENCE_CONNECTOR_LABELS_TO_SKIP
 from danswer.configs.app_configs import CONTINUE_ON_CONNECTOR_FAILURE
 from danswer.configs.app_configs import INDEX_BATCH_SIZE
 from danswer.configs.constants import DocumentSource
-from danswer.connectors.confluence.onyx_confluence import build_confluence_client
 from danswer.connectors.confluence.onyx_confluence import OnyxConfluence
 from danswer.connectors.confluence.utils import attachment_to_content
+from danswer.connectors.confluence.utils import build_confluence_client
 from danswer.connectors.confluence.utils import build_confluence_document_id
 from danswer.connectors.confluence.utils import datetime_from_string
 from danswer.connectors.confluence.utils import extract_text_from_confluence_html
@@ -70,7 +70,7 @@ class ConfluenceConnector(LoadConnector, PollConnector, SlimConnector):
     ) -> None:
         self.batch_size = batch_size
         self.continue_on_failure = continue_on_failure
-        self._confluence_client: OnyxConfluence | None = None
+        self.confluence_client: OnyxConfluence | None = None
         self.is_cloud = is_cloud
 
         # Remove trailing slash from wiki_base if present
@@ -81,15 +81,15 @@ class ConfluenceConnector(LoadConnector, PollConnector, SlimConnector):
         if cql_query:
             # if a cql_query is provided, we will use it to fetch the pages
             cql_page_query = cql_query
+        elif space:
+            # if no cql_query is provided, we will use the space to fetch the pages
+            cql_page_query += f" and space='{quote(space)}'"
         elif page_id:
-            # if a cql_query is not provided, we will use the page_id to fetch the page
             if index_recursively:
                 cql_page_query += f" and ancestor='{page_id}'"
             else:
+                # if neither a space nor a cql_query is provided, we will use the page_id to fetch the page
                 cql_page_query += f" and id='{page_id}'"
-        elif space:
-            # if no cql_query or page_id is provided, we will use the space to fetch the pages
-            cql_page_query += f" and space='{quote(space)}'"
 
         self.cql_page_query = cql_page_query
         self.cql_time_filter = ""
@@ -97,44 +97,39 @@ class ConfluenceConnector(LoadConnector, PollConnector, SlimConnector):
         self.cql_label_filter = ""
         if labels_to_skip:
             labels_to_skip = list(set(labels_to_skip))
-            comma_separated_labels = ",".join(
-                f"'{quote(label)}'" for label in labels_to_skip
-            )
+            comma_separated_labels = ",".join(f"'{label}'" for label in labels_to_skip)
             self.cql_label_filter = f" and label not in ({comma_separated_labels})"
 
-    @property
-    def confluence_client(self) -> OnyxConfluence:
-        if self._confluence_client is None:
-            raise ConnectorMissingCredentialError("Confluence")
-        return self._confluence_client
-
     def load_credentials(self, credentials: dict[str, Any]) -> dict[str, Any] | None:
         # see https://github.com/atlassian-api/atlassian-python-api/blob/master/atlassian/rest_client.py
         # for a list of other hidden constructor args
-        self._confluence_client = build_confluence_client(
-            credentials=credentials,
+        self.confluence_client = build_confluence_client(
+            credentials_json=credentials,
             is_cloud=self.is_cloud,
             wiki_base=self.wiki_base,
         )
         return None
 
     def _get_comment_string_for_page_id(self, page_id: str) -> str:
+        if self.confluence_client is None:
+            raise ConnectorMissingCredentialError("Confluence")
+
         comment_string = ""
 
         comment_cql = f"type=comment and container='{page_id}'"
         comment_cql += self.cql_label_filter
 
         expand = ",".join(_COMMENT_EXPANSION_FIELDS)
-        for comment in self.confluence_client.paginated_cql_retrieval(
+        for comments in self.confluence_client.paginated_cql_page_retrieval(
             cql=comment_cql,
             expand=expand,
         ):
-            comment_string += "\nComment:\n"
-            comment_string += extract_text_from_confluence_html(
-                confluence_client=self.confluence_client,
-                confluence_object=comment,
-                fetched_titles=set(),
-            )
+            for comment in comments:
+                comment_string += "\nComment:\n"
+                comment_string += extract_text_from_confluence_html(
+                    confluence_client=self.confluence_client,
+                    confluence_object=comment,
+                )
 
         return comment_string
 
@@ -146,6 +141,9 @@ class ConfluenceConnector(LoadConnector, PollConnector, SlimConnector):
         If its a page, it extracts the text, adds the comments for the document text.
         If its an attachment, it just downloads the attachment and converts that into a document.
         """
+        if self.confluence_client is None:
+            raise ConnectorMissingCredentialError("Confluence")
+
         # The url and the id are the same
         object_url = build_confluence_document_id(
             self.wiki_base, confluence_object["_links"]["webui"], self.is_cloud
@@ -155,19 +153,16 @@ class ConfluenceConnector(LoadConnector, PollConnector, SlimConnector):
         # Extract text from page
         if confluence_object["type"] == "page":
             object_text = extract_text_from_confluence_html(
-                confluence_client=self.confluence_client,
-                confluence_object=confluence_object,
-                fetched_titles={confluence_object.get("title", "")},
+                self.confluence_client, confluence_object
             )
             # Add comments to text
             object_text += self._get_comment_string_for_page_id(confluence_object["id"])
         elif confluence_object["type"] == "attachment":
             object_text = attachment_to_content(
-                confluence_client=self.confluence_client, attachment=confluence_object
+                self.confluence_client, confluence_object
             )
 
         if object_text is None:
-            # This only happens for attachments that are not parseable
             return None
 
         # Get space name
@@ -198,39 +193,44 @@ class ConfluenceConnector(LoadConnector, PollConnector, SlimConnector):
         )
 
     def _fetch_document_batches(self) -> GenerateDocumentsOutput:
+        if self.confluence_client is None:
+            raise ConnectorMissingCredentialError("Confluence")
+
         doc_batch: list[Document] = []
         confluence_page_ids: list[str] = []
 
         page_query = self.cql_page_query + self.cql_label_filter + self.cql_time_filter
         # Fetch pages as Documents
-        for page in self.confluence_client.paginated_cql_retrieval(
+        for page_batch in self.confluence_client.paginated_cql_page_retrieval(
             cql=page_query,
             expand=",".join(_PAGE_EXPANSION_FIELDS),
             limit=self.batch_size,
         ):
-            confluence_page_ids.append(page["id"])
-            doc = self._convert_object_to_document(page)
-            if doc is not None:
-                doc_batch.append(doc)
-            if len(doc_batch) >= self.batch_size:
-                yield doc_batch
-                doc_batch = []
+            for page in page_batch:
+                confluence_page_ids.append(page["id"])
+                doc = self._convert_object_to_document(page)
+                if doc is not None:
+                    doc_batch.append(doc)
+                if len(doc_batch) >= self.batch_size:
+                    yield doc_batch
+                    doc_batch = []
 
         # Fetch attachments as Documents
         for confluence_page_id in confluence_page_ids:
             attachment_cql = f"type=attachment and container='{confluence_page_id}'"
             attachment_cql += self.cql_label_filter
             # TODO: maybe should add time filter as well?
-            for attachment in self.confluence_client.paginated_cql_retrieval(
+            for attachments in self.confluence_client.paginated_cql_page_retrieval(
                 cql=attachment_cql,
                 expand=",".join(_ATTACHMENT_EXPANSION_FIELDS),
             ):
-                doc = self._convert_object_to_document(attachment)
-                if doc is not None:
-                    doc_batch.append(doc)
-                if len(doc_batch) >= self.batch_size:
-                    yield doc_batch
-                    doc_batch = []
+                for attachment in attachments:
+                    doc = self._convert_object_to_document(attachment)
+                    if doc is not None:
+                        doc_batch.append(doc)
+                    if len(doc_batch) >= self.batch_size:
+                        yield doc_batch
+                        doc_batch = []
 
         if doc_batch:
             yield doc_batch
@@ -255,47 +255,52 @@ class ConfluenceConnector(LoadConnector, PollConnector, SlimConnector):
         start: SecondsSinceUnixEpoch | None = None,
         end: SecondsSinceUnixEpoch | None = None,
     ) -> GenerateSlimDocumentOutput:
+        if self.confluence_client is None:
+            raise ConnectorMissingCredentialError("Confluence")
+
         doc_metadata_list: list[SlimDocument] = []
 
         restrictions_expand = ",".join(_RESTRICTIONS_EXPANSION_FIELDS)
 
         page_query = self.cql_page_query + self.cql_label_filter
-        for page in self.confluence_client.cql_paginate_all_expansions(
+        for pages in self.confluence_client.cql_paginate_all_expansions(
             cql=page_query,
             expand=restrictions_expand,
         ):
-            # If the page has restrictions, add them to the perm_sync_data
-            # These will be used by doc_sync.py to sync permissions
-            perm_sync_data = {
-                "restrictions": page.get("restrictions", {}),
-                "space_key": page.get("space", {}).get("key"),
-            }
+            for page in pages:
+                # If the page has restrictions, add them to the perm_sync_data
+                # These will be used by doc_sync.py to sync permissions
+                perm_sync_data = {
+                    "restrictions": page.get("restrictions", {}),
+                    "space_key": page.get("space", {}).get("key"),
+                }
 
-            doc_metadata_list.append(
-                SlimDocument(
-                    id=build_confluence_document_id(
-                        self.wiki_base,
-                        page["_links"]["webui"],
-                        self.is_cloud,
-                    ),
-                    perm_sync_data=perm_sync_data,
-                )
-            )
-            attachment_cql = f"type=attachment and container='{page['id']}'"
-            attachment_cql += self.cql_label_filter
-            for attachment in self.confluence_client.cql_paginate_all_expansions(
-                cql=attachment_cql,
-                expand=restrictions_expand,
-            ):
                 doc_metadata_list.append(
                     SlimDocument(
                         id=build_confluence_document_id(
                             self.wiki_base,
-                            attachment["_links"]["webui"],
+                            page["_links"]["webui"],
                             self.is_cloud,
                         ),
                         perm_sync_data=perm_sync_data,
                     )
                 )
-            yield doc_metadata_list
-            doc_metadata_list = []
+                attachment_cql = f"type=attachment and container='{page['id']}'"
+                attachment_cql += self.cql_label_filter
+                for attachments in self.confluence_client.cql_paginate_all_expansions(
+                    cql=attachment_cql,
+                    expand=restrictions_expand,
+                ):
+                    for attachment in attachments:
+                        doc_metadata_list.append(
+                            SlimDocument(
+                                id=build_confluence_document_id(
+                                    self.wiki_base,
+                                    attachment["_links"]["webui"],
+                                    self.is_cloud,
+                                ),
+                                perm_sync_data=perm_sync_data,
+                            )
+                        )
+                yield doc_metadata_list
+                doc_metadata_list = []

backend/danswer/connectors/confluence/onyx_confluence.py

@@ -20,10 +20,6 @@ F = TypeVar("F", bound=Callable[..., Any])
 
 RATE_LIMIT_MESSAGE_LOWERCASE = "Rate limit exceeded".lower()
 
-# https://jira.atlassian.com/browse/CONFCLOUD-76433
-_PROBLEMATIC_EXPANSIONS = "body.storage.value"
-_REPLACEMENT_EXPANSIONS = "body.view.value"
-
 
 class ConfluenceRateLimitError(Exception):
     pass
@@ -84,7 +80,7 @@ def handle_confluence_rate_limit(confluence_call: F) -> F:
     def wrapped_call(*args: list[Any], **kwargs: Any) -> Any:
         MAX_RETRIES = 5
 
-        TIMEOUT = 600
+        TIMEOUT = 3600
         timeout_at = time.monotonic() + TIMEOUT
 
         for attempt in range(MAX_RETRIES):
@@ -99,10 +95,6 @@ def handle_confluence_rate_limit(confluence_call: F) -> F:
                 return confluence_call(*args, **kwargs)
             except HTTPError as e:
                 delay_until = _handle_http_error(e, attempt)
-                logger.warning(
-                    f"HTTPError in confluence call. "
-                    f"Retrying in {delay_until} seconds..."
-                )
                 while time.monotonic() < delay_until:
                     # in the future, check a signal here to exit
                     time.sleep(1)
@@ -149,7 +141,7 @@ class OnyxConfluence(Confluence):
 
     def _paginate_url(
         self, url_suffix: str, limit: int | None = None
-    ) -> Iterator[dict[str, Any]]:
+    ) -> Iterator[list[dict[str, Any]]]:
         """
         This will paginate through the top level query.
         """
@@ -161,43 +153,46 @@ class OnyxConfluence(Confluence):
 
         while url_suffix:
             try:
-                logger.debug(f"Making confluence call to {url_suffix}")
                 next_response = self.get(url_suffix)
             except Exception as e:
-                logger.warning(f"Error in confluence call to {url_suffix}")
-
-                # If the problematic expansion is in the url, replace it
-                # with the replacement expansion and try again
-                # If that fails, raise the error
-                if _PROBLEMATIC_EXPANSIONS not in url_suffix:
-                    logger.exception(f"Error in confluence call to {url_suffix}")
-                    raise e
-                logger.warning(
-                    f"Replacing {_PROBLEMATIC_EXPANSIONS} with {_REPLACEMENT_EXPANSIONS}"
-                    " and trying again."
-                )
-                url_suffix = url_suffix.replace(
-                    _PROBLEMATIC_EXPANSIONS,
-                    _REPLACEMENT_EXPANSIONS,
-                )
-                continue
-
-            # yield the results individually
-            yield from next_response.get("results", [])
-
+                logger.exception("Error in danswer_cql: \n")
+                raise e
+            yield next_response.get("results", [])
             url_suffix = next_response.get("_links", {}).get("next")
 
-    def paginated_cql_retrieval(
+    def paginated_groups_retrieval(
+        self,
+        limit: int | None = None,
+    ) -> Iterator[list[dict[str, Any]]]:
+        return self._paginate_url("rest/api/group", limit)
+
+    def paginated_group_members_retrieval(
+        self,
+        group_name: str,
+        limit: int | None = None,
+    ) -> Iterator[list[dict[str, Any]]]:
+        group_name = quote(group_name)
+        return self._paginate_url(f"rest/api/group/{group_name}/member", limit)
+
+    def paginated_cql_user_retrieval(
         self,
         cql: str,
         expand: str | None = None,
         limit: int | None = None,
-    ) -> Iterator[dict[str, Any]]:
-        """
-        The content/search endpoint can be used to fetch pages, attachments, and comments.
-        """
+    ) -> Iterator[list[dict[str, Any]]]:
         expand_string = f"&expand={expand}" if expand else ""
-        yield from self._paginate_url(
+        return self._paginate_url(
+            f"rest/api/search/user?cql={cql}{expand_string}", limit
+        )
+
+    def paginated_cql_page_retrieval(
+        self,
+        cql: str,
+        expand: str | None = None,
+        limit: int | None = None,
+    ) -> Iterator[list[dict[str, Any]]]:
+        expand_string = f"&expand={expand}" if expand else ""
+        return self._paginate_url(
             f"rest/api/content/search?cql={cql}{expand_string}", limit
         )
 
@@ -206,7 +201,7 @@ class OnyxConfluence(Confluence):
         cql: str,
         expand: str | None = None,
         limit: int | None = None,
-    ) -> Iterator[dict[str, Any]]:
+    ) -> Iterator[list[dict[str, Any]]]:
         """
         This function will paginate through the top level query first, then
         paginate through all of the expansions.
@@ -226,110 +221,6 @@ class OnyxConfluence(Confluence):
                 for item in data:
                     _traverse_and_update(item)
 
-        for confluence_object in self.paginated_cql_retrieval(cql, expand, limit):
-            _traverse_and_update(confluence_object)
-            yield confluence_object
-
-    def paginated_cql_user_retrieval(
-        self,
-        expand: str | None = None,
-        limit: int | None = None,
-    ) -> Iterator[dict[str, Any]]:
-        """
-        The search/user endpoint can be used to fetch users.
-        It's a seperate endpoint from the content/search endpoint used only for users.
-        Otherwise it's very similar to the content/search endpoint.
-        """
-        cql = "type=user"
-        url = "rest/api/search/user" if self.cloud else "rest/api/search"
-        expand_string = f"&expand={expand}" if expand else ""
-        url += f"?cql={cql}{expand_string}"
-        yield from self._paginate_url(url, limit)
-
-    def paginated_groups_by_user_retrieval(
-        self,
-        user: dict[str, Any],
-        limit: int | None = None,
-    ) -> Iterator[dict[str, Any]]:
-        """
-        This is not an SQL like query.
-        It's a confluence specific endpoint that can be used to fetch groups.
-        """
-        user_field = "accountId" if self.cloud else "key"
-        user_value = user["accountId"] if self.cloud else user["userKey"]
-        # Server uses userKey (but calls it key during the API call), Cloud uses accountId
-        user_query = f"{user_field}={quote(user_value)}"
-
-        url = f"rest/api/user/memberof?{user_query}"
-        yield from self._paginate_url(url, limit)
-
-    def paginated_groups_retrieval(
-        self,
-        limit: int | None = None,
-    ) -> Iterator[dict[str, Any]]:
-        """
-        This is not an SQL like query.
-        It's a confluence specific endpoint that can be used to fetch groups.
-        """
-        yield from self._paginate_url("rest/api/group", limit)
-
-    def paginated_group_members_retrieval(
-        self,
-        group_name: str,
-        limit: int | None = None,
-    ) -> Iterator[dict[str, Any]]:
-        """
-        This is not an SQL like query.
-        It's a confluence specific endpoint that can be used to fetch the members of a group.
-        THIS DOESN'T WORK FOR SERVER because it breaks when there is a slash in the group name.
-        E.g. neither "test/group" nor "test%2Fgroup" works for confluence.
-        """
-        group_name = quote(group_name)
-        yield from self._paginate_url(f"rest/api/group/{group_name}/member", limit)
-
-
-def _validate_connector_configuration(
-    credentials: dict[str, Any],
-    is_cloud: bool,
-    wiki_base: str,
-) -> None:
-    # test connection with direct client, no retries
-    confluence_client_without_retries = Confluence(
-        api_version="cloud" if is_cloud else "latest",
-        url=wiki_base.rstrip("/"),
-        username=credentials["confluence_username"] if is_cloud else None,
-        password=credentials["confluence_access_token"] if is_cloud else None,
-        token=credentials["confluence_access_token"] if not is_cloud else None,
-    )
-    spaces = confluence_client_without_retries.get_all_spaces(limit=1)
-
-    if not spaces:
-        raise RuntimeError(
-            f"No spaces found at {wiki_base}! "
-            "Check your credentials and wiki_base and make sure "
-            "is_cloud is set correctly."
-        )
-
-
-def build_confluence_client(
-    credentials: dict[str, Any],
-    is_cloud: bool,
-    wiki_base: str,
-) -> OnyxConfluence:
-    _validate_connector_configuration(
-        credentials=credentials,
-        is_cloud=is_cloud,
-        wiki_base=wiki_base,
-    )
-    return OnyxConfluence(
-        api_version="cloud" if is_cloud else "latest",
-        # Remove trailing slash from wiki_base if present
-        url=wiki_base.rstrip("/"),
-        # passing in username causes issues for Confluence data center
-        username=credentials["confluence_username"] if is_cloud else None,
-        password=credentials["confluence_access_token"] if is_cloud else None,
-        token=credentials["confluence_access_token"] if not is_cloud else None,
-        backoff_and_retry=True,
-        max_backoff_retries=10,
-        max_backoff_seconds=60,
-    )
+        for results in self.paginated_cql_page_retrieval(cql, expand, limit):
+            _traverse_and_update(results)
+            yield results

backend/danswer/connectors/confluence/utils.py

@@ -2,7 +2,6 @@ import io
 from datetime import datetime
 from datetime import timezone
 from typing import Any
-from urllib.parse import quote
 
 import bs4
 
@@ -72,9 +71,7 @@ def _get_user(confluence_client: OnyxConfluence, user_id: str) -> str:
 
 
 def extract_text_from_confluence_html(
-    confluence_client: OnyxConfluence,
-    confluence_object: dict[str, Any],
-    fetched_titles: set[str],
+    confluence_client: OnyxConfluence, confluence_object: dict[str, Any]
 ) -> str:
     """Parse a Confluence html page and replace the 'user Id' by the real
         User Display Name
@@ -82,7 +79,7 @@ def extract_text_from_confluence_html(
     Args:
         confluence_object (dict): The confluence object as a dict
         confluence_client (Confluence): Confluence client
-        fetched_titles (set[str]): The titles of the pages that have already been fetched
+
     Returns:
         str: loaded and formated Confluence page
     """
@@ -104,72 +101,38 @@ def extract_text_from_confluence_html(
         # Include @ sign for tagging, more clear for LLM
         user.replaceWith("@" + _get_user(confluence_client, user_id))
 
-    for html_page_reference in soup.findAll("ac:structured-macro"):
-        # Here, we only want to process page within page macros
-        if html_page_reference.attrs.get("ac:name") != "include":
-            continue
-
-        page_data = html_page_reference.find("ri:page")
-        if not page_data:
-            logger.warning(
-                f"Skipping retrieval of {html_page_reference} because because page data is missing"
-            )
-            continue
-
-        page_title = page_data.attrs.get("ri:content-title")
-        if not page_title:
-            # only fetch pages that have a title
-            logger.warning(
-                f"Skipping retrieval of {html_page_reference} because it has no title"
-            )
-            continue
-
-        if page_title in fetched_titles:
-            # prevent recursive fetching of pages
-            logger.debug(f"Skipping {page_title} because it has already been fetched")
-            continue
-
-        fetched_titles.add(page_title)
-
+    for html_page_reference in soup.findAll("ri:page"):
         # Wrap this in a try-except because there are some pages that might not exist
         try:
-            page_query = f"type=page and title='{quote(page_title)}'"
+            page_title = html_page_reference.attrs["ri:content-title"]
+            if not page_title:
+                continue
+
+            page_query = f"type=page and title='{page_title}'"
 
             page_contents: dict[str, Any] | None = None
             # Confluence enforces title uniqueness, so we should only get one result here
-            for page in confluence_client.paginated_cql_retrieval(
+            for page_batch in confluence_client.paginated_cql_page_retrieval(
                 cql=page_query,
                 expand="body.storage.value",
                 limit=1,
             ):
-                page_contents = page
+                page_contents = page_batch[0]
                 break
-        except Exception as e:
+        except Exception:
             logger.warning(
-                f"Error getting page contents for object {confluence_object}: {e}"
+                f"Error getting page contents for object {confluence_object}"
             )
             continue
 
         if not page_contents:
             continue
-
         text_from_page = extract_text_from_confluence_html(
-            confluence_client=confluence_client,
-            confluence_object=page_contents,
-            fetched_titles=fetched_titles,
+            confluence_client, page_contents
         )
 
         html_page_reference.replaceWith(text_from_page)
 
-    for html_link_body in soup.findAll("ac:link-body"):
-        # This extracts the text from inline links in the page so they can be
-        # represented in the document text as plain text
-        try:
-            text_from_link = html_link_body.text
-            html_link_body.replaceWith(f"(LINK TEXT: {text_from_link})")
-        except Exception as e:
-            logger.warning(f"Error processing ac:link-body: {e}")
-
     return format_document_soup(soup)
 
 
@@ -269,3 +232,20 @@ def datetime_from_string(datetime_string: str) -> datetime:
         datetime_object = datetime_object.astimezone(timezone.utc)
 
     return datetime_object
+
+
+def build_confluence_client(
+    credentials_json: dict[str, Any], is_cloud: bool, wiki_base: str
+) -> OnyxConfluence:
+    return OnyxConfluence(
+        api_version="cloud" if is_cloud else "latest",
+        # Remove trailing slash from wiki_base if present
+        url=wiki_base.rstrip("/"),
+        # passing in username causes issues for Confluence data center
+        username=credentials_json["confluence_username"] if is_cloud else None,
+        password=credentials_json["confluence_access_token"] if is_cloud else None,
+        token=credentials_json["confluence_access_token"] if not is_cloud else None,
+        backoff_and_retry=True,
+        max_backoff_retries=60,
+        max_backoff_seconds=60,
+    )

backend/danswer/connectors/danswer_jira/connector.py

@@ -1,8 +1,8 @@
 import os
-from collections.abc import Iterable
 from datetime import datetime
 from datetime import timezone
 from typing import Any
+from urllib.parse import urlparse
 
 from jira import JIRA
 from jira.resources import Issue
@@ -12,93 +12,129 @@ from danswer.configs.app_configs import JIRA_CONNECTOR_LABELS_TO_SKIP
 from danswer.configs.app_configs import JIRA_CONNECTOR_MAX_TICKET_SIZE
 from danswer.configs.constants import DocumentSource
 from danswer.connectors.cross_connector_utils.miscellaneous_utils import time_str_to_utc
-from danswer.connectors.danswer_jira.utils import best_effort_basic_expert_info
-from danswer.connectors.danswer_jira.utils import best_effort_get_field_from_issue
-from danswer.connectors.danswer_jira.utils import build_jira_client
-from danswer.connectors.danswer_jira.utils import build_jira_url
-from danswer.connectors.danswer_jira.utils import extract_jira_project
-from danswer.connectors.danswer_jira.utils import extract_text_from_adf
-from danswer.connectors.danswer_jira.utils import get_comment_strs
 from danswer.connectors.interfaces import GenerateDocumentsOutput
-from danswer.connectors.interfaces import GenerateSlimDocumentOutput
 from danswer.connectors.interfaces import LoadConnector
 from danswer.connectors.interfaces import PollConnector
 from danswer.connectors.interfaces import SecondsSinceUnixEpoch
-from danswer.connectors.interfaces import SlimConnector
+from danswer.connectors.models import BasicExpertInfo
 from danswer.connectors.models import ConnectorMissingCredentialError
 from danswer.connectors.models import Document
 from danswer.connectors.models import Section
-from danswer.connectors.models import SlimDocument
 from danswer.utils.logger import setup_logger
 
 
 logger = setup_logger()
-
+PROJECT_URL_PAT = "projects"
 JIRA_API_VERSION = os.environ.get("JIRA_API_VERSION") or "2"
-_JIRA_SLIM_PAGE_SIZE = 500
-_JIRA_FULL_PAGE_SIZE = 50
 
 
-def _paginate_jql_search(
-    jira_client: JIRA,
-    jql: str,
-    max_results: int,
-    fields: str | None = None,
-) -> Iterable[Issue]:
-    start = 0
-    while True:
-        logger.debug(
-            f"Fetching Jira issues with JQL: {jql}, "
-            f"starting at {start}, max results: {max_results}"
-        )
-        issues = jira_client.search_issues(
-            jql_str=jql,
-            startAt=start,
-            maxResults=max_results,
-            fields=fields,
-        )
+def extract_jira_project(url: str) -> tuple[str, str]:
+    parsed_url = urlparse(url)
+    jira_base = parsed_url.scheme + "://" + parsed_url.netloc
 
-        for issue in issues:
-            if isinstance(issue, Issue):
-                yield issue
-            else:
-                raise Exception(f"Found Jira object not of type Issue: {issue}")
+    # Split the path by '/' and find the position of 'projects' to get the project name
+    split_path = parsed_url.path.split("/")
+    if PROJECT_URL_PAT in split_path:
+        project_pos = split_path.index(PROJECT_URL_PAT)
+        if len(split_path) > project_pos + 1:
+            jira_project = split_path[project_pos + 1]
+        else:
+            raise ValueError("No project name found in the URL")
+    else:
+        raise ValueError("'projects' not found in the URL")
 
-        if len(issues) < max_results:
-            break
+    return jira_base, jira_project
 
-        start += max_results
+
+def extract_text_from_adf(adf: dict | None) -> str:
+    """Extracts plain text from Atlassian Document Format:
+    https://developer.atlassian.com/cloud/jira/platform/apis/document/structure/
+
+    WARNING: This function is incomplete and will e.g. skip lists!
+    """
+    texts = []
+    if adf is not None and "content" in adf:
+        for block in adf["content"]:
+            if "content" in block:
+                for item in block["content"]:
+                    if item["type"] == "text":
+                        texts.append(item["text"])
+    return " ".join(texts)
+
+
+def best_effort_get_field_from_issue(jira_issue: Issue, field: str) -> Any:
+    if hasattr(jira_issue.fields, field):
+        return getattr(jira_issue.fields, field)
+
+    try:
+        return jira_issue.raw["fields"][field]
+    except Exception:
+        return None
+
+
+def _get_comment_strs(
+    jira: Issue, comment_email_blacklist: tuple[str, ...] = ()
+) -> list[str]:
+    comment_strs = []
+    for comment in jira.fields.comment.comments:
+        try:
+            body_text = (
+                comment.body
+                if JIRA_API_VERSION == "2"
+                else extract_text_from_adf(comment.raw["body"])
+            )
+
+            if (
+                hasattr(comment, "author")
+                and hasattr(comment.author, "emailAddress")
+                and comment.author.emailAddress in comment_email_blacklist
+            ):
+                continue  # Skip adding comment if author's email is in blacklist
+
+            comment_strs.append(body_text)
+        except Exception as e:
+            logger.error(f"Failed to process comment due to an error: {e}")
+            continue
+
+    return comment_strs
 
 
 def fetch_jira_issues_batch(
-    jira_client: JIRA,
     jql: str,
-    batch_size: int,
+    start_index: int,
+    jira_client: JIRA,
+    batch_size: int = INDEX_BATCH_SIZE,
     comment_email_blacklist: tuple[str, ...] = (),
     labels_to_skip: set[str] | None = None,
-) -> Iterable[Document]:
-    for issue in _paginate_jql_search(
-        jira_client=jira_client,
-        jql=jql,
-        max_results=batch_size,
-    ):
-        if labels_to_skip:
-            if any(label in issue.fields.labels for label in labels_to_skip):
-                logger.info(
-                    f"Skipping {issue.key} because it has a label to skip. Found "
-                    f"labels: {issue.fields.labels}. Labels to skip: {labels_to_skip}."
-                )
-                continue
+) -> tuple[list[Document], int]:
+    doc_batch = []
+
+    batch = jira_client.search_issues(
+        jql,
+        startAt=start_index,
+        maxResults=batch_size,
+    )
+
+    for jira in batch:
+        if type(jira) != Issue:
+            logger.warning(f"Found Jira object not of type Issue {jira}")
+            continue
+
+        if labels_to_skip and any(
+            label in jira.fields.labels for label in labels_to_skip
+        ):
+            logger.info(
+                f"Skipping {jira.key} because it has a label to skip. Found "
+                f"labels: {jira.fields.labels}. Labels to skip: {labels_to_skip}."
+            )
+            continue
 
         description = (
-            issue.fields.description
+            jira.fields.description
             if JIRA_API_VERSION == "2"
-            else extract_text_from_adf(issue.raw["fields"]["description"])
-        )
-        comments = get_comment_strs(
-            issue=issue,
-            comment_email_blacklist=comment_email_blacklist,
+            else extract_text_from_adf(jira.raw["fields"]["description"])
         )
+        comments = _get_comment_strs(jira, comment_email_blacklist)
         ticket_content = f"{description}\n" + "\n".join(
             [f"Comment: {comment}" for comment in comments if comment]
         )
@@ -106,53 +142,66 @@ def fetch_jira_issues_batch(
         # Check ticket size
         if len(ticket_content.encode("utf-8")) > JIRA_CONNECTOR_MAX_TICKET_SIZE:
             logger.info(
-                f"Skipping {issue.key} because it exceeds the maximum size of "
+                f"Skipping {jira.key} because it exceeds the maximum size of "
                 f"{JIRA_CONNECTOR_MAX_TICKET_SIZE} bytes."
             )
             continue
 
-        page_url = f"{jira_client.client_info()}/browse/{issue.key}"
+        page_url = f"{jira_client.client_info()}/browse/{jira.key}"
 
         people = set()
         try:
-            creator = best_effort_get_field_from_issue(issue, "creator")
-            if basic_expert_info := best_effort_basic_expert_info(creator):
-                people.add(basic_expert_info)
+            people.add(
+                BasicExpertInfo(
+                    display_name=jira.fields.creator.displayName,
+                    email=jira.fields.creator.emailAddress,
+                )
+            )
         except Exception:
             # Author should exist but if not, doesn't matter
             pass
 
         try:
-            assignee = best_effort_get_field_from_issue(issue, "assignee")
-            if basic_expert_info := best_effort_basic_expert_info(assignee):
-                people.add(basic_expert_info)
+            people.add(
+                BasicExpertInfo(
+                    display_name=jira.fields.assignee.displayName,  # type: ignore
+                    email=jira.fields.assignee.emailAddress,  # type: ignore
+                )
+            )
         except Exception:
             # Author should exist but if not, doesn't matter
             pass
 
         metadata_dict = {}
-        if priority := best_effort_get_field_from_issue(issue, "priority"):
+        priority = best_effort_get_field_from_issue(jira, "priority")
+        if priority:
             metadata_dict["priority"] = priority.name
-        if status := best_effort_get_field_from_issue(issue, "status"):
+        status = best_effort_get_field_from_issue(jira, "status")
+        if status:
             metadata_dict["status"] = status.name
-        if resolution := best_effort_get_field_from_issue(issue, "resolution"):
+        resolution = best_effort_get_field_from_issue(jira, "resolution")
+        if resolution:
             metadata_dict["resolution"] = resolution.name
-        if labels := best_effort_get_field_from_issue(issue, "labels"):
+        labels = best_effort_get_field_from_issue(jira, "labels")
+        if labels:
             metadata_dict["label"] = labels
 
-        yield Document(
-            id=page_url,
-            sections=[Section(link=page_url, text=ticket_content)],
-            source=DocumentSource.JIRA,
-            semantic_identifier=issue.fields.summary,
-            doc_updated_at=time_str_to_utc(issue.fields.updated),
-            primary_owners=list(people) or None,
-            # TODO add secondary_owners (commenters) if needed
-            metadata=metadata_dict,
+        doc_batch.append(
+            Document(
+                id=page_url,
+                sections=[Section(link=page_url, text=ticket_content)],
+                source=DocumentSource.JIRA,
+                semantic_identifier=jira.fields.summary,
+                doc_updated_at=time_str_to_utc(jira.fields.updated),
+                primary_owners=list(people) or None,
+                # TODO add secondary_owners (commenters) if needed
+                metadata=metadata_dict,
+            )
         )
+    return doc_batch, len(batch)
 
 
-class JiraConnector(LoadConnector, PollConnector, SlimConnector):
+class JiraConnector(LoadConnector, PollConnector):
     def __init__(
         self,
         jira_project_url: str,
@@ -164,8 +213,8 @@ class JiraConnector(LoadConnector, PollConnector, SlimConnector):
         labels_to_skip: list[str] = JIRA_CONNECTOR_LABELS_TO_SKIP,
     ) -> None:
         self.batch_size = batch_size
-        self.jira_base, self._jira_project = extract_jira_project(jira_project_url)
-        self._jira_client: JIRA | None = None
+        self.jira_base, self.jira_project = extract_jira_project(jira_project_url)
+        self.jira_client: JIRA | None = None
         self._comment_email_blacklist = comment_email_blacklist or []
 
         self.labels_to_skip = set(labels_to_skip)
@@ -174,45 +223,54 @@ class JiraConnector(LoadConnector, PollConnector, SlimConnector):
     def comment_email_blacklist(self) -> tuple:
         return tuple(email.strip() for email in self._comment_email_blacklist)
 
-    @property
-    def jira_client(self) -> JIRA:
-        if self._jira_client is None:
-            raise ConnectorMissingCredentialError("Jira")
-        return self._jira_client
-
-    @property
-    def quoted_jira_project(self) -> str:
-        # Quote the project name to handle reserved words
-        return f'"{self._jira_project}"'
-
     def load_credentials(self, credentials: dict[str, Any]) -> dict[str, Any] | None:
-        self._jira_client = build_jira_client(
-            credentials=credentials,
-            jira_base=self.jira_base,
-        )
+        api_token = credentials["jira_api_token"]
+        # if user provide an email we assume it's cloud
+        if "jira_user_email" in credentials:
+            email = credentials["jira_user_email"]
+            self.jira_client = JIRA(
+                basic_auth=(email, api_token),
+                server=self.jira_base,
+                options={"rest_api_version": JIRA_API_VERSION},
+            )
+        else:
+            self.jira_client = JIRA(
+                token_auth=api_token,
+                server=self.jira_base,
+                options={"rest_api_version": JIRA_API_VERSION},
+            )
         return None
 
     def load_from_state(self) -> GenerateDocumentsOutput:
-        jql = f"project = {self.quoted_jira_project}"
+        if self.jira_client is None:
+            raise ConnectorMissingCredentialError("Jira")
 
-        document_batch = []
-        for doc in fetch_jira_issues_batch(
-            jira_client=self.jira_client,
-            jql=jql,
-            batch_size=_JIRA_FULL_PAGE_SIZE,
-            comment_email_blacklist=self.comment_email_blacklist,
-            labels_to_skip=self.labels_to_skip,
-        ):
-            document_batch.append(doc)
-            if len(document_batch) >= self.batch_size:
-                yield document_batch
-                document_batch = []
+        # Quote the project name to handle reserved words
+        quoted_project = f'"{self.jira_project}"'
+        start_ind = 0
+        while True:
+            doc_batch, fetched_batch_size = fetch_jira_issues_batch(
+                jql=f"project = {quoted_project}",
+                start_index=start_ind,
+                jira_client=self.jira_client,
+                batch_size=self.batch_size,
+                comment_email_blacklist=self.comment_email_blacklist,
+                labels_to_skip=self.labels_to_skip,
+            )
 
-        yield document_batch
+            if doc_batch:
+                yield doc_batch
+
+            start_ind += fetched_batch_size
+            if fetched_batch_size < self.batch_size:
+                break
 
     def poll_source(
         self, start: SecondsSinceUnixEpoch, end: SecondsSinceUnixEpoch
     ) -> GenerateDocumentsOutput:
+        if self.jira_client is None:
+            raise ConnectorMissingCredentialError("Jira")
+
         start_date_str = datetime.fromtimestamp(start, tz=timezone.utc).strftime(
             "%Y-%m-%d %H:%M"
         )
@@ -220,54 +278,31 @@ class JiraConnector(LoadConnector, PollConnector, SlimConnector):
             "%Y-%m-%d %H:%M"
         )
 
+        # Quote the project name to handle reserved words
+        quoted_project = f'"{self.jira_project}"'
         jql = (
-            f"project = {self.quoted_jira_project} AND "
+            f"project = {quoted_project} AND "
             f"updated >= '{start_date_str}' AND "
             f"updated <= '{end_date_str}'"
         )
 
-        document_batch = []
-        for doc in fetch_jira_issues_batch(
-            jira_client=self.jira_client,
-            jql=jql,
-            batch_size=_JIRA_FULL_PAGE_SIZE,
-            comment_email_blacklist=self.comment_email_blacklist,
-            labels_to_skip=self.labels_to_skip,
-        ):
-            document_batch.append(doc)
-            if len(document_batch) >= self.batch_size:
-                yield document_batch
-                document_batch = []
-
-        yield document_batch
-
-    def retrieve_all_slim_documents(
-        self,
-        start: SecondsSinceUnixEpoch | None = None,
-        end: SecondsSinceUnixEpoch | None = None,
-    ) -> GenerateSlimDocumentOutput:
-        jql = f"project = {self.quoted_jira_project}"
-
-        slim_doc_batch = []
-        for issue in _paginate_jql_search(
-            jira_client=self.jira_client,
-            jql=jql,
-            max_results=_JIRA_SLIM_PAGE_SIZE,
-            fields="key",
-        ):
-            issue_key = best_effort_get_field_from_issue(issue, "key")
-            id = build_jira_url(self.jira_client, issue_key)
-            slim_doc_batch.append(
-                SlimDocument(
-                    id=id,
-                    perm_sync_data=None,
-                )
+        start_ind = 0
+        while True:
+            doc_batch, fetched_batch_size = fetch_jira_issues_batch(
+                jql=jql,
+                start_index=start_ind,
+                jira_client=self.jira_client,
+                batch_size=self.batch_size,
+                comment_email_blacklist=self.comment_email_blacklist,
+                labels_to_skip=self.labels_to_skip,
             )
-            if len(slim_doc_batch) >= _JIRA_SLIM_PAGE_SIZE:
-                yield slim_doc_batch
-                slim_doc_batch = []
 
-        yield slim_doc_batch
+            if doc_batch:
+                yield doc_batch
+
+            start_ind += fetched_batch_size
+            if fetched_batch_size < self.batch_size:
+                break
 
 
 if __name__ == "__main__":

backend/danswer/connectors/danswer_jira/utils.py

@@ -1,136 +1,17 @@
 """Module with custom fields processing functions"""
-import os
 from typing import Any
 from typing import List
-from urllib.parse import urlparse
 
 from jira import JIRA
 from jira.resources import CustomFieldOption
 from jira.resources import Issue
 from jira.resources import User
 
-from danswer.connectors.models import BasicExpertInfo
 from danswer.utils.logger import setup_logger
 
 logger = setup_logger()
 
 
-PROJECT_URL_PAT = "projects"
-JIRA_API_VERSION = os.environ.get("JIRA_API_VERSION") or "2"
-
-
-def best_effort_basic_expert_info(obj: Any) -> BasicExpertInfo | None:
-    display_name = None
-    email = None
-    if hasattr(obj, "display_name"):
-        display_name = obj.display_name
-    else:
-        display_name = obj.get("displayName")
-
-    if hasattr(obj, "emailAddress"):
-        email = obj.emailAddress
-    else:
-        email = obj.get("emailAddress")
-
-    if not email and not display_name:
-        return None
-
-    return BasicExpertInfo(display_name=display_name, email=email)
-
-
-def best_effort_get_field_from_issue(jira_issue: Issue, field: str) -> Any:
-    if hasattr(jira_issue.fields, field):
-        return getattr(jira_issue.fields, field)
-
-    try:
-        return jira_issue.raw["fields"][field]
-    except Exception:
-        return None
-
-
-def extract_text_from_adf(adf: dict | None) -> str:
-    """Extracts plain text from Atlassian Document Format:
-    https://developer.atlassian.com/cloud/jira/platform/apis/document/structure/
-
-    WARNING: This function is incomplete and will e.g. skip lists!
-    """
-    texts = []
-    if adf is not None and "content" in adf:
-        for block in adf["content"]:
-            if "content" in block:
-                for item in block["content"]:
-                    if item["type"] == "text":
-                        texts.append(item["text"])
-    return " ".join(texts)
-
-
-def build_jira_url(jira_client: JIRA, issue_key: str) -> str:
-    return f"{jira_client.client_info()}/browse/{issue_key}"
-
-
-def build_jira_client(credentials: dict[str, Any], jira_base: str) -> JIRA:
-    api_token = credentials["jira_api_token"]
-    # if user provide an email we assume it's cloud
-    if "jira_user_email" in credentials:
-        email = credentials["jira_user_email"]
-        return JIRA(
-            basic_auth=(email, api_token),
-            server=jira_base,
-            options={"rest_api_version": JIRA_API_VERSION},
-        )
-    else:
-        return JIRA(
-            token_auth=api_token,
-            server=jira_base,
-            options={"rest_api_version": JIRA_API_VERSION},
-        )
-
-
-def extract_jira_project(url: str) -> tuple[str, str]:
-    parsed_url = urlparse(url)
-    jira_base = parsed_url.scheme + "://" + parsed_url.netloc
-
-    # Split the path by '/' and find the position of 'projects' to get the project name
-    split_path = parsed_url.path.split("/")
-    if PROJECT_URL_PAT in split_path:
-        project_pos = split_path.index(PROJECT_URL_PAT)
-        if len(split_path) > project_pos + 1:
-            jira_project = split_path[project_pos + 1]
-        else:
-            raise ValueError("No project name found in the URL")
-    else:
-        raise ValueError("'projects' not found in the URL")
-
-    return jira_base, jira_project
-
-
-def get_comment_strs(
-    issue: Issue, comment_email_blacklist: tuple[str, ...] = ()
-) -> list[str]:
-    comment_strs = []
-    for comment in issue.fields.comment.comments:
-        try:
-            body_text = (
-                comment.body
-                if JIRA_API_VERSION == "2"
-                else extract_text_from_adf(comment.raw["body"])
-            )
-
-            if (
-                hasattr(comment, "author")
-                and hasattr(comment.author, "emailAddress")
-                and comment.author.emailAddress in comment_email_blacklist
-            ):
-                continue  # Skip adding comment if author's email is in blacklist
-
-            comment_strs.append(body_text)
-        except Exception as e:
-            logger.error(f"Failed to process comment due to an error: {e}")
-            continue
-
-    return comment_strs
-
-
 class CustomFieldExtractor:
     @staticmethod
     def _process_custom_field_value(value: Any) -> str:

backend/danswer/connectors/google_drive/connector.py

@@ -15,7 +15,6 @@ from danswer.connectors.google_drive.doc_conversion import (
     convert_drive_item_to_document,
 )
 from danswer.connectors.google_drive.file_retrieval import crawl_folders_for_files
-from danswer.connectors.google_drive.file_retrieval import get_all_files_for_oauth
 from danswer.connectors.google_drive.file_retrieval import get_all_files_in_my_drive
 from danswer.connectors.google_drive.file_retrieval import get_files_in_shared_drive
 from danswer.connectors.google_drive.models import GoogleDriveFileType
@@ -83,31 +82,12 @@ def _process_files_batch(
         yield doc_batch
 
 
-def _clean_requested_drive_ids(
-    requested_drive_ids: set[str],
-    requested_folder_ids: set[str],
-    all_drive_ids_available: set[str],
-) -> tuple[set[str], set[str]]:
-    invalid_requested_drive_ids = requested_drive_ids - all_drive_ids_available
-    filtered_folder_ids = requested_folder_ids - all_drive_ids_available
-    if invalid_requested_drive_ids:
-        logger.warning(
-            f"Some shared drive IDs were not found. IDs: {invalid_requested_drive_ids}"
-        )
-        logger.warning("Checking for folder access instead...")
-        filtered_folder_ids.update(invalid_requested_drive_ids)
-
-    valid_requested_drive_ids = requested_drive_ids - invalid_requested_drive_ids
-    return valid_requested_drive_ids, filtered_folder_ids
-
-
 class GoogleDriveConnector(LoadConnector, PollConnector, SlimConnector):
     def __init__(
         self,
-        include_shared_drives: bool = False,
-        include_my_drives: bool = False,
-        include_files_shared_with_me: bool = False,
+        include_shared_drives: bool = True,
         shared_drive_urls: str | None = None,
+        include_my_drives: bool = True,
         my_drive_emails: str | None = None,
         shared_folder_urls: str | None = None,
         batch_size: int = INDEX_BATCH_SIZE,
@@ -140,36 +120,22 @@ class GoogleDriveConnector(LoadConnector, PollConnector, SlimConnector):
         if (
             not include_shared_drives
             and not include_my_drives
-            and not include_files_shared_with_me
             and not shared_folder_urls
-            and not my_drive_emails
-            and not shared_drive_urls
         ):
             raise ValueError(
-                "Nothing to index. Please specify at least one of the following: "
-                "include_shared_drives, include_my_drives, include_files_shared_with_me, "
-                "shared_folder_urls, or my_drive_emails"
+                "At least one of include_shared_drives, include_my_drives,"
+                " or shared_folder_urls must be true"
             )
 
         self.batch_size = batch_size
 
-        specific_requests_made = False
-        if bool(shared_drive_urls) or bool(my_drive_emails) or bool(shared_folder_urls):
-            specific_requests_made = True
-
-        self.include_files_shared_with_me = (
-            False if specific_requests_made else include_files_shared_with_me
-        )
-        self.include_my_drives = False if specific_requests_made else include_my_drives
-        self.include_shared_drives = (
-            False if specific_requests_made else include_shared_drives
-        )
-
+        self.include_shared_drives = include_shared_drives
         shared_drive_url_list = _extract_str_list_from_comma_str(shared_drive_urls)
         self._requested_shared_drive_ids = set(
             _extract_ids_from_urls(shared_drive_url_list)
         )
 
+        self.include_my_drives = include_my_drives
         self._requested_my_drive_emails = set(
             _extract_str_list_from_comma_str(my_drive_emails)
         )
@@ -259,20 +225,26 @@ class GoogleDriveConnector(LoadConnector, PollConnector, SlimConnector):
             creds=self.creds,
             user_email=self.primary_admin_email,
         )
-        is_service_account = isinstance(self.creds, ServiceAccountCredentials)
         all_drive_ids = set()
+        # We don't want to fail if we're using OAuth because you can
+        # access your my drive as a non admin user in an org still
+        ignore_fetch_failure = isinstance(self.creds, OAuthCredentials)
         for drive in execute_paginated_retrieval(
             retrieval_function=primary_drive_service.drives().list,
             list_key="drives",
-            useDomainAdminAccess=is_service_account,
+            continue_on_404_or_403=ignore_fetch_failure,
+            useDomainAdminAccess=True,
             fields="drives(id)",
         ):
             all_drive_ids.add(drive["id"])
 
         if not all_drive_ids:
             logger.warning(
-                "No drives found even though we are indexing shared drives was requested."
+                "No drives found. This is likely because oauth user "
+                "is not an admin and cannot view all drive IDs. "
+                "Continuing with only the shared drive IDs specified in the config."
             )
+            all_drive_ids = set(self._requested_shared_drive_ids)
 
         return all_drive_ids
 
@@ -289,9 +261,14 @@ class GoogleDriveConnector(LoadConnector, PollConnector, SlimConnector):
 
         # if we are including my drives, try to get the current user's my
         # drive if any of the following are true:
-        # - include_my_drives is true
+        # - no specific emails were requested
         # - the current user's email is in the requested emails
-        if self.include_my_drives or user_email in self._requested_my_drive_emails:
+        # - we are using OAuth (in which case we assume that is the only email we will try)
+        if self.include_my_drives and (
+            not self._requested_my_drive_emails
+            or user_email in self._requested_my_drive_emails
+            or isinstance(self.creds, OAuthCredentials)
+        ):
             yield from get_all_files_in_my_drive(
                 service=drive_service,
                 update_traversed_ids_func=self._update_traversed_parent_ids,
@@ -322,7 +299,7 @@ class GoogleDriveConnector(LoadConnector, PollConnector, SlimConnector):
                 end=end,
             )
 
-    def _manage_service_account_retrieval(
+    def _fetch_drive_items(
         self,
         is_slim: bool,
         start: SecondsSinceUnixEpoch | None = None,
@@ -332,16 +309,29 @@ class GoogleDriveConnector(LoadConnector, PollConnector, SlimConnector):
 
         all_drive_ids: set[str] = self._get_all_drive_ids()
 
-        drive_ids_to_retrieve: set[str] = set()
-        folder_ids_to_retrieve: set[str] = set()
-        if self._requested_shared_drive_ids or self._requested_folder_ids:
-            drive_ids_to_retrieve, folder_ids_to_retrieve = _clean_requested_drive_ids(
-                requested_drive_ids=self._requested_shared_drive_ids,
-                requested_folder_ids=self._requested_folder_ids,
-                all_drive_ids_available=all_drive_ids,
+        # remove drive ids from the folder ids because they are queried differently
+        filtered_folder_ids = self._requested_folder_ids - all_drive_ids
+
+        # Remove drive_ids that are not in the all_drive_ids and check them as folders instead
+        invalid_drive_ids = self._requested_shared_drive_ids - all_drive_ids
+        if invalid_drive_ids:
+            logger.warning(
+                f"Some shared drive IDs were not found. IDs: {invalid_drive_ids}"
             )
-        elif self.include_shared_drives:
-            drive_ids_to_retrieve = all_drive_ids
+            logger.warning("Checking for folder access instead...")
+            filtered_folder_ids.update(invalid_drive_ids)
+
+        # If including shared drives, use the requested IDs if provided,
+        # otherwise use all drive IDs
+        filtered_drive_ids = set()
+        if self.include_shared_drives:
+            if self._requested_shared_drive_ids:
+                # Remove invalid drive IDs from requested IDs
+                filtered_drive_ids = (
+                    self._requested_shared_drive_ids - invalid_drive_ids
+                )
+            else:
+                filtered_drive_ids = all_drive_ids
 
         # Process users in parallel using ThreadPoolExecutor
         with ThreadPoolExecutor(max_workers=10) as executor:
@@ -350,8 +340,8 @@ class GoogleDriveConnector(LoadConnector, PollConnector, SlimConnector):
                     self._impersonate_user_for_retrieval,
                     email,
                     is_slim,
-                    drive_ids_to_retrieve,
-                    folder_ids_to_retrieve,
+                    filtered_drive_ids,
+                    filtered_folder_ids,
                     start,
                     end,
                 ): email
@@ -363,101 +353,13 @@ class GoogleDriveConnector(LoadConnector, PollConnector, SlimConnector):
                 yield from future.result()
 
         remaining_folders = (
-            drive_ids_to_retrieve | folder_ids_to_retrieve
+            filtered_drive_ids | filtered_folder_ids
         ) - self._retrieved_ids
         if remaining_folders:
             logger.warning(
                 f"Some folders/drives were not retrieved. IDs: {remaining_folders}"
             )
 
-    def _manage_oauth_retrieval(
-        self,
-        is_slim: bool,
-        start: SecondsSinceUnixEpoch | None = None,
-        end: SecondsSinceUnixEpoch | None = None,
-    ) -> Iterator[GoogleDriveFileType]:
-        drive_service = get_drive_service(self.creds, self.primary_admin_email)
-
-        if self.include_files_shared_with_me or self.include_my_drives:
-            yield from get_all_files_for_oauth(
-                service=drive_service,
-                include_files_shared_with_me=self.include_files_shared_with_me,
-                include_my_drives=self.include_my_drives,
-                include_shared_drives=self.include_shared_drives,
-                is_slim=is_slim,
-                start=start,
-                end=end,
-            )
-
-        all_requested = (
-            self.include_files_shared_with_me
-            and self.include_my_drives
-            and self.include_shared_drives
-        )
-        if all_requested:
-            # If all 3 are true, we already yielded from get_all_files_for_oauth
-            return
-
-        all_drive_ids = self._get_all_drive_ids()
-        drive_ids_to_retrieve: set[str] = set()
-        folder_ids_to_retrieve: set[str] = set()
-        if self._requested_shared_drive_ids or self._requested_folder_ids:
-            drive_ids_to_retrieve, folder_ids_to_retrieve = _clean_requested_drive_ids(
-                requested_drive_ids=self._requested_shared_drive_ids,
-                requested_folder_ids=self._requested_folder_ids,
-                all_drive_ids_available=all_drive_ids,
-            )
-        elif self.include_shared_drives:
-            drive_ids_to_retrieve = all_drive_ids
-
-        for drive_id in drive_ids_to_retrieve:
-            yield from get_files_in_shared_drive(
-                service=drive_service,
-                drive_id=drive_id,
-                is_slim=is_slim,
-                update_traversed_ids_func=self._update_traversed_parent_ids,
-                start=start,
-                end=end,
-            )
-
-        # Even if no folders were requested, we still check if any drives were requested
-        # that could be folders.
-        remaining_folders = folder_ids_to_retrieve - self._retrieved_ids
-        for folder_id in remaining_folders:
-            yield from crawl_folders_for_files(
-                service=drive_service,
-                parent_id=folder_id,
-                traversed_parent_ids=self._retrieved_ids,
-                update_traversed_ids_func=self._update_traversed_parent_ids,
-                start=start,
-                end=end,
-            )
-
-        remaining_folders = (
-            drive_ids_to_retrieve | folder_ids_to_retrieve
-        ) - self._retrieved_ids
-        if remaining_folders:
-            logger.warning(
-                f"Some folders/drives were not retrieved. IDs: {remaining_folders}"
-            )
-
-    def _fetch_drive_items(
-        self,
-        is_slim: bool,
-        start: SecondsSinceUnixEpoch | None = None,
-        end: SecondsSinceUnixEpoch | None = None,
-    ) -> Iterator[GoogleDriveFileType]:
-        retrieval_method = (
-            self._manage_service_account_retrieval
-            if isinstance(self.creds, ServiceAccountCredentials)
-            else self._manage_oauth_retrieval
-        )
-        return retrieval_method(
-            is_slim=is_slim,
-            start=start,
-            end=end,
-        )
-
     def _extract_docs_from_google_drive(
         self,
         start: SecondsSinceUnixEpoch | None = None,

backend/danswer/connectors/google_drive/doc_conversion.py

@@ -2,7 +2,6 @@ import io
 from datetime import datetime
 from datetime import timezone
 
-from googleapiclient.discovery import build  # type: ignore
 from googleapiclient.errors import HttpError  # type: ignore
 
 from danswer.configs.app_configs import CONTINUE_ON_CONNECTOR_FAILURE
@@ -49,67 +48,6 @@ def _extract_sections_basic(
         return [Section(link=link, text=UNSUPPORTED_FILE_TYPE_CONTENT)]
 
     try:
-        if mime_type == GDriveMimeType.SPREADSHEET.value:
-            try:
-                sheets_service = build(
-                    "sheets", "v4", credentials=service._http.credentials
-                )
-                spreadsheet = (
-                    sheets_service.spreadsheets()
-                    .get(spreadsheetId=file["id"])
-                    .execute()
-                )
-
-                sections = []
-                for sheet in spreadsheet["sheets"]:
-                    sheet_name = sheet["properties"]["title"]
-                    sheet_id = sheet["properties"]["sheetId"]
-
-                    # Get sheet dimensions
-                    grid_properties = sheet["properties"].get("gridProperties", {})
-                    row_count = grid_properties.get("rowCount", 1000)
-                    column_count = grid_properties.get("columnCount", 26)
-
-                    # Convert column count to letter (e.g., 26 -> Z, 27 -> AA)
-                    end_column = ""
-                    while column_count:
-                        column_count, remainder = divmod(column_count - 1, 26)
-                        end_column = chr(65 + remainder) + end_column
-
-                    range_name = f"'{sheet_name}'!A1:{end_column}{row_count}"
-
-                    try:
-                        result = (
-                            sheets_service.spreadsheets()
-                            .values()
-                            .get(spreadsheetId=file["id"], range=range_name)
-                            .execute()
-                        )
-                        values = result.get("values", [])
-
-                        if values:
-                            text = f"Sheet: {sheet_name}\n"
-                            for row in values:
-                                text += "\t".join(str(cell) for cell in row) + "\n"
-                            sections.append(
-                                Section(
-                                    link=f"{link}#gid={sheet_id}",
-                                    text=text,
-                                )
-                            )
-                    except HttpError as e:
-                        logger.warning(
-                            f"Error fetching data for sheet '{sheet_name}': {e}"
-                        )
-                        continue
-                return sections
-
-            except Exception as e:
-                logger.warning(
-                    f"Ran into exception '{e}' when pulling data from Google Sheet '{file['name']}'."
-                    " Falling back to basic extraction."
-                )
-
         if mime_type in [
             GDriveMimeType.DOC.value,
             GDriveMimeType.PPT.value,
@@ -127,7 +65,6 @@ def _extract_sections_basic(
                 .decode("utf-8")
             )
             return [Section(link=link, text=text)]
-
         elif mime_type in [
             GDriveMimeType.PLAIN_TEXT.value,
             GDriveMimeType.MARKDOWN.value,

backend/danswer/connectors/google_drive/file_retrieval.py

@@ -140,8 +140,8 @@ def get_files_in_shared_drive(
 ) -> Iterator[GoogleDriveFileType]:
     # If we know we are going to folder crawl later, we can cache the folders here
     # Get all folders being queried and add them to the traversed set
-    folder_query = f"mimeType = '{DRIVE_FOLDER_TYPE}'"
-    folder_query += " and trashed = false"
+    query = f"mimeType = '{DRIVE_FOLDER_TYPE}'"
+    query += " and trashed = false"
     found_folders = False
     for file in execute_paginated_retrieval(
         retrieval_function=service.files().list,
@@ -152,7 +152,7 @@ def get_files_in_shared_drive(
         supportsAllDrives=True,
         includeItemsFromAllDrives=True,
         fields="nextPageToken, files(id)",
-        q=folder_query,
+        q=query,
     ):
         update_traversed_ids_func(file["id"])
         found_folders = True
@@ -160,9 +160,9 @@ def get_files_in_shared_drive(
         update_traversed_ids_func(drive_id)
 
     # Get all files in the shared drive
-    file_query = f"mimeType != '{DRIVE_FOLDER_TYPE}'"
-    file_query += " and trashed = false"
-    file_query += _generate_time_range_filter(start, end)
+    query = f"mimeType != '{DRIVE_FOLDER_TYPE}'"
+    query += " and trashed = false"
+    query += _generate_time_range_filter(start, end)
     yield from execute_paginated_retrieval(
         retrieval_function=service.files().list,
         list_key="files",
@@ -172,7 +172,7 @@ def get_files_in_shared_drive(
         supportsAllDrives=True,
         includeItemsFromAllDrives=True,
         fields=SLIM_FILE_FIELDS if is_slim else FILE_FIELDS,
-        q=file_query,
+        q=query,
     )
 
 
@@ -185,16 +185,14 @@ def get_all_files_in_my_drive(
 ) -> Iterator[GoogleDriveFileType]:
     # If we know we are going to folder crawl later, we can cache the folders here
     # Get all folders being queried and add them to the traversed set
-    folder_query = f"mimeType = '{DRIVE_FOLDER_TYPE}'"
-    folder_query += " and trashed = false"
-    folder_query += " and 'me' in owners"
+    query = "trashed = false and 'me' in owners"
     found_folders = False
     for file in execute_paginated_retrieval(
         retrieval_function=service.files().list,
         list_key="files",
         corpora="user",
         fields=SLIM_FILE_FIELDS if is_slim else FILE_FIELDS,
-        q=folder_query,
+        q=query,
     ):
         update_traversed_ids_func(file["id"])
         found_folders = True
@@ -202,52 +200,18 @@ def get_all_files_in_my_drive(
         update_traversed_ids_func(get_root_folder_id(service))
 
     # Then get the files
-    file_query = f"mimeType != '{DRIVE_FOLDER_TYPE}'"
-    file_query += " and trashed = false"
-    file_query += " and 'me' in owners"
-    file_query += _generate_time_range_filter(start, end)
+    query = "trashed = false and 'me' in owners"
+    query += _generate_time_range_filter(start, end)
+    fields = "files(id, name, mimeType, webViewLink, modifiedTime, createdTime)"
+    if not is_slim:
+        fields += ", files(permissions, permissionIds, owners)"
+
     yield from execute_paginated_retrieval(
         retrieval_function=service.files().list,
         list_key="files",
         corpora="user",
         fields=SLIM_FILE_FIELDS if is_slim else FILE_FIELDS,
-        q=file_query,
-    )
-
-
-def get_all_files_for_oauth(
-    service: Any,
-    include_files_shared_with_me: bool,
-    include_my_drives: bool,
-    # One of the above 2 should be true
-    include_shared_drives: bool,
-    is_slim: bool = False,
-    start: SecondsSinceUnixEpoch | None = None,
-    end: SecondsSinceUnixEpoch | None = None,
-) -> Iterator[GoogleDriveFileType]:
-    should_get_all = (
-        include_shared_drives and include_my_drives and include_files_shared_with_me
-    )
-    corpora = "allDrives" if should_get_all else "user"
-
-    file_query = f"mimeType != '{DRIVE_FOLDER_TYPE}'"
-    file_query += " and trashed = false"
-    file_query += _generate_time_range_filter(start, end)
-
-    if not should_get_all:
-        if include_files_shared_with_me and not include_my_drives:
-            file_query += " and not 'me' in owners"
-        if not include_files_shared_with_me and include_my_drives:
-            file_query += " and 'me' in owners"
-
-    yield from execute_paginated_retrieval(
-        retrieval_function=service.files().list,
-        list_key="files",
-        corpora=corpora,
-        includeItemsFromAllDrives=should_get_all,
-        supportsAllDrives=should_get_all,
-        fields=SLIM_FILE_FIELDS if is_slim else FILE_FIELDS,
-        q=file_query,
+        q=query,
     )
 
 

backend/danswer/danswerbot/slack/config.py

@@ -2,8 +2,8 @@ import os
 
 from sqlalchemy.orm import Session
 
-from danswer.db.models import SlackChannelConfig
-from danswer.db.slack_channel_config import fetch_slack_channel_configs
+from danswer.db.models import SlackBotConfig
+from danswer.db.slack_bot_config import fetch_slack_bot_configs
 
 
 VALID_SLACK_FILTERS = [
@@ -13,52 +13,46 @@ VALID_SLACK_FILTERS = [
 ]
 
 
-def get_slack_channel_config_for_bot_and_channel(
-    db_session: Session,
-    slack_bot_id: int,
-    channel_name: str | None,
-) -> SlackChannelConfig | None:
+def get_slack_bot_config_for_channel(
+    channel_name: str | None, db_session: Session
+) -> SlackBotConfig | None:
     if not channel_name:
         return None
 
-    slack_bot_configs = fetch_slack_channel_configs(
-        db_session=db_session, slack_bot_id=slack_bot_id
-    )
+    slack_bot_configs = fetch_slack_bot_configs(db_session=db_session)
     for config in slack_bot_configs:
-        if channel_name in config.channel_config["channel_name"]:
+        if channel_name in config.channel_config["channel_names"]:
             return config
 
     return None
 
 
-def validate_channel_name(
+def validate_channel_names(
+    channel_names: list[str],
+    current_slack_bot_config_id: int | None,
     db_session: Session,
-    current_slack_bot_id: int,
-    channel_name: str,
-    current_slack_channel_config_id: int | None,
-) -> str:
-    """Make sure that this channel_name does not exist in other Slack channel configs.
-    Returns a cleaned up channel name (e.g. '#' removed if present)"""
-    slack_bot_configs = fetch_slack_channel_configs(
-        db_session=db_session,
-        slack_bot_id=current_slack_bot_id,
-    )
-    cleaned_channel_name = channel_name.lstrip("#").lower()
-    for slack_channel_config in slack_bot_configs:
-        if slack_channel_config.id == current_slack_channel_config_id:
+) -> list[str]:
+    """Make sure that these channel_names don't exist in other slack bot configs.
+    Returns a list of cleaned up channel names (e.g. '#' removed if present)"""
+    slack_bot_configs = fetch_slack_bot_configs(db_session=db_session)
+    cleaned_channel_names = [
+        channel_name.lstrip("#").lower() for channel_name in channel_names
+    ]
+    for slack_bot_config in slack_bot_configs:
+        if slack_bot_config.id == current_slack_bot_config_id:
             continue
 
-        if cleaned_channel_name == slack_channel_config.channel_config["channel_name"]:
-            raise ValueError(
-                f"Channel name '{channel_name}' already exists in "
-                "another Slack channel config with in Slack Bot with name: "
-                f"{slack_channel_config.slack_bot.name}"
-            )
+        for channel_name in cleaned_channel_names:
+            if channel_name in slack_bot_config.channel_config["channel_names"]:
+                raise ValueError(
+                    f"Channel name '{channel_name}' already exists in "
+                    "another slack bot config"
+                )
 
-    return cleaned_channel_name
+    return cleaned_channel_names
 
 
-# Scaling configurations for multi-tenant Slack channel handling
+# Scaling configurations for multi-tenant Slack bot handling
 TENANT_LOCK_EXPIRATION = 1800  # How long a pod can hold exclusive access to a tenant before other pods can acquire it
 TENANT_HEARTBEAT_INTERVAL = (
     15  # How often pods send heartbeats to indicate they are still processing a tenant

backend/danswer/danswerbot/slack/handlers/handle_buttons.py

@@ -13,7 +13,7 @@ from danswer.connectors.slack.utils import expert_info_from_slack_id
 from danswer.connectors.slack.utils import make_slack_api_rate_limited
 from danswer.danswerbot.slack.blocks import build_follow_up_resolved_blocks
 from danswer.danswerbot.slack.blocks import get_document_feedback_blocks
-from danswer.danswerbot.slack.config import get_slack_channel_config_for_bot_and_channel
+from danswer.danswerbot.slack.config import get_slack_bot_config_for_channel
 from danswer.danswerbot.slack.constants import DISLIKE_BLOCK_ACTION_ID
 from danswer.danswerbot.slack.constants import FeedbackVisibility
 from danswer.danswerbot.slack.constants import LIKE_BLOCK_ACTION_ID
@@ -117,10 +117,8 @@ def handle_generate_answer_button(
     )
 
     with get_session_with_tenant(client.tenant_id) as db_session:
-        slack_channel_config = get_slack_channel_config_for_bot_and_channel(
-            db_session=db_session,
-            slack_bot_id=client.slack_bot_id,
-            channel_name=channel_name,
+        slack_bot_config = get_slack_bot_config_for_channel(
+            channel_name=channel_name, db_session=db_session
         )
 
         handle_regular_answer(
@@ -135,7 +133,7 @@ def handle_generate_answer_button(
                 is_bot_msg=False,
                 is_bot_dm=False,
             ),
-            slack_channel_config=slack_channel_config,
+            slack_bot_config=slack_bot_config,
             receiver_ids=None,
             client=client.web_client,
             tenant_id=client.tenant_id,
@@ -258,13 +256,11 @@ def handle_followup_button(
         channel_name, is_dm = get_channel_name_from_id(
             client=client.web_client, channel_id=channel_id
         )
-        slack_channel_config = get_slack_channel_config_for_bot_and_channel(
-            db_session=db_session,
-            slack_bot_id=client.slack_bot_id,
-            channel_name=channel_name,
+        slack_bot_config = get_slack_bot_config_for_channel(
+            channel_name=channel_name, db_session=db_session
         )
-        if slack_channel_config:
-            tag_names = slack_channel_config.channel_config.get("follow_up_tags")
+        if slack_bot_config:
+            tag_names = slack_bot_config.channel_config.get("follow_up_tags")
             remaining = None
             if tag_names:
                 tag_ids, remaining = fetch_user_ids_from_emails(

backend/danswer/danswerbot/slack/handlers/handle_message.py

@@ -19,8 +19,8 @@ from danswer.danswerbot.slack.utils import respond_in_thread
 from danswer.danswerbot.slack.utils import slack_usage_report
 from danswer.danswerbot.slack.utils import update_emote_react
 from danswer.db.engine import get_session_with_tenant
-from danswer.db.models import SlackChannelConfig
-from danswer.db.users import add_slack_user_if_not_exists
+from danswer.db.models import SlackBotConfig
+from danswer.db.users import add_non_web_user_if_not_exists
 from danswer.utils.logger import setup_logger
 from shared_configs.configs import SLACK_CHANNEL_ID
 
@@ -106,7 +106,7 @@ def remove_scheduled_feedback_reminder(
 
 def handle_message(
     message_info: SlackMessageInfo,
-    slack_channel_config: SlackChannelConfig | None,
+    slack_bot_config: SlackBotConfig | None,
     client: WebClient,
     feedback_reminder_id: str | None,
     tenant_id: str | None,
@@ -140,7 +140,7 @@ def handle_message(
     )
 
     document_set_names: list[str] | None = None
-    persona = slack_channel_config.persona if slack_channel_config else None
+    persona = slack_bot_config.persona if slack_bot_config else None
     prompt = None
     if persona:
         document_set_names = [
@@ -152,8 +152,8 @@ def handle_message(
     respond_member_group_list = None
 
     channel_conf = None
-    if slack_channel_config and slack_channel_config.channel_config:
-        channel_conf = slack_channel_config.channel_config
+    if slack_bot_config and slack_bot_config.channel_config:
+        channel_conf = slack_bot_config.channel_config
         if not bypass_filters and "answer_filters" in channel_conf:
             if (
                 "questionmark_prefilter" in channel_conf["answer_filters"]
@@ -213,13 +213,13 @@ def handle_message(
 
     with get_session_with_tenant(tenant_id) as db_session:
         if message_info.email:
-            add_slack_user_if_not_exists(db_session, message_info.email)
+            add_non_web_user_if_not_exists(db_session, message_info.email)
 
         # first check if we need to respond with a standard answer
         used_standard_answer = handle_standard_answers(
             message_info=message_info,
             receiver_ids=send_to,
-            slack_channel_config=slack_channel_config,
+            slack_bot_config=slack_bot_config,
             prompt=prompt,
             logger=logger,
             client=client,
@@ -231,7 +231,7 @@ def handle_message(
         # if no standard answer applies, try a regular answer
         issue_with_regular_answer = handle_regular_answer(
             message_info=message_info,
-            slack_channel_config=slack_channel_config,
+            slack_bot_config=slack_bot_config,
             receiver_ids=send_to,
             client=client,
             channel=channel,

backend/danswer/danswerbot/slack/handlers/handle_regular_answer.py

@@ -34,8 +34,8 @@ from danswer.danswerbot.slack.utils import SlackRateLimiter
 from danswer.danswerbot.slack.utils import update_emote_react
 from danswer.db.engine import get_session_with_tenant
 from danswer.db.models import Persona
+from danswer.db.models import SlackBotConfig
 from danswer.db.models import SlackBotResponseType
-from danswer.db.models import SlackChannelConfig
 from danswer.db.persona import fetch_persona_by_id
 from danswer.db.search_settings import get_current_search_settings
 from danswer.db.users import get_user_by_email
@@ -81,7 +81,7 @@ def rate_limits(
 
 def handle_regular_answer(
     message_info: SlackMessageInfo,
-    slack_channel_config: SlackChannelConfig | None,
+    slack_bot_config: SlackBotConfig | None,
     receiver_ids: list[str] | None,
     client: WebClient,
     channel: str,
@@ -96,7 +96,7 @@ def handle_regular_answer(
     disable_cot: bool = DANSWER_BOT_DISABLE_COT,
     reflexion: bool = ENABLE_DANSWERBOT_REFLEXION,
 ) -> bool:
-    channel_conf = slack_channel_config.channel_config if slack_channel_config else None
+    channel_conf = slack_bot_config.channel_config if slack_bot_config else None
 
     messages = message_info.thread_messages
     message_ts_to_respond_to = message_info.msg_to_respond
@@ -108,7 +108,7 @@ def handle_regular_answer(
                 user = get_user_by_email(message_info.email, db_session)
 
     document_set_names: list[str] | None = None
-    persona = slack_channel_config.persona if slack_channel_config else None
+    persona = slack_bot_config.persona if slack_bot_config else None
     prompt = None
     if persona:
         document_set_names = [
@@ -120,9 +120,9 @@ def handle_regular_answer(
 
     bypass_acl = False
     if (
-        slack_channel_config
-        and slack_channel_config.persona
-        and slack_channel_config.persona.document_sets
+        slack_bot_config
+        and slack_bot_config.persona
+        and slack_bot_config.persona.document_sets
     ):
         # For Slack channels, use the full document set, admin will be warned when configuring it
         # with non-public document sets
@@ -131,8 +131,8 @@ def handle_regular_answer(
     # figure out if we want to use citations or quotes
     use_citations = (
         not DANSWER_BOT_USE_QUOTES
-        if slack_channel_config is None
-        else slack_channel_config.response_type == SlackBotResponseType.CITATIONS
+        if slack_bot_config is None
+        else slack_bot_config.response_type == SlackBotResponseType.CITATIONS
     )
 
     if not message_ts_to_respond_to and not is_bot_msg:
@@ -234,8 +234,8 @@ def handle_regular_answer(
         #     persona.llm_filter_extraction if persona is not None else True
         # )
         auto_detect_filters = (
-            slack_channel_config.enable_auto_filters
-            if slack_channel_config is not None
+            slack_bot_config.enable_auto_filters
+            if slack_bot_config is not None
             else False
         )
         retrieval_details = RetrievalDetails(

backend/danswer/danswerbot/slack/handlers/handle_standard_answers.py

@@ -3,7 +3,7 @@ from sqlalchemy.orm import Session
 
 from danswer.danswerbot.slack.models import SlackMessageInfo
 from danswer.db.models import Prompt
-from danswer.db.models import SlackChannelConfig
+from danswer.db.models import SlackBotConfig
 from danswer.utils.logger import DanswerLoggingAdapter
 from danswer.utils.logger import setup_logger
 from danswer.utils.variable_functionality import fetch_versioned_implementation
@@ -14,7 +14,7 @@ logger = setup_logger()
 def handle_standard_answers(
     message_info: SlackMessageInfo,
     receiver_ids: list[str] | None,
-    slack_channel_config: SlackChannelConfig | None,
+    slack_bot_config: SlackBotConfig | None,
     prompt: Prompt | None,
     logger: DanswerLoggingAdapter,
     client: WebClient,
@@ -29,7 +29,7 @@ def handle_standard_answers(
     return versioned_handle_standard_answers(
         message_info=message_info,
         receiver_ids=receiver_ids,
-        slack_channel_config=slack_channel_config,
+        slack_bot_config=slack_bot_config,
         prompt=prompt,
         logger=logger,
         client=client,
@@ -40,7 +40,7 @@ def handle_standard_answers(
 def _handle_standard_answers(
     message_info: SlackMessageInfo,
     receiver_ids: list[str] | None,
-    slack_channel_config: SlackChannelConfig | None,
+    slack_bot_config: SlackBotConfig | None,
     prompt: Prompt | None,
     logger: DanswerLoggingAdapter,
     client: WebClient,

backend/danswer/danswerbot/slack/listener.py

@@ -4,7 +4,6 @@ import signal
 import sys
 import threading
 import time
-from collections.abc import Callable
 from threading import Event
 from types import FrameType
 from typing import Any
@@ -17,7 +16,6 @@ from prometheus_client import start_http_server
 from slack_sdk import WebClient
 from slack_sdk.socket_mode.request import SocketModeRequest
 from slack_sdk.socket_mode.response import SocketModeResponse
-from sqlalchemy.orm import Session
 
 from danswer.configs.app_configs import POD_NAME
 from danswer.configs.app_configs import POD_NAMESPACE
@@ -27,7 +25,7 @@ from danswer.configs.danswerbot_configs import DANSWER_BOT_REPHRASE_MESSAGE
 from danswer.configs.danswerbot_configs import DANSWER_BOT_RESPOND_EVERY_CHANNEL
 from danswer.configs.danswerbot_configs import NOTIFY_SLACKBOT_NO_ANSWER
 from danswer.connectors.slack.utils import expert_info_from_slack_id
-from danswer.danswerbot.slack.config import get_slack_channel_config_for_bot_and_channel
+from danswer.danswerbot.slack.config import get_slack_bot_config_for_channel
 from danswer.danswerbot.slack.config import MAX_TENANTS_PER_POD
 from danswer.danswerbot.slack.config import TENANT_ACQUISITION_INTERVAL
 from danswer.danswerbot.slack.config import TENANT_HEARTBEAT_EXPIRATION
@@ -56,20 +54,20 @@ from danswer.danswerbot.slack.handlers.handle_message import (
 )
 from danswer.danswerbot.slack.handlers.handle_message import schedule_feedback_reminder
 from danswer.danswerbot.slack.models import SlackMessageInfo
+from danswer.danswerbot.slack.tokens import fetch_tokens
 from danswer.danswerbot.slack.utils import check_message_limit
 from danswer.danswerbot.slack.utils import decompose_action_id
 from danswer.danswerbot.slack.utils import get_channel_name_from_id
-from danswer.danswerbot.slack.utils import get_danswer_bot_slack_bot_id
+from danswer.danswerbot.slack.utils import get_danswer_bot_app_id
 from danswer.danswerbot.slack.utils import read_slack_thread
 from danswer.danswerbot.slack.utils import remove_danswer_bot_tag
 from danswer.danswerbot.slack.utils import rephrase_slack_message
 from danswer.danswerbot.slack.utils import respond_in_thread
 from danswer.danswerbot.slack.utils import TenantSocketModeClient
+from danswer.db.engine import CURRENT_TENANT_ID_CONTEXTVAR
 from danswer.db.engine import get_all_tenant_ids
 from danswer.db.engine import get_session_with_tenant
-from danswer.db.models import SlackBot
 from danswer.db.search_settings import get_current_search_settings
-from danswer.db.slack_bot import fetch_slack_bots
 from danswer.key_value_store.interface import KvKeyNotFoundError
 from danswer.natural_language_processing.search_nlp_models import EmbeddingModel
 from danswer.natural_language_processing.search_nlp_models import warm_up_bi_encoder
@@ -84,8 +82,6 @@ from shared_configs.configs import MODEL_SERVER_HOST
 from shared_configs.configs import MODEL_SERVER_PORT
 from shared_configs.configs import POSTGRES_DEFAULT_SCHEMA
 from shared_configs.configs import SLACK_CHANNEL_ID
-from shared_configs.contextvars import CURRENT_TENANT_ID_CONTEXTVAR
-
 
 logger = setup_logger()
 
@@ -117,10 +113,8 @@ class SlackbotHandler:
     def __init__(self) -> None:
         logger.info("Initializing SlackbotHandler")
         self.tenant_ids: Set[str | None] = set()
-        # The keys for these dictionaries are tuples of (tenant_id, slack_bot_id)
-        self.socket_clients: Dict[tuple[str | None, int], TenantSocketModeClient] = {}
-        self.slack_bot_tokens: Dict[tuple[str | None, int], SlackBotTokens] = {}
-
+        self.socket_clients: Dict[str | None, TenantSocketModeClient] = {}
+        self.slack_bot_tokens: Dict[str | None, SlackBotTokens] = {}
         self.running = True
         self.pod_id = self.get_pod_id()
         self._shutdown_event = Event()
@@ -175,52 +169,6 @@ class SlackbotHandler:
                 logger.exception(f"Error in heartbeat loop: {e}")
             self._shutdown_event.wait(timeout=TENANT_HEARTBEAT_INTERVAL)
 
-    def _manage_clients_per_tenant(
-        self, db_session: Session, tenant_id: str | None, bot: SlackBot
-    ) -> None:
-        slack_bot_tokens = SlackBotTokens(
-            bot_token=bot.bot_token,
-            app_token=bot.app_token,
-        )
-        tenant_bot_pair = (tenant_id, bot.id)
-
-        # If the tokens are not set, we need to close the socket client and delete the tokens
-        # for the tenant and app
-        if not slack_bot_tokens:
-            logger.debug(
-                f"No Slack bot token found for tenant {tenant_id}, bot {bot.id}"
-            )
-            if tenant_bot_pair in self.socket_clients:
-                asyncio.run(self.socket_clients[tenant_bot_pair].close())
-                del self.socket_clients[tenant_bot_pair]
-                del self.slack_bot_tokens[tenant_bot_pair]
-            return
-
-        tokens_exist = tenant_bot_pair in self.slack_bot_tokens
-        tokens_changed = (
-            tokens_exist and slack_bot_tokens != self.slack_bot_tokens[tenant_bot_pair]
-        )
-        if not tokens_exist or tokens_changed:
-            if tokens_exist:
-                logger.info(
-                    f"Slack Bot tokens have changed for tenant {tenant_id}, bot {bot.id} - reconnecting"
-                )
-            else:
-                search_settings = get_current_search_settings(db_session)
-                embedding_model = EmbeddingModel.from_db_model(
-                    search_settings=search_settings,
-                    server_host=MODEL_SERVER_HOST,
-                    server_port=MODEL_SERVER_PORT,
-                )
-                warm_up_bi_encoder(embedding_model=embedding_model)
-
-            self.slack_bot_tokens[tenant_bot_pair] = slack_bot_tokens
-
-            if tenant_bot_pair in self.socket_clients:
-                asyncio.run(self.socket_clients[tenant_bot_pair].close())
-
-            self.start_socket_client(bot.id, tenant_id, slack_bot_tokens)
-
     def acquire_tenants(self) -> None:
         tenant_ids = get_all_tenant_ids()
 
@@ -255,7 +203,6 @@ class SlackbotHandler:
                 continue
 
             logger.debug(f"Acquired lock for tenant {tenant_id}")
-
             self.tenant_ids.add(tenant_id)
 
         for tenant_id in self.tenant_ids:
@@ -265,20 +212,57 @@ class SlackbotHandler:
             try:
                 with get_session_with_tenant(tenant_id) as db_session:
                     try:
-                        bots = fetch_slack_bots(db_session=db_session)
-                        for bot in bots:
-                            self._manage_clients_per_tenant(
-                                db_session=db_session,
-                                tenant_id=tenant_id,
-                                bot=bot,
+                        logger.debug(
+                            f"Setting tenant ID context variable for tenant {tenant_id}"
+                        )
+                        slack_bot_tokens = fetch_tokens()
+                        logger.debug(f"Fetched Slack bot tokens for tenant {tenant_id}")
+                        logger.debug(
+                            f"Reset tenant ID context variable for tenant {tenant_id}"
+                        )
+
+                        if not slack_bot_tokens:
+                            logger.debug(
+                                f"No Slack bot token found for tenant {tenant_id}"
                             )
+                            if tenant_id in self.socket_clients:
+                                asyncio.run(self.socket_clients[tenant_id].close())
+                                del self.socket_clients[tenant_id]
+                                del self.slack_bot_tokens[tenant_id]
+                            continue
+
+                        if (
+                            tenant_id not in self.slack_bot_tokens
+                            or slack_bot_tokens != self.slack_bot_tokens[tenant_id]
+                        ):
+                            if tenant_id in self.slack_bot_tokens:
+                                logger.info(
+                                    f"Slack Bot tokens have changed for tenant {tenant_id} - reconnecting"
+                                )
+                            else:
+                                search_settings = get_current_search_settings(
+                                    db_session
+                                )
+                                embedding_model = EmbeddingModel.from_db_model(
+                                    search_settings=search_settings,
+                                    server_host=MODEL_SERVER_HOST,
+                                    server_port=MODEL_SERVER_PORT,
+                                )
+                                warm_up_bi_encoder(embedding_model=embedding_model)
+
+                            self.slack_bot_tokens[tenant_id] = slack_bot_tokens
+
+                            if self.socket_clients.get(tenant_id):
+                                asyncio.run(self.socket_clients[tenant_id].close())
+
+                            self.start_socket_client(tenant_id, slack_bot_tokens)
 
                     except KvKeyNotFoundError:
                         logger.debug(f"Missing Slack Bot tokens for tenant {tenant_id}")
-                        if (tenant_id, bot.id) in self.socket_clients:
-                            asyncio.run(self.socket_clients[tenant_id, bot.id].close())
-                            del self.socket_clients[tenant_id, bot.id]
-                            del self.slack_bot_tokens[tenant_id, bot.id]
+                        if self.socket_clients.get(tenant_id):
+                            asyncio.run(self.socket_clients[tenant_id].close())
+                            del self.socket_clients[tenant_id]
+                            del self.slack_bot_tokens[tenant_id]
                     except Exception as e:
                         logger.exception(f"Error handling tenant {tenant_id}: {e}")
             finally:
@@ -297,37 +281,26 @@ class SlackbotHandler:
             )
 
     def start_socket_client(
-        self, slack_bot_id: int, tenant_id: str | None, slack_bot_tokens: SlackBotTokens
+        self, tenant_id: str | None, slack_bot_tokens: SlackBotTokens
     ) -> None:
-        logger.info(
-            f"Starting socket client for tenant: {tenant_id}, app: {slack_bot_id}"
-        )
-        socket_client: TenantSocketModeClient = _get_socket_client(
-            slack_bot_tokens, tenant_id, slack_bot_id
-        )
+        logger.info(f"Starting socket client for tenant {tenant_id}")
+        socket_client = _get_socket_client(slack_bot_tokens, tenant_id)
 
         # Append the event handler
-        process_slack_event = create_process_slack_event()
         socket_client.socket_mode_request_listeners.append(process_slack_event)  # type: ignore
 
         # Establish a WebSocket connection to the Socket Mode servers
-        logger.info(
-            f"Connecting socket client for tenant: {tenant_id}, app: {slack_bot_id}"
-        )
+        logger.info(f"Connecting socket client for tenant {tenant_id}")
         socket_client.connect()
-        self.socket_clients[tenant_id, slack_bot_id] = socket_client
-        self.tenant_ids.add(tenant_id)
-        logger.info(
-            f"Started SocketModeClient for tenant: {tenant_id}, app: {slack_bot_id}"
-        )
+        self.socket_clients[tenant_id] = socket_client
+        logger.info(f"Started SocketModeClient for tenant {tenant_id}")
 
     def stop_socket_clients(self) -> None:
         logger.info(f"Stopping {len(self.socket_clients)} socket clients")
-        for (tenant_id, slack_bot_id), client in self.socket_clients.items():
-            asyncio.run(client.close())
-            logger.info(
-                f"Stopped SocketModeClient for tenant: {tenant_id}, app: {slack_bot_id}"
-            )
+        for tenant_id, client in self.socket_clients.items():
+            if client:
+                asyncio.run(client.close())
+                logger.info(f"Stopped SocketModeClient for tenant {tenant_id}")
 
     def shutdown(self, signum: int | None, frame: FrameType | None) -> None:
         if not self.running:
@@ -411,7 +384,7 @@ def prefilter_requests(req: SocketModeRequest, client: TenantSocketModeClient) -
             )
             return False
 
-        bot_tag_id = get_danswer_bot_slack_bot_id(client.web_client)
+        bot_tag_id = get_danswer_bot_app_id(client.web_client)
         if event_type == "message":
             is_dm = event.get("channel_type") == "im"
             is_tagged = bot_tag_id and bot_tag_id in msg
@@ -434,15 +407,13 @@ def prefilter_requests(req: SocketModeRequest, client: TenantSocketModeClient) -
             )
 
             with get_session_with_tenant(client.tenant_id) as db_session:
-                slack_channel_config = get_slack_channel_config_for_bot_and_channel(
-                    db_session=db_session,
-                    slack_bot_id=client.slack_bot_id,
-                    channel_name=channel_name,
+                slack_bot_config = get_slack_bot_config_for_channel(
+                    channel_name=channel_name, db_session=db_session
                 )
             # If DanswerBot is not specifically tagged and the channel is not set to respond to bots, ignore the message
             if (not bot_tag_id or bot_tag_id not in msg) and (
-                not slack_channel_config
-                or not slack_channel_config.channel_config.get("respond_to_bots")
+                not slack_bot_config
+                or not slack_bot_config.channel_config.get("respond_to_bots")
             ):
                 channel_specific_logger.info("Ignoring message from bot")
                 return False
@@ -647,16 +618,14 @@ def process_message(
         token = CURRENT_TENANT_ID_CONTEXTVAR.set(client.tenant_id)
     try:
         with get_session_with_tenant(client.tenant_id) as db_session:
-            slack_channel_config = get_slack_channel_config_for_bot_and_channel(
-                db_session=db_session,
-                slack_bot_id=client.slack_bot_id,
-                channel_name=channel_name,
+            slack_bot_config = get_slack_bot_config_for_channel(
+                channel_name=channel_name, db_session=db_session
             )
 
             # Be careful about this default, don't want to accidentally spam every channel
             # Users should be able to DM slack bot in their private channels though
             if (
-                slack_channel_config is None
+                slack_bot_config is None
                 and not respond_every_channel
                 # Can't have configs for DMs so don't toss them out
                 and not is_dm
@@ -667,10 +636,9 @@ def process_message(
                 return
 
             follow_up = bool(
-                slack_channel_config
-                and slack_channel_config.channel_config
-                and slack_channel_config.channel_config.get("follow_up_tags")
-                is not None
+                slack_bot_config
+                and slack_bot_config.channel_config
+                and slack_bot_config.channel_config.get("follow_up_tags") is not None
             )
             feedback_reminder_id = schedule_feedback_reminder(
                 details=details, client=client.web_client, include_followup=follow_up
@@ -678,7 +646,7 @@ def process_message(
 
             failed = handle_message(
                 message_info=details,
-                slack_channel_config=slack_channel_config,
+                slack_bot_config=slack_bot_config,
                 client=client.web_client,
                 feedback_reminder_id=feedback_reminder_id,
                 tenant_id=client.tenant_id,
@@ -730,32 +698,26 @@ def view_routing(req: SocketModeRequest, client: TenantSocketModeClient) -> None
             return process_feedback(req, client)
 
 
-def create_process_slack_event() -> (
-    Callable[[TenantSocketModeClient, SocketModeRequest], None]
-):
-    def process_slack_event(
-        client: TenantSocketModeClient, req: SocketModeRequest
-    ) -> None:
-        # Always respond right away, if Slack doesn't receive these frequently enough
-        # it will assume the Bot is DEAD!!! :(
-        acknowledge_message(req, client)
+def process_slack_event(client: TenantSocketModeClient, req: SocketModeRequest) -> None:
+    # Always respond right away, if Slack doesn't receive these frequently enough
+    # it will assume the Bot is DEAD!!! :(
+    acknowledge_message(req, client)
 
-        try:
-            if req.type == "interactive":
-                if req.payload.get("type") == "block_actions":
-                    return action_routing(req, client)
-                elif req.payload.get("type") == "view_submission":
-                    return view_routing(req, client)
-            elif req.type == "events_api" or req.type == "slash_commands":
-                return process_message(req, client)
-        except Exception:
-            logger.exception("Failed to process slack event")
-
-    return process_slack_event
+    try:
+        if req.type == "interactive":
+            if req.payload.get("type") == "block_actions":
+                return action_routing(req, client)
+            elif req.payload.get("type") == "view_submission":
+                return view_routing(req, client)
+        elif req.type == "events_api" or req.type == "slash_commands":
+            return process_message(req, client)
+    except Exception as e:
+        logger.exception(f"Failed to process slack event. Error: {e}")
+        logger.error(f"Slack request payload: {req.payload}")
 
 
 def _get_socket_client(
-    slack_bot_tokens: SlackBotTokens, tenant_id: str | None, slack_bot_id: int
+    slack_bot_tokens: SlackBotTokens, tenant_id: str | None
 ) -> TenantSocketModeClient:
     # For more info on how to set this up, checkout the docs:
     # https://docs.danswer.dev/slack_bot_setup
@@ -764,7 +726,6 @@ def _get_socket_client(
         app_token=slack_bot_tokens.app_token,
         web_client=WebClient(token=slack_bot_tokens.bot_token),
         tenant_id=tenant_id,
-        slack_bot_id=slack_bot_id,
     )
 
 

backend/danswer/danswerbot/slack/tokens.py

@@ -0,0 +1,28 @@
+import os
+from typing import cast
+
+from danswer.configs.constants import KV_SLACK_BOT_TOKENS_CONFIG_KEY
+from danswer.key_value_store.factory import get_kv_store
+from danswer.server.manage.models import SlackBotTokens
+
+
+def fetch_tokens() -> SlackBotTokens:
+    # first check env variables
+    app_token = os.environ.get("DANSWER_BOT_SLACK_APP_TOKEN")
+    bot_token = os.environ.get("DANSWER_BOT_SLACK_BOT_TOKEN")
+    if app_token and bot_token:
+        return SlackBotTokens(app_token=app_token, bot_token=bot_token)
+
+    dynamic_config_store = get_kv_store()
+    return SlackBotTokens(
+        **cast(dict, dynamic_config_store.load(key=KV_SLACK_BOT_TOKENS_CONFIG_KEY))
+    )
+
+
+def save_tokens(
+    tokens: SlackBotTokens,
+) -> None:
+    dynamic_config_store = get_kv_store()
+    dynamic_config_store.store(
+        key=KV_SLACK_BOT_TOKENS_CONFIG_KEY, val=dict(tokens), encrypt=True
+    )

backend/danswer/danswerbot/slack/utils.py

@@ -30,6 +30,7 @@ from danswer.configs.danswerbot_configs import (
 from danswer.connectors.slack.utils import make_slack_api_rate_limited
 from danswer.connectors.slack.utils import SlackTextCleaner
 from danswer.danswerbot.slack.constants import FeedbackVisibility
+from danswer.danswerbot.slack.tokens import fetch_tokens
 from danswer.db.engine import get_session_with_tenant
 from danswer.db.users import get_user_by_email
 from danswer.llm.exceptions import GenAIDisabledException
@@ -46,16 +47,16 @@ from danswer.utils.text_processing import replace_whitespaces_w_space
 logger = setup_logger()
 
 
-_DANSWER_BOT_SLACK_BOT_ID: str | None = None
+_DANSWER_BOT_APP_ID: str | None = None
 _DANSWER_BOT_MESSAGE_COUNT: int = 0
 _DANSWER_BOT_COUNT_START_TIME: float = time.time()
 
 
-def get_danswer_bot_slack_bot_id(web_client: WebClient) -> Any:
-    global _DANSWER_BOT_SLACK_BOT_ID
-    if _DANSWER_BOT_SLACK_BOT_ID is None:
-        _DANSWER_BOT_SLACK_BOT_ID = web_client.auth_test().get("user_id")
-    return _DANSWER_BOT_SLACK_BOT_ID
+def get_danswer_bot_app_id(web_client: WebClient) -> Any:
+    global _DANSWER_BOT_APP_ID
+    if _DANSWER_BOT_APP_ID is None:
+        _DANSWER_BOT_APP_ID = web_client.auth_test().get("user_id")
+    return _DANSWER_BOT_APP_ID
 
 
 def check_message_limit() -> bool:
@@ -136,10 +137,15 @@ def update_emote_react(
 
 
 def remove_danswer_bot_tag(message_str: str, client: WebClient) -> str:
-    bot_tag_id = get_danswer_bot_slack_bot_id(web_client=client)
+    bot_tag_id = get_danswer_bot_app_id(web_client=client)
     return re.sub(rf"<@{bot_tag_id}>\s", "", message_str)
 
 
+def get_web_client() -> WebClient:
+    slack_tokens = fetch_tokens()
+    return WebClient(token=slack_tokens.bot_token)
+
+
 @retry(
     tries=DANSWER_BOT_NUM_RETRIES,
     delay=0.25,
@@ -431,9 +437,9 @@ def read_slack_thread(
             )
             message_type = MessageType.USER
         else:
-            self_slack_bot_id = get_danswer_bot_slack_bot_id(client)
+            self_app_id = get_danswer_bot_app_id(client)
 
-            if reply.get("user") == self_slack_bot_id:
+            if reply.get("user") == self_app_id:
                 # DanswerBot response
                 message_type = MessageType.ASSISTANT
                 user_sem_id = "Assistant"
@@ -576,9 +582,6 @@ def get_feedback_visibility() -> FeedbackVisibility:
 
 
 class TenantSocketModeClient(SocketModeClient):
-    def __init__(
-        self, tenant_id: str | None, slack_bot_id: int, *args: Any, **kwargs: Any
-    ):
+    def __init__(self, tenant_id: str | None, *args: Any, **kwargs: Any):
         super().__init__(*args, **kwargs)
         self.tenant_id = tenant_id
-        self.slack_bot_id = slack_bot_id

backend/danswer/db/auth.py

@@ -4,7 +4,6 @@ from typing import Any
 from typing import Dict
 
 from fastapi import Depends
-from fastapi_users.models import ID
 from fastapi_users.models import UP
 from fastapi_users_db_sqlalchemy import SQLAlchemyUserDatabase
 from fastapi_users_db_sqlalchemy.access_token import SQLAlchemyAccessTokenDatabase
@@ -44,10 +43,7 @@ def get_total_users_count(db_session: Session) -> int:
     """
     user_count = (
         db_session.query(User)
-        .filter(
-            ~User.email.endswith(get_api_key_email_pattern()),  # type: ignore
-            User.role != UserRole.EXT_PERM_USER,
-        )
+        .filter(~User.email.endswith(get_api_key_email_pattern()))  # type: ignore
         .count()
     )
     invited_users = len(get_invited_users())
@@ -65,7 +61,7 @@ async def get_user_count() -> int:
 
 
 # Need to override this because FastAPI Users doesn't give flexibility for backend field creation logic in OAuth flow
-class SQLAlchemyUserAdminDB(SQLAlchemyUserDatabase[UP, ID]):
+class SQLAlchemyUserAdminDB(SQLAlchemyUserDatabase):
     async def create(
         self,
         create_dict: Dict[str, Any],

backend/danswer/db/document.py

@@ -209,7 +209,6 @@ def get_document_connector_counts(
 def get_document_counts_for_cc_pairs(
     db_session: Session, cc_pair_identifiers: list[ConnectorCredentialPairIdentifier]
 ) -> Sequence[tuple[int, int, int]]:
-    """Returns a sequence of tuples of (connector_id, credential_id, document count)"""
     stmt = (
         select(
             DocumentByConnectorCredentialPair.connector_id,
@@ -309,7 +308,7 @@ def get_access_info_for_documents(
     return db_session.execute(stmt).all()  # type: ignore
 
 
-def upsert_documents(
+def _upsert_documents(
     db_session: Session,
     document_metadata_batch: list[DocumentMetadata],
     initial_boost: int = DEFAULT_BOOST,
@@ -365,24 +364,24 @@ def upsert_documents(
     db_session.commit()
 
 
-def upsert_document_by_connector_credential_pair(
-    db_session: Session, connector_id: int, credential_id: int, document_ids: list[str]
+def _upsert_document_by_connector_credential_pair(
+    db_session: Session, document_metadata_batch: list[DocumentMetadata]
 ) -> None:
     """NOTE: this function is Postgres specific. Not all DBs support the ON CONFLICT clause."""
-    if not document_ids:
-        logger.info("`document_ids` is empty. Skipping.")
+    if not document_metadata_batch:
+        logger.info("`document_metadata_batch` is empty. Skipping.")
         return
 
     insert_stmt = insert(DocumentByConnectorCredentialPair).values(
         [
             model_to_dict(
                 DocumentByConnectorCredentialPair(
-                    id=doc_id,
-                    connector_id=connector_id,
-                    credential_id=credential_id,
+                    id=document_metadata.document_id,
+                    connector_id=document_metadata.connector_id,
+                    credential_id=document_metadata.credential_id,
                 )
             )
-            for doc_id in document_ids
+            for document_metadata in document_metadata_batch
         ]
     )
     # for now, there are no columns to update. If more metadata is added, then this
@@ -443,6 +442,17 @@ def mark_document_as_synced(document_id: str, db_session: Session) -> None:
     db_session.commit()
 
 
+def upsert_documents_complete(
+    db_session: Session,
+    document_metadata_batch: list[DocumentMetadata],
+) -> None:
+    _upsert_documents(db_session, document_metadata_batch)
+    _upsert_document_by_connector_credential_pair(db_session, document_metadata_batch)
+    logger.info(
+        f"Upserted {len(document_metadata_batch)} document store entries into DB"
+    )
+
+
 def delete_document_by_connector_credential_pair__no_commit(
     db_session: Session,
     document_id: str,
@@ -551,7 +561,7 @@ def prepare_to_modify_documents(
     db_session.commit()  # ensure that we're not in a transaction
 
     lock_acquired = False
-    for i in range(_NUM_LOCK_ATTEMPTS):
+    for _ in range(_NUM_LOCK_ATTEMPTS):
         try:
             with db_session.begin() as transaction:
                 lock_acquired = acquire_document_locks(
@@ -562,7 +572,7 @@ def prepare_to_modify_documents(
                     break
         except OperationalError as e:
             logger.warning(
-                f"Failed to acquire locks for documents on attempt {i}, retrying. Error: {e}"
+                f"Failed to acquire locks for documents, retrying. Error: {e}"
             )
 
         time.sleep(retry_delay)

backend/danswer/db/index_attempt.py

@@ -67,13 +67,6 @@ def create_index_attempt(
     return new_attempt.id
 
 
-def delete_index_attempt(db_session: Session, index_attempt_id: int) -> None:
-    index_attempt = get_index_attempt(db_session, index_attempt_id)
-    if index_attempt:
-        db_session.delete(index_attempt)
-        db_session.commit()
-
-
 def mock_successful_index_attempt(
     connector_credential_pair_id: int,
     search_settings_id: int,

backend/danswer/db/models.py

@@ -53,7 +53,7 @@ from danswer.db.enums import IndexingStatus
 from danswer.db.enums import IndexModelStatus
 from danswer.db.enums import TaskStatus
 from danswer.db.pydantic_type import PydanticType
-from danswer.utils.special_types import JSON_ro
+from danswer.key_value_store.interface import JSON_ro
 from danswer.file_store.models import FileDescriptor
 from danswer.llm.override_models import LLMOverride
 from danswer.llm.override_models import PromptOverride
@@ -171,6 +171,8 @@ class User(SQLAlchemyBaseUserTableUUID, Base):
     notifications: Mapped[list["Notification"]] = relationship(
         "Notification", back_populates="user"
     )
+    # Whether the user has logged in via web. False if user has only used Danswer through Slack bot
+    has_web_login: Mapped[bool] = mapped_column(Boolean, default=True)
     cc_pairs: Mapped[list["ConnectorCredentialPair"]] = relationship(
         "ConnectorCredentialPair",
         back_populates="creator",
@@ -350,11 +352,11 @@ class StandardAnswer__StandardAnswerCategory(Base):
     )
 
 
-class SlackChannelConfig__StandardAnswerCategory(Base):
-    __tablename__ = "slack_channel_config__standard_answer_category"
+class SlackBotConfig__StandardAnswerCategory(Base):
+    __tablename__ = "slack_bot_config__standard_answer_category"
 
-    slack_channel_config_id: Mapped[int] = mapped_column(
-        ForeignKey("slack_channel_config.id"), primary_key=True
+    slack_bot_config_id: Mapped[int] = mapped_column(
+        ForeignKey("slack_bot_config.id"), primary_key=True
     )
     standard_answer_category_id: Mapped[int] = mapped_column(
         ForeignKey("standard_answer_category.id"), primary_key=True
@@ -1181,7 +1183,7 @@ class LLMProvider(Base):
     default_model_name: Mapped[str] = mapped_column(String)
     fast_default_model_name: Mapped[str | None] = mapped_column(String, nullable=True)
 
-    # Models to actually display to users
+    # Models to actually disp;aly to users
     # If nulled out, we assume in the application logic we should present all
     display_model_names: Mapped[list[str] | None] = mapped_column(
         postgresql.ARRAY(String), nullable=True
@@ -1363,9 +1365,6 @@ class Persona(Base):
     recency_bias: Mapped[RecencyBiasSetting] = mapped_column(
         Enum(RecencyBiasSetting, native_enum=False)
     )
-    category_id: Mapped[int | None] = mapped_column(
-        ForeignKey("persona_category.id"), nullable=True
-    )
     # Allows the Persona to specify a different LLM version than is controlled
     # globablly via env variables. For flexibility, validity is not currently enforced
     # NOTE: only is applied on the actual response generation - is not used for things like
@@ -1437,9 +1436,6 @@ class Persona(Base):
         secondary="persona__user_group",
         viewonly=True,
     )
-    category: Mapped["PersonaCategory"] = relationship(
-        "PersonaCategory", back_populates="personas"
-    )
 
     # Default personas loaded via yaml cannot have the same name
     __table_args__ = (
@@ -1452,17 +1448,6 @@ class Persona(Base):
     )
 
 
-class PersonaCategory(Base):
-    __tablename__ = "persona_category"
-
-    id: Mapped[int] = mapped_column(primary_key=True)
-    name: Mapped[str] = mapped_column(String, unique=True)
-    description: Mapped[str | None] = mapped_column(String, nullable=True)
-    personas: Mapped[list["Persona"]] = relationship(
-        "Persona", back_populates="category"
-    )
-
-
 AllowedAnswerFilters = (
     Literal["well_answered_postfilter"] | Literal["questionmark_prefilter"]
 )
@@ -1472,7 +1457,7 @@ class ChannelConfig(TypedDict):
     """NOTE: is a `TypedDict` so it can be used as a type hint for a JSONB column
     in Postgres"""
 
-    channel_name: str
+    channel_names: list[str]
     respond_tag_only: NotRequired[bool]  # defaults to False
     respond_to_bots: NotRequired[bool]  # defaults to False
     respond_member_group_list: NotRequired[list[str]]
@@ -1487,11 +1472,10 @@ class SlackBotResponseType(str, PyEnum):
     CITATIONS = "citations"
 
 
-class SlackChannelConfig(Base):
-    __tablename__ = "slack_channel_config"
+class SlackBotConfig(Base):
+    __tablename__ = "slack_bot_config"
 
     id: Mapped[int] = mapped_column(primary_key=True)
-    slack_bot_id: Mapped[int] = mapped_column(ForeignKey("slack_bot.id"), nullable=True)
     persona_id: Mapped[int | None] = mapped_column(
         ForeignKey("persona.id"), nullable=True
     )
@@ -1508,30 +1492,10 @@ class SlackChannelConfig(Base):
     )
 
     persona: Mapped[Persona | None] = relationship("Persona")
-    slack_bot: Mapped["SlackBot"] = relationship(
-        "SlackBot",
-        back_populates="slack_channel_configs",
-    )
     standard_answer_categories: Mapped[list["StandardAnswerCategory"]] = relationship(
         "StandardAnswerCategory",
-        secondary=SlackChannelConfig__StandardAnswerCategory.__table__,
-        back_populates="slack_channel_configs",
-    )
-
-
-class SlackBot(Base):
-    __tablename__ = "slack_bot"
-
-    id: Mapped[int] = mapped_column(primary_key=True)
-    name: Mapped[str] = mapped_column(String)
-    enabled: Mapped[bool] = mapped_column(Boolean, default=True)
-
-    bot_token: Mapped[str] = mapped_column(EncryptedString(), unique=True)
-    app_token: Mapped[str] = mapped_column(EncryptedString(), unique=True)
-
-    slack_channel_configs: Mapped[list[SlackChannelConfig]] = relationship(
-        "SlackChannelConfig",
-        back_populates="slack_bot",
+        secondary=SlackBotConfig__StandardAnswerCategory.__table__,
+        back_populates="slack_bot_configs",
     )
 
 
@@ -1770,9 +1734,9 @@ class StandardAnswerCategory(Base):
         secondary=StandardAnswer__StandardAnswerCategory.__table__,
         back_populates="categories",
     )
-    slack_channel_configs: Mapped[list["SlackChannelConfig"]] = relationship(
-        "SlackChannelConfig",
-        secondary=SlackChannelConfig__StandardAnswerCategory.__table__,
+    slack_bot_configs: Mapped[list["SlackBotConfig"]] = relationship(
+        "SlackBotConfig",
+        secondary=SlackBotConfig__StandardAnswerCategory.__table__,
         back_populates="standard_answer_categories",
     )
 

backend/danswer/db/persona.py

@@ -26,7 +26,6 @@ from danswer.db.models import DocumentSet
 from danswer.db.models import Persona
 from danswer.db.models import Persona__User
 from danswer.db.models import Persona__UserGroup
-from danswer.db.models import PersonaCategory
 from danswer.db.models import Prompt
 from danswer.db.models import StarterMessage
 from danswer.db.models import Tool
@@ -259,6 +258,7 @@ def get_personas(
 ) -> Sequence[Persona]:
     stmt = select(Persona).distinct()
     stmt = _add_user_filters(stmt=stmt, user=user, get_editable=get_editable)
+
     if not include_default:
         stmt = stmt.where(Persona.builtin_persona.is_(False))
     if not include_slack_bot_personas:
@@ -417,7 +417,6 @@ def upsert_persona(
     search_start_date: datetime | None = None,
     builtin_persona: bool = False,
     is_default_persona: bool = False,
-    category_id: int | None = None,
     chunks_above: int = CONTEXT_CHUNKS_ABOVE,
     chunks_below: int = CONTEXT_CHUNKS_BELOW,
 ) -> Persona:
@@ -488,7 +487,7 @@ def upsert_persona(
         persona.is_visible = is_visible
         persona.search_start_date = search_start_date
         persona.is_default_persona = is_default_persona
-        persona.category_id = category_id
+
         # Do not delete any associations manually added unless
         # a new updated list is provided
         if document_sets is not None:
@@ -529,7 +528,6 @@ def upsert_persona(
             is_visible=is_visible,
             search_start_date=search_start_date,
             is_default_persona=is_default_persona,
-            category_id=category_id,
         )
         db_session.add(persona)
 
@@ -746,39 +744,3 @@ def delete_persona_by_name(
 
     db_session.execute(stmt)
     db_session.commit()
-
-
-def get_assistant_categories(db_session: Session) -> list[PersonaCategory]:
-    return db_session.query(PersonaCategory).all()
-
-
-def create_assistant_category(
-    db_session: Session, name: str, description: str
-) -> PersonaCategory:
-    category = PersonaCategory(name=name, description=description)
-    db_session.add(category)
-    db_session.commit()
-    return category
-
-
-def update_persona_category(
-    category_id: int,
-    category_description: str,
-    category_name: str,
-    db_session: Session,
-) -> None:
-    persona_category = (
-        db_session.query(PersonaCategory)
-        .filter(PersonaCategory.id == category_id)
-        .one_or_none()
-    )
-    if persona_category is None:
-        raise ValueError(f"Persona category with ID {category_id} does not exist")
-    persona_category.description = category_description
-    persona_category.name = category_name
-    db_session.commit()
-
-
-def delete_persona_category(category_id: int, db_session: Session) -> None:
-    db_session.query(PersonaCategory).filter(PersonaCategory.id == category_id).delete()
-    db_session.commit()

backend/danswer/db/slack_bot.py

@@ -1,76 +0,0 @@
-from collections.abc import Sequence
-
-from sqlalchemy import select
-from sqlalchemy.orm import Session
-
-from danswer.db.models import SlackBot
-
-
-def insert_slack_bot(
-    db_session: Session,
-    name: str,
-    enabled: bool,
-    bot_token: str,
-    app_token: str,
-) -> SlackBot:
-    slack_bot = SlackBot(
-        name=name,
-        enabled=enabled,
-        bot_token=bot_token,
-        app_token=app_token,
-    )
-    db_session.add(slack_bot)
-    db_session.commit()
-
-    return slack_bot
-
-
-def update_slack_bot(
-    db_session: Session,
-    slack_bot_id: int,
-    name: str,
-    enabled: bool,
-    bot_token: str,
-    app_token: str,
-) -> SlackBot:
-    slack_bot = db_session.scalar(select(SlackBot).where(SlackBot.id == slack_bot_id))
-    if slack_bot is None:
-        raise ValueError(f"Unable to find Slack Bot with ID {slack_bot_id}")
-
-    # update the app
-    slack_bot.name = name
-    slack_bot.enabled = enabled
-    slack_bot.bot_token = bot_token
-    slack_bot.app_token = app_token
-
-    db_session.commit()
-
-    return slack_bot
-
-
-def fetch_slack_bot(
-    db_session: Session,
-    slack_bot_id: int,
-) -> SlackBot:
-    slack_bot = db_session.scalar(select(SlackBot).where(SlackBot.id == slack_bot_id))
-    if slack_bot is None:
-        raise ValueError(f"Unable to find Slack Bot with ID {slack_bot_id}")
-
-    return slack_bot
-
-
-def remove_slack_bot(
-    db_session: Session,
-    slack_bot_id: int,
-) -> None:
-    slack_bot = fetch_slack_bot(
-        db_session=db_session,
-        slack_bot_id=slack_bot_id,
-    )
-
-    db_session.delete(slack_bot)
-    db_session.commit()
-
-
-def fetch_slack_bots(db_session: Session) -> Sequence[SlackBot]:
-    return db_session.scalars(select(SlackBot)).all()

backend/danswer/db/slack_bot_config.py

@@ -9,8 +9,8 @@ from danswer.db.constants import SLACK_BOT_PERSONA_PREFIX
 from danswer.db.models import ChannelConfig
 from danswer.db.models import Persona
 from danswer.db.models import Persona__DocumentSet
+from danswer.db.models import SlackBotConfig
 from danswer.db.models import SlackBotResponseType
-from danswer.db.models import SlackChannelConfig
 from danswer.db.models import User
 from danswer.db.persona import get_default_prompt
 from danswer.db.persona import mark_persona_as_deleted
@@ -22,8 +22,8 @@ from danswer.utils.variable_functionality import (
 )
 
 
-def _build_persona_name(channel_name: str) -> str:
-    return f"{SLACK_BOT_PERSONA_PREFIX}{channel_name}"
+def _build_persona_name(channel_names: list[str]) -> str:
+    return f"{SLACK_BOT_PERSONA_PREFIX}{'-'.join(channel_names)}"
 
 
 def _cleanup_relationships(db_session: Session, persona_id: int) -> None:
@@ -38,9 +38,9 @@ def _cleanup_relationships(db_session: Session, persona_id: int) -> None:
         db_session.delete(rel)
 
 
-def create_slack_channel_persona(
+def create_slack_bot_persona(
     db_session: Session,
-    channel_name: str,
+    channel_names: list[str],
     document_set_ids: list[int],
     existing_persona_id: int | None = None,
     num_chunks: float = MAX_CHUNKS_FED_TO_CHAT,
@@ -48,11 +48,11 @@ def create_slack_channel_persona(
 ) -> Persona:
     """NOTE: does not commit changes"""
 
-    # create/update persona associated with the Slack channel
-    persona_name = _build_persona_name(channel_name)
+    # create/update persona associated with the slack bot
+    persona_name = _build_persona_name(channel_names)
     default_prompt = get_default_prompt(db_session)
     persona = upsert_persona(
-        user=None,  # Slack channel Personas are not attached to users
+        user=None,  # Slack Bot Personas are not attached to users
         persona_id=existing_persona_id,
         name=persona_name,
         description="",
@@ -78,15 +78,14 @@ def _no_ee_standard_answer_categories(*args: Any, **kwargs: Any) -> list:
     return []
 
 
-def insert_slack_channel_config(
-    db_session: Session,
-    slack_bot_id: int,
+def insert_slack_bot_config(
     persona_id: int | None,
     channel_config: ChannelConfig,
     response_type: SlackBotResponseType,
     standard_answer_category_ids: list[int],
     enable_auto_filters: bool,
-) -> SlackChannelConfig:
+    db_session: Session,
+) -> SlackBotConfig:
     versioned_fetch_standard_answer_categories_by_ids = (
         fetch_versioned_implementation_with_fallback(
             "danswer.db.standard_answer",
@@ -111,37 +110,34 @@ def insert_slack_channel_config(
                 f"Some or all categories with ids {standard_answer_category_ids} do not exist"
             )
 
-    slack_channel_config = SlackChannelConfig(
-        slack_bot_id=slack_bot_id,
+    slack_bot_config = SlackBotConfig(
         persona_id=persona_id,
         channel_config=channel_config,
         response_type=response_type,
         standard_answer_categories=existing_standard_answer_categories,
         enable_auto_filters=enable_auto_filters,
     )
-    db_session.add(slack_channel_config)
+    db_session.add(slack_bot_config)
     db_session.commit()
 
-    return slack_channel_config
+    return slack_bot_config
 
 
-def update_slack_channel_config(
-    db_session: Session,
-    slack_channel_config_id: int,
+def update_slack_bot_config(
+    slack_bot_config_id: int,
     persona_id: int | None,
     channel_config: ChannelConfig,
     response_type: SlackBotResponseType,
     standard_answer_category_ids: list[int],
     enable_auto_filters: bool,
-) -> SlackChannelConfig:
-    slack_channel_config = db_session.scalar(
-        select(SlackChannelConfig).where(
-            SlackChannelConfig.id == slack_channel_config_id
-        )
+    db_session: Session,
+) -> SlackBotConfig:
+    slack_bot_config = db_session.scalar(
+        select(SlackBotConfig).where(SlackBotConfig.id == slack_bot_config_id)
     )
-    if slack_channel_config is None:
+    if slack_bot_config is None:
         raise ValueError(
-            f"Unable to find Slack channel config with ID {slack_channel_config_id}"
+            f"Unable to find slack bot config with ID {slack_bot_config_id}"
         )
 
     versioned_fetch_standard_answer_categories_by_ids = (
@@ -163,25 +159,25 @@ def update_slack_channel_config(
         )
 
     # get the existing persona id before updating the object
-    existing_persona_id = slack_channel_config.persona_id
+    existing_persona_id = slack_bot_config.persona_id
 
     # update the config
     # NOTE: need to do this before cleaning up the old persona or else we
     # will encounter `violates foreign key constraint` errors
-    slack_channel_config.persona_id = persona_id
-    slack_channel_config.channel_config = channel_config
-    slack_channel_config.response_type = response_type
-    slack_channel_config.standard_answer_categories = list(
+    slack_bot_config.persona_id = persona_id
+    slack_bot_config.channel_config = channel_config
+    slack_bot_config.response_type = response_type
+    slack_bot_config.standard_answer_categories = list(
         existing_standard_answer_categories
     )
-    slack_channel_config.enable_auto_filters = enable_auto_filters
+    slack_bot_config.enable_auto_filters = enable_auto_filters
 
     # if the persona has changed, then clean up the old persona
     if persona_id != existing_persona_id and existing_persona_id:
         existing_persona = db_session.scalar(
             select(Persona).where(Persona.id == existing_persona_id)
         )
-        # if the existing persona was one created just for use with this Slack channel,
+        # if the existing persona was one created just for use with this Slack Bot,
         # then clean it up
         if existing_persona and existing_persona.name.startswith(
             SLACK_BOT_PERSONA_PREFIX
@@ -192,30 +188,28 @@ def update_slack_channel_config(
 
     db_session.commit()
 
-    return slack_channel_config
+    return slack_bot_config
 
 
-def remove_slack_channel_config(
-    db_session: Session,
-    slack_channel_config_id: int,
+def remove_slack_bot_config(
+    slack_bot_config_id: int,
     user: User | None,
+    db_session: Session,
 ) -> None:
-    slack_channel_config = db_session.scalar(
-        select(SlackChannelConfig).where(
-            SlackChannelConfig.id == slack_channel_config_id
-        )
+    slack_bot_config = db_session.scalar(
+        select(SlackBotConfig).where(SlackBotConfig.id == slack_bot_config_id)
     )
-    if slack_channel_config is None:
+    if slack_bot_config is None:
         raise ValueError(
-            f"Unable to find Slack channel config with ID {slack_channel_config_id}"
+            f"Unable to find slack bot config with ID {slack_bot_config_id}"
         )
 
-    existing_persona_id = slack_channel_config.persona_id
+    existing_persona_id = slack_bot_config.persona_id
     if existing_persona_id:
         existing_persona = db_session.scalar(
             select(Persona).where(Persona.id == existing_persona_id)
         )
-        # if the existing persona was one created just for use with this Slack channel,
+        # if the existing persona was one created just for use with this Slack Bot,
         # then clean it up
         if existing_persona and existing_persona.name.startswith(
             SLACK_BOT_PERSONA_PREFIX
@@ -227,28 +221,17 @@ def remove_slack_channel_config(
                 persona_id=existing_persona_id, user=user, db_session=db_session
             )
 
-    db_session.delete(slack_channel_config)
+    db_session.delete(slack_bot_config)
     db_session.commit()
 
 
-def fetch_slack_channel_configs(
-    db_session: Session, slack_bot_id: int | None = None
-) -> Sequence[SlackChannelConfig]:
-    if not slack_bot_id:
-        return db_session.scalars(select(SlackChannelConfig)).all()
-
-    return db_session.scalars(
-        select(SlackChannelConfig).where(
-            SlackChannelConfig.slack_bot_id == slack_bot_id
-        )
-    ).all()
-
-
-def fetch_slack_channel_config(
-    db_session: Session, slack_channel_config_id: int
-) -> SlackChannelConfig | None:
+def fetch_slack_bot_config(
+    db_session: Session, slack_bot_config_id: int
+) -> SlackBotConfig | None:
     return db_session.scalar(
-        select(SlackChannelConfig).where(
-            SlackChannelConfig.id == slack_channel_config_id
-        )
+        select(SlackBotConfig).where(SlackBotConfig.id == slack_bot_config_id)
     )
+
+
+def fetch_slack_bot_configs(db_session: Session) -> Sequence[SlackBotConfig]:
+    return db_session.scalars(select(SlackBotConfig)).all()

backend/danswer/db/users.py

@@ -1,7 +1,6 @@
 from collections.abc import Sequence
 from uuid import UUID
 
-from fastapi import HTTPException
 from fastapi_users.password import PasswordHelper
 from sqlalchemy import func
 from sqlalchemy import select
@@ -11,94 +10,15 @@ from danswer.auth.schemas import UserRole
 from danswer.db.models import User
 
 
-def validate_user_role_update(requested_role: UserRole, current_role: UserRole) -> None:
-    """
-    Validate that a user role update is valid.
-    Assumed only admins can hit this endpoint.
-    raise if:
-    - requested role is a curator
-    - requested role is a slack user
-    - requested role is an external permissioned user
-    - requested role is a limited user
-    - current role is a slack user
-    - current role is an external permissioned user
-    - current role is a limited user
-    """
-
-    if current_role == UserRole.SLACK_USER:
-        raise HTTPException(
-            status_code=400,
-            detail="To change a Slack User's role, they must first login to Danswer via the web app.",
-        )
-
-    if current_role == UserRole.EXT_PERM_USER:
-        # This shouldn't happen, but just in case
-        raise HTTPException(
-            status_code=400,
-            detail="To change an External Permissioned User's role, they must first login to Danswer via the web app.",
-        )
-
-    if current_role == UserRole.LIMITED:
-        raise HTTPException(
-            status_code=400,
-            detail="To change a Limited User's role, they must first login to Danswer via the web app.",
-        )
-
-    if requested_role == UserRole.CURATOR:
-        # This shouldn't happen, but just in case
-        raise HTTPException(
-            status_code=400,
-            detail="Curator role must be set via the User Group Menu",
-        )
-
-    if requested_role == UserRole.LIMITED:
-        # This shouldn't happen, but just in case
-        raise HTTPException(
-            status_code=400,
-            detail=(
-                "A user cannot be set to a Limited User role. "
-                "This role is automatically assigned to users through certain endpoints in the API."
-            ),
-        )
-
-    if requested_role == UserRole.SLACK_USER:
-        # This shouldn't happen, but just in case
-        raise HTTPException(
-            status_code=400,
-            detail=(
-                "A user cannot be set to a Slack User role. "
-                "This role is automatically assigned to users who only use Danswer via Slack."
-            ),
-        )
-
-    if requested_role == UserRole.EXT_PERM_USER:
-        # This shouldn't happen, but just in case
-        raise HTTPException(
-            status_code=400,
-            detail=(
-                "A user cannot be set to an External Permissioned User role. "
-                "This role is automatically assigned to users who have been "
-                "pulled in to the system via an external permissions system."
-            ),
-        )
-
-
 def list_users(
-    db_session: Session, email_filter_string: str = "", include_external: bool = False
+    db_session: Session, email_filter_string: str = "", user: User | None = None
 ) -> Sequence[User]:
     """List all users. No pagination as of now, as the # of users
     is assumed to be relatively small (<< 1 million)"""
     stmt = select(User)
 
-    where_clause = []
-
-    if not include_external:
-        where_clause.append(User.role != UserRole.EXT_PERM_USER)
-
     if email_filter_string:
-        where_clause.append(User.email.ilike(f"%{email_filter_string}%"))  # type: ignore
-
-    stmt = stmt.where(*where_clause)
+        stmt = stmt.where(User.email.ilike(f"%{email_filter_string}%"))  # type: ignore
 
     return db_session.scalars(stmt).unique().all()
 
@@ -125,56 +45,68 @@ def get_user_by_email(email: str, db_session: Session) -> User | None:
 
 
 def fetch_user_by_id(db_session: Session, user_id: UUID) -> User | None:
-    return db_session.query(User).filter(User.id == user_id).first()  # type: ignore
+    user = db_session.query(User).filter(User.id == user_id).first()  # type: ignore
+
+    return user
 
 
-def _generate_non_web_slack_user(email: str) -> User:
+def _generate_non_web_user(email: str) -> User:
     fastapi_users_pw_helper = PasswordHelper()
     password = fastapi_users_pw_helper.generate()
     hashed_pass = fastapi_users_pw_helper.hash(password)
     return User(
         email=email,
         hashed_password=hashed_pass,
-        role=UserRole.SLACK_USER,
+        has_web_login=False,
+        role=UserRole.BASIC,
     )
 
 
-def add_slack_user_if_not_exists(db_session: Session, email: str) -> User:
-    email = email.lower()
+def add_non_web_user_if_not_exists(db_session: Session, email: str) -> User:
     user = get_user_by_email(email, db_session)
     if user is not None:
-        # If the user is an external permissioned user, we update it to a slack user
-        if user.role == UserRole.EXT_PERM_USER:
-            user.role = UserRole.SLACK_USER
-            db_session.commit()
         return user
 
-    user = _generate_non_web_slack_user(email=email)
+    user = _generate_non_web_user(email=email)
     db_session.add(user)
     db_session.commit()
     return user
 
 
-def _generate_non_web_permissioned_user(email: str) -> User:
-    fastapi_users_pw_helper = PasswordHelper()
-    password = fastapi_users_pw_helper.generate()
-    hashed_pass = fastapi_users_pw_helper.hash(password)
-    return User(
-        email=email,
-        hashed_password=hashed_pass,
-        role=UserRole.EXT_PERM_USER,
-    )
+def add_non_web_user_if_not_exists__no_commit(db_session: Session, email: str) -> User:
+    user = get_user_by_email(email, db_session)
+    if user is not None:
+        return user
+
+    user = _generate_non_web_user(email=email)
+    db_session.add(user)
+    db_session.flush()  # generate id
+    return user
 
 
-def batch_add_ext_perm_user_if_not_exists(
+def batch_add_non_web_user_if_not_exists__no_commit(
     db_session: Session, emails: list[str]
 ) -> list[User]:
-    emails = [email.lower() for email in emails]
     found_users, missing_user_emails = get_users_by_emails(db_session, emails)
 
     new_users: list[User] = []
     for email in missing_user_emails:
-        new_users.append(_generate_non_web_permissioned_user(email=email))
+        new_users.append(_generate_non_web_user(email=email))
+
+    db_session.add_all(new_users)
+    db_session.flush()  # generate ids
+
+    return found_users + new_users
+
+
+def batch_add_non_web_user_if_not_exists(
+    db_session: Session, emails: list[str]
+) -> list[User]:
+    found_users, missing_user_emails = get_users_by_emails(db_session, emails)
+
+    new_users: list[User] = []
+    for email in missing_user_emails:
+        new_users.append(_generate_non_web_user(email=email))
 
     db_session.add_all(new_users)
     db_session.commit()

backend/danswer/document_index/vespa/app_config/schemas/danswer_chunk.sd

@@ -15,7 +15,7 @@ schema DANSWER_CHUNK_NAME {
         # Must have an additional field for whether to skip title embeddings
         # This information cannot be extracted from either the title field nor title embedding
         field skip_title type bool {
-            indexing: attribute 
+            indexing: attribute
         }
         # May not always match the `semantic_identifier` e.g. for Slack docs the
         # `semantic_identifier` will be the channel name, but the `title` will be empty
@@ -36,7 +36,7 @@ schema DANSWER_CHUNK_NAME {
         }
         # Title embedding (x1)
         field title_embedding type tensor<float>(x[VARIABLE_DIM]) {
-            indexing: attribute | index
+            indexing: attribute
             attribute {
                 distance-metric: angular
             }
@@ -44,7 +44,7 @@ schema DANSWER_CHUNK_NAME {
         # Content embeddings (chunk + optional mini chunks embeddings)
         # "t" and "x" are arbitrary names, not special keywords
         field embeddings type tensor<float>(t{},x[VARIABLE_DIM]) {
-            indexing: attribute | index
+            indexing: attribute
             attribute {
                 distance-metric: angular
             }

backend/danswer/document_index/vespa/indexing_utils.py

@@ -2,7 +2,6 @@ import concurrent.futures
 import json
 from datetime import datetime
 from datetime import timezone
-from http import HTTPStatus
 
 import httpx
 from retry import retry
@@ -195,14 +194,6 @@ def _index_vespa_chunk(
         logger.exception(
             f"Failed to index document: '{document.id}'. Got response: '{res.text}'"
         )
-        if isinstance(e, httpx.HTTPStatusError):
-            if e.response.status_code == HTTPStatus.INSUFFICIENT_STORAGE:
-                logger.error(
-                    "NOTE: HTTP Status 507 Insufficient Storage usually means "
-                    "you need to allocate more memory or disk space to the "
-                    "Vespa/index container."
-                )
-
         raise e
 
 

backend/danswer/indexing/chunker.py

@@ -10,7 +10,7 @@ from danswer.connectors.cross_connector_utils.miscellaneous_utils import (
     get_metadata_keys_to_ignore,
 )
 from danswer.connectors.models import Document
-from danswer.indexing.indexing_heartbeat import IndexingHeartbeatInterface
+from danswer.indexing.indexing_heartbeat import Heartbeat
 from danswer.indexing.models import DocAwareChunk
 from danswer.natural_language_processing.utils import BaseTokenizer
 from danswer.utils.logger import setup_logger
@@ -125,7 +125,7 @@ class Chunker:
         chunk_token_limit: int = DOC_EMBEDDING_CONTEXT_SIZE,
         chunk_overlap: int = CHUNK_OVERLAP,
         mini_chunk_size: int = MINI_CHUNK_SIZE,
-        callback: IndexingHeartbeatInterface | None = None,
+        heartbeat: Heartbeat | None = None,
     ) -> None:
         from llama_index.text_splitter import SentenceSplitter
 
@@ -134,7 +134,7 @@ class Chunker:
         self.enable_multipass = enable_multipass
         self.enable_large_chunks = enable_large_chunks
         self.tokenizer = tokenizer
-        self.callback = callback
+        self.heartbeat = heartbeat
 
         self.blurb_splitter = SentenceSplitter(
             tokenizer=tokenizer.tokenize,
@@ -356,14 +356,9 @@ class Chunker:
     def chunk(self, documents: list[Document]) -> list[DocAwareChunk]:
         final_chunks: list[DocAwareChunk] = []
         for document in documents:
-            if self.callback:
-                if self.callback.should_stop():
-                    raise RuntimeError("Chunker.chunk: Stop signal detected")
+            final_chunks.extend(self._handle_single_document(document))
 
-            chunks = self._handle_single_document(document)
-            final_chunks.extend(chunks)
-
-            if self.callback:
-                self.callback.progress("Chunker.chunk", len(chunks))
+            if self.heartbeat:
+                self.heartbeat.heartbeat()
 
         return final_chunks

backend/danswer/indexing/embedder.py

@@ -2,7 +2,7 @@ from abc import ABC
 from abc import abstractmethod
 
 from danswer.db.models import SearchSettings
-from danswer.indexing.indexing_heartbeat import IndexingHeartbeatInterface
+from danswer.indexing.indexing_heartbeat import Heartbeat
 from danswer.indexing.models import ChunkEmbedding
 from danswer.indexing.models import DocAwareChunk
 from danswer.indexing.models import IndexChunk
@@ -34,7 +34,7 @@ class IndexingEmbedder(ABC):
         api_url: str | None,
         api_version: str | None,
         deployment_name: str | None,
-        callback: IndexingHeartbeatInterface | None,
+        heartbeat: Heartbeat | None,
     ):
         self.model_name = model_name
         self.normalize = normalize
@@ -60,7 +60,7 @@ class IndexingEmbedder(ABC):
             server_host=INDEXING_MODEL_SERVER_HOST,
             server_port=INDEXING_MODEL_SERVER_PORT,
             retrim_content=True,
-            callback=callback,
+            heartbeat=heartbeat,
         )
 
     @abstractmethod
@@ -83,7 +83,7 @@ class DefaultIndexingEmbedder(IndexingEmbedder):
         api_url: str | None = None,
         api_version: str | None = None,
         deployment_name: str | None = None,
-        callback: IndexingHeartbeatInterface | None = None,
+        heartbeat: Heartbeat | None = None,
     ):
         super().__init__(
             model_name,
@@ -95,7 +95,7 @@ class DefaultIndexingEmbedder(IndexingEmbedder):
             api_url,
             api_version,
             deployment_name,
-            callback,
+            heartbeat,
         )
 
     @log_function_time()
@@ -201,9 +201,7 @@ class DefaultIndexingEmbedder(IndexingEmbedder):
 
     @classmethod
     def from_db_search_settings(
-        cls,
-        search_settings: SearchSettings,
-        callback: IndexingHeartbeatInterface | None = None,
+        cls, search_settings: SearchSettings, heartbeat: Heartbeat | None = None
     ) -> "DefaultIndexingEmbedder":
         return cls(
             model_name=search_settings.model_name,
@@ -215,5 +213,5 @@ class DefaultIndexingEmbedder(IndexingEmbedder):
             api_url=search_settings.api_url,
             api_version=search_settings.api_version,
             deployment_name=search_settings.deployment_name,
-            callback=callback,
+            heartbeat=heartbeat,
         )

backend/danswer/indexing/indexing_heartbeat.py

@@ -1,15 +1,41 @@
-from abc import ABC
-from abc import abstractmethod
+import abc
+from typing import Any
+
+from sqlalchemy import func
+from sqlalchemy.orm import Session
+
+from danswer.db.index_attempt import get_index_attempt
+from danswer.utils.logger import setup_logger
+
+logger = setup_logger()
 
 
-class IndexingHeartbeatInterface(ABC):
-    """Defines a callback interface to be passed to
-    to run_indexing_entrypoint."""
+class Heartbeat(abc.ABC):
+    """Useful for any long-running work that goes through a bunch of items
+    and needs to occasionally give updates on progress.
+    e.g. chunking, embedding, updating vespa, etc."""
 
-    @abstractmethod
-    def should_stop(self) -> bool:
-        """Signal to stop the looping function in flight."""
+    @abc.abstractmethod
+    def heartbeat(self, metadata: Any = None) -> None:
+        raise NotImplementedError
 
-    @abstractmethod
-    def progress(self, tag: str, amount: int) -> None:
-        """Send progress updates to the caller."""
+
+class IndexingHeartbeat(Heartbeat):
+    def __init__(self, index_attempt_id: int, db_session: Session, freq: int):
+        self.cnt = 0
+
+        self.index_attempt_id = index_attempt_id
+        self.db_session = db_session
+        self.freq = freq
+
+    def heartbeat(self, metadata: Any = None) -> None:
+        self.cnt += 1
+        if self.cnt % self.freq == 0:
+            index_attempt = get_index_attempt(
+                db_session=self.db_session, index_attempt_id=self.index_attempt_id
+            )
+            if index_attempt:
+                index_attempt.time_updated = func.now()
+                self.db_session.commit()
+            else:
+                logger.error("Index attempt not found, this should not happen!")

backend/danswer/indexing/indexing_pipeline.py

@@ -1,9 +1,7 @@
 import traceback
 from functools import partial
-from http import HTTPStatus
 from typing import Protocol
 
-import httpx
 from pydantic import BaseModel
 from pydantic import ConfigDict
 from sqlalchemy.orm import Session
@@ -22,8 +20,7 @@ from danswer.db.document import get_documents_by_ids
 from danswer.db.document import prepare_to_modify_documents
 from danswer.db.document import update_docs_last_modified__no_commit
 from danswer.db.document import update_docs_updated_at__no_commit
-from danswer.db.document import upsert_document_by_connector_credential_pair
-from danswer.db.document import upsert_documents
+from danswer.db.document import upsert_documents_complete
 from danswer.db.document_set import fetch_document_sets_for_documents
 from danswer.db.index_attempt import create_index_attempt_error
 from danswer.db.models import Document as DBDocument
@@ -34,7 +31,7 @@ from danswer.document_index.interfaces import DocumentIndex
 from danswer.document_index.interfaces import DocumentMetadata
 from danswer.indexing.chunker import Chunker
 from danswer.indexing.embedder import IndexingEmbedder
-from danswer.indexing.indexing_heartbeat import IndexingHeartbeatInterface
+from danswer.indexing.indexing_heartbeat import IndexingHeartbeat
 from danswer.indexing.models import DocAwareChunk
 from danswer.indexing.models import DocMetadataAwareIndexChunk
 from danswer.utils.logger import setup_logger
@@ -65,7 +62,7 @@ def _upsert_documents_in_db(
     db_session: Session,
 ) -> None:
     # Metadata here refers to basic document info, not metadata about the actual content
-    document_metadata_list: list[DocumentMetadata] = []
+    doc_m_batch: list[DocumentMetadata] = []
     for doc in documents:
         first_link = next(
             (section.link for section in doc.sections if section.link), ""
@@ -80,9 +77,12 @@ def _upsert_documents_in_db(
             secondary_owners=get_experts_stores_representations(doc.secondary_owners),
             from_ingestion_api=doc.from_ingestion_api,
         )
-        document_metadata_list.append(db_doc_metadata)
+        doc_m_batch.append(db_doc_metadata)
 
-    upsert_documents(db_session, document_metadata_list)
+    upsert_documents_complete(
+        db_session=db_session,
+        document_metadata_batch=doc_m_batch,
+    )
 
     # Insert document content metadata
     for doc in documents:
@@ -95,25 +95,21 @@ def _upsert_documents_in_db(
                     document_id=doc.id,
                     db_session=db_session,
                 )
-                continue
-
-            create_or_add_document_tag(
-                tag_key=k,
-                tag_value=v,
-                source=doc.source,
-                document_id=doc.id,
-                db_session=db_session,
-            )
+            else:
+                create_or_add_document_tag(
+                    tag_key=k,
+                    tag_value=v,
+                    source=doc.source,
+                    document_id=doc.id,
+                    db_session=db_session,
+                )
 
 
 def get_doc_ids_to_update(
     documents: list[Document], db_docs: list[DBDocument]
 ) -> list[Document]:
     """Figures out which documents actually need to be updated. If a document is already present
-    and the `updated_at` hasn't changed, we shouldn't need to do anything with it.
-
-    NB: Still need to associate the document in the DB if multiple connectors are
-    indexing the same doc."""
+    and the `updated_at` hasn't changed, we shouldn't need to do anything with it."""
     id_update_time_map = {
         doc.id: doc.doc_updated_at for doc in db_docs if doc.doc_updated_at
     }
@@ -156,14 +152,6 @@ def index_doc_batch_with_handler(
             tenant_id=tenant_id,
         )
     except Exception as e:
-        if isinstance(e, httpx.HTTPStatusError):
-            if e.response.status_code == HTTPStatus.INSUFFICIENT_STORAGE:
-                logger.error(
-                    "NOTE: HTTP Status 507 Insufficient Storage indicates "
-                    "you need to allocate more memory or disk space to the "
-                    "Vespa/index container."
-                )
-
         if INDEXING_EXCEPTION_LIMIT == 0:
             raise
 
@@ -207,9 +195,9 @@ def index_doc_batch_prepare(
     db_session: Session,
     ignore_time_skip: bool = False,
 ) -> DocumentBatchPrepareContext | None:
-    """Sets up the documents in the relational DB (source of truth) for permissions, metadata, etc.
+    """This sets up the documents in the relational DB (source of truth) for permissions, metadata, etc.
     This preceeds indexing it into the actual document index."""
-    documents: list[Document] = []
+    documents = []
     for document in document_batch:
         empty_contents = not any(section.text.strip() for section in document.sections)
         if (
@@ -224,58 +212,43 @@ def index_doc_batch_prepare(
             logger.warning(
                 f"Skipping document with ID {document.id} as it has neither title nor content."
             )
-            continue
-
-        if document.title is not None and not document.title.strip() and empty_contents:
+        elif (
+            document.title is not None and not document.title.strip() and empty_contents
+        ):
             # The title is explicitly empty ("" and not None) and the document is empty
             # so when building the chunk text representation, it will be empty and unuseable
             logger.warning(
                 f"Skipping document with ID {document.id} as the chunks will be empty."
             )
-            continue
+        else:
+            documents.append(document)
 
-        documents.append(document)
-
-    # Create a trimmed list of docs that don't have a newer updated at
-    # Shortcuts the time-consuming flow on connector index retries
-    document_ids: list[str] = [document.id for document in documents]
+    document_ids = [document.id for document in documents]
     db_docs: list[DBDocument] = get_documents_by_ids(
         db_session=db_session,
         document_ids=document_ids,
     )
 
+    # Skip indexing docs that don't have a newer updated at
+    # Shortcuts the time-consuming flow on connector index retries
     updatable_docs = (
         get_doc_ids_to_update(documents=documents, db_docs=db_docs)
         if not ignore_time_skip
         else documents
     )
 
-    # for all updatable docs, upsert into the DB
-    # Does not include doc_updated_at which is also used to indicate a successful update
-    if updatable_docs:
-        _upsert_documents_in_db(
-            documents=updatable_docs,
-            index_attempt_metadata=index_attempt_metadata,
-            db_session=db_session,
-        )
-
-    logger.info(
-        f"Upserted {len(updatable_docs)} changed docs out of "
-        f"{len(documents)} total docs into the DB"
-    )
-
-    # for all docs, upsert the document to cc pair relationship
-    upsert_document_by_connector_credential_pair(
-        db_session,
-        index_attempt_metadata.connector_id,
-        index_attempt_metadata.credential_id,
-        document_ids,
-    )
-
-    # No docs to process because the batch is empty or every doc was already indexed
+    # No docs to update either because the batch is empty or every doc was already indexed
     if not updatable_docs:
         return None
 
+    # Create records in the source of truth about these documents,
+    # does not include doc_updated_at which is also used to indicate a successful update
+    _upsert_documents_in_db(
+        documents=documents,
+        index_attempt_metadata=index_attempt_metadata,
+        db_session=db_session,
+    )
+
     id_to_db_doc_map = {doc.id: doc for doc in db_docs}
     return DocumentBatchPrepareContext(
         updatable_docs=updatable_docs, id_to_db_doc_map=id_to_db_doc_map
@@ -296,10 +269,7 @@ def index_doc_batch(
 ) -> tuple[int, int]:
     """Takes different pieces of the indexing pipeline and applies it to a batch of documents
     Note that the documents should already be batched at this point so that it does not inflate the
-    memory requirements
-
-    Returns a tuple where the first element is the number of new docs and the
-    second element is the number of chunks."""
+    memory requirements"""
 
     no_access = DocumentAccess.build(
         user_emails=[],
@@ -342,9 +312,9 @@ def index_doc_batch(
 
         # we're concerned about race conditions where multiple simultaneous indexings might result
         # in one set of metadata overwriting another one in vespa.
-        # we still write data here for the immediate and most likely correct sync, but
+        # we still write data here for immediate and most likely correct sync, but
         # to resolve this, an update of the last modified field at the end of this loop
-        # always triggers a final metadata sync via the celery queue
+        # always triggers a final metadata sync
         access_aware_chunks = [
             DocMetadataAwareIndexChunk.from_index_chunk(
                 index_chunk=chunk,
@@ -381,8 +351,7 @@ def index_doc_batch(
         ids_to_new_updated_at = {}
         for doc in successful_docs:
             last_modified_ids.append(doc.id)
-            # doc_updated_at is the source's idea (on the other end of the connector)
-            # of when the doc was last modified
+            # doc_updated_at is the connector source's idea of when the doc was last modified
             if doc.doc_updated_at is None:
                 continue
             ids_to_new_updated_at[doc.id] = doc.doc_updated_at
@@ -397,13 +366,10 @@ def index_doc_batch(
 
         db_session.commit()
 
-    result = (
-        len([r for r in insertion_records if r.already_existed is False]),
-        len(access_aware_chunks),
+    return len([r for r in insertion_records if r.already_existed is False]), len(
+        access_aware_chunks
     )
 
-    return result
-
 
 def build_indexing_pipeline(
     *,
@@ -414,7 +380,6 @@ def build_indexing_pipeline(
     ignore_time_skip: bool = False,
     attempt_id: int | None = None,
     tenant_id: str | None = None,
-    callback: IndexingHeartbeatInterface | None = None,
 ) -> IndexingPipelineProtocol:
     """Builds a pipeline which takes in a list (batch) of docs and indexes them."""
     search_settings = get_current_search_settings(db_session)
@@ -441,8 +406,13 @@ def build_indexing_pipeline(
         tokenizer=embedder.embedding_model.tokenizer,
         enable_multipass=multipass,
         enable_large_chunks=enable_large_chunks,
-        # after every doc, update status in case there are a bunch of really long docs
-        callback=callback,
+        # after every doc, update status in case there are a bunch of
+        # really long docs
+        heartbeat=IndexingHeartbeat(
+            index_attempt_id=attempt_id, db_session=db_session, freq=1
+        )
+        if attempt_id
+        else None,
     )
 
     return partial(

backend/danswer/key_value_store/interface.py

@@ -1,6 +1,12 @@
 import abc
+from collections.abc import Mapping
+from collections.abc import Sequence
+from typing import TypeAlias
 
-from danswer.utils.special_types import JSON_ro
+
+JSON_ro: TypeAlias = (
+    Mapping[str, "JSON_ro"] | Sequence["JSON_ro"] | str | int | float | bool | None
+)
 
 
 class KvKeyNotFoundError(Exception):

backend/danswer/key_value_store/store.py

@@ -11,11 +11,11 @@ from sqlalchemy.orm import Session
 from danswer.db.engine import get_sqlalchemy_engine
 from danswer.db.engine import is_valid_schema_name
 from danswer.db.models import KVStore
+from danswer.key_value_store.interface import JSON_ro
 from danswer.key_value_store.interface import KeyValueStore
 from danswer.key_value_store.interface import KvKeyNotFoundError
 from danswer.redis.redis_pool import get_redis_client
 from danswer.utils.logger import setup_logger
-from danswer.utils.special_types import JSON_ro
 from shared_configs.configs import MULTI_TENANT
 from shared_configs.configs import POSTGRES_DEFAULT_SCHEMA
 from shared_configs.contextvars import CURRENT_TENANT_ID_CONTEXTVAR

backend/danswer/llm/answering/answer.py

@@ -263,7 +263,6 @@ class Answer:
             message_history=self.message_history,
             llm_config=self.llm.config,
             single_message_history=self.single_message_history,
-            raw_user_text=self.question,
         )
         prompt_builder.update_system_prompt(
             default_build_system_message(self.prompt_config)

backend/danswer/llm/answering/prompts/build.py

@@ -59,7 +59,6 @@ class AnswerPromptBuilder:
         message_history: list[PreviousMessage],
         llm_config: LLMConfig,
         single_message_history: str | None = None,
-        raw_user_text: str | None = None,
     ) -> None:
         self.max_tokens = compute_max_llm_input_tokens(llm_config)
 
@@ -89,12 +88,6 @@ class AnswerPromptBuilder:
 
         self.new_messages_and_token_cnts: list[tuple[BaseMessage, int]] = []
 
-        self.raw_user_message = (
-            HumanMessage(content=raw_user_text)
-            if raw_user_text is not None
-            else user_message
-        )
-
     def update_system_prompt(self, system_message: SystemMessage | None) -> None:
         if not system_message:
             self.system_message_and_token_cnt = None

backend/danswer/llm/answering/stream_processing/quotes_processing.py

@@ -231,16 +231,16 @@ class QuotesProcessor:
 
         model_previous = self.model_output
         self.model_output += token
+
         if not self.found_answer_start:
             m = answer_pattern.search(self.model_output)
             if m:
                 self.found_answer_start = True
 
                 # Prevent heavy cases of hallucinations
-                if self.is_json_prompt and len(self.model_output) > 400:
-                    self.found_answer_end = True
+                if self.is_json_prompt and len(self.model_output) > 70:
                     logger.warning("LLM did not produce json as prompted")
-                    logger.debug("Model output thus far:", self.model_output)
+                    self.found_answer_end = True
                     return
 
                 remaining = self.model_output[m.end() :]

backend/danswer/llm/chat_llm.py

@@ -1,8 +1,6 @@
 import json
 import os
-import traceback
 from collections.abc import Iterator
-from collections.abc import Sequence
 from typing import Any
 from typing import cast
 
@@ -23,8 +21,8 @@ from langchain_core.messages import SystemMessage
 from langchain_core.messages import SystemMessageChunk
 from langchain_core.messages.tool import ToolCallChunk
 from langchain_core.messages.tool import ToolMessage
-from langchain_core.prompt_values import PromptValue
 
+from danswer.configs.app_configs import LOG_ALL_MODEL_INTERACTIONS
 from danswer.configs.app_configs import LOG_DANSWER_MODEL_INTERACTIONS
 from danswer.configs.model_configs import DISABLE_LITELLM_STREAMING
 from danswer.configs.model_configs import GEN_AI_TEMPERATURE
@@ -32,9 +30,7 @@ from danswer.configs.model_configs import LITELLM_EXTRA_BODY
 from danswer.llm.interfaces import LLM
 from danswer.llm.interfaces import LLMConfig
 from danswer.llm.interfaces import ToolChoiceOptions
-from danswer.server.utils import mask_string
 from danswer.utils.logger import setup_logger
-from danswer.utils.long_term_log import LongTermLogger
 
 
 logger = setup_logger()
@@ -44,7 +40,7 @@ logger = setup_logger()
 litellm.drop_params = True
 litellm.telemetry = False
 
-_LLM_PROMPT_LONG_TERM_LOG_CATEGORY = "llm_prompt"
+litellm.set_verbose = LOG_ALL_MODEL_INTERACTIONS
 
 
 def _base_msg_to_role(msg: BaseMessage) -> str:
@@ -200,23 +196,6 @@ def _convert_delta_to_message_chunk(
     raise ValueError(f"Unknown role: {role}")
 
 
-def _prompt_to_dict(
-    prompt: LanguageModelInput,
-) -> Sequence[str | list[str] | dict[str, Any] | tuple[str, str]]:
-    # NOTE: this must go first, since it is also a Sequence
-    if isinstance(prompt, str):
-        return [_convert_message_to_dict(HumanMessage(content=prompt))]
-
-    if isinstance(prompt, (list, Sequence)):
-        return [
-            _convert_message_to_dict(msg) if isinstance(msg, BaseMessage) else msg
-            for msg in prompt
-        ]
-
-    if isinstance(prompt, PromptValue):
-        return [_convert_message_to_dict(message) for message in prompt.to_messages()]
-
-
 class DefaultMultiLLM(LLM):
     """Uses Litellm library to allow easy configuration to use a multitude of LLMs
     See https://python.langchain.com/docs/integrations/chat/litellm"""
@@ -236,7 +215,6 @@ class DefaultMultiLLM(LLM):
         custom_config: dict[str, str] | None = None,
         extra_headers: dict[str, str] | None = None,
         extra_body: dict | None = LITELLM_EXTRA_BODY,
-        long_term_logger: LongTermLogger | None = None,
     ):
         self._timeout = timeout
         self._model_provider = model_provider
@@ -247,7 +225,6 @@ class DefaultMultiLLM(LLM):
         self._api_base = api_base
         self._api_version = api_version
         self._custom_llm_provider = custom_llm_provider
-        self._long_term_logger = long_term_logger
 
         # This can be used to store the maximum output tokens for this model.
         # self._max_output_tokens = (
@@ -279,52 +256,6 @@ class DefaultMultiLLM(LLM):
     def log_model_configs(self) -> None:
         logger.debug(f"Config: {self.config}")
 
-    def _safe_model_config(self) -> dict:
-        dump = self.config.model_dump()
-        dump["api_key"] = mask_string(dump.get("api_key", ""))
-        return dump
-
-    def _record_call(self, prompt: LanguageModelInput) -> None:
-        if self._long_term_logger:
-            self._long_term_logger.record(
-                {"prompt": _prompt_to_dict(prompt), "model": self._safe_model_config()},
-                category=_LLM_PROMPT_LONG_TERM_LOG_CATEGORY,
-            )
-
-    def _record_result(
-        self, prompt: LanguageModelInput, model_output: BaseMessage
-    ) -> None:
-        if self._long_term_logger:
-            self._long_term_logger.record(
-                {
-                    "prompt": _prompt_to_dict(prompt),
-                    "content": model_output.content,
-                    "tool_calls": (
-                        model_output.tool_calls
-                        if hasattr(model_output, "tool_calls")
-                        else []
-                    ),
-                    "model": self._safe_model_config(),
-                },
-                category=_LLM_PROMPT_LONG_TERM_LOG_CATEGORY,
-            )
-
-    def _record_error(self, prompt: LanguageModelInput, error: Exception) -> None:
-        if self._long_term_logger:
-            self._long_term_logger.record(
-                {
-                    "prompt": _prompt_to_dict(prompt),
-                    "error": str(error),
-                    "traceback": "".join(
-                        traceback.format_exception(
-                            type(error), error, error.__traceback__
-                        )
-                    ),
-                    "model": self._safe_model_config(),
-                },
-                category=_LLM_PROMPT_LONG_TERM_LOG_CATEGORY,
-            )
-
     # def _calculate_max_output_tokens(self, prompt: LanguageModelInput) -> int:
     #     # NOTE: This method can be used for calculating the maximum tokens for the stream,
     #     # but it isn't used in practice due to the computational cost of counting tokens
@@ -357,10 +288,14 @@ class DefaultMultiLLM(LLM):
         stream: bool,
         structured_response_format: dict | None = None,
     ) -> litellm.ModelResponse | litellm.CustomStreamWrapper:
-        # litellm doesn't accept LangChain BaseMessage objects, so we need to convert them
-        # to a dict representation
-        processed_prompt = _prompt_to_dict(prompt)
-        self._record_call(processed_prompt)
+        if isinstance(prompt, list):
+            prompt = [
+                _convert_message_to_dict(msg) if isinstance(msg, BaseMessage) else msg
+                for msg in prompt
+            ]
+
+        elif isinstance(prompt, str):
+            prompt = [_convert_message_to_dict(HumanMessage(content=prompt))]
 
         try:
             return litellm.completion(
@@ -373,7 +308,7 @@ class DefaultMultiLLM(LLM):
                 api_version=self._api_version or None,
                 custom_llm_provider=self._custom_llm_provider or None,
                 # actual input
-                messages=processed_prompt,
+                messages=prompt,
                 tools=tools,
                 tool_choice=tool_choice if tools else None,
                 # streaming choice
@@ -393,7 +328,6 @@ class DefaultMultiLLM(LLM):
                 **self._model_kwargs,
             )
         except Exception as e:
-            self._record_error(processed_prompt, e)
             # for break pointing
             raise e
 
@@ -427,10 +361,7 @@ class DefaultMultiLLM(LLM):
         )
         choice = response.choices[0]
         if hasattr(choice, "message"):
-            output = _convert_litellm_message_to_langchain_message(choice.message)
-            if output:
-                self._record_result(prompt, output)
-            return output
+            return _convert_litellm_message_to_langchain_message(choice.message)
         else:
             raise ValueError("Unexpected response choice type")
 
@@ -479,9 +410,6 @@ class DefaultMultiLLM(LLM):
                 "The AI model failed partway through generation, please try again."
             )
 
-        if output:
-            self._record_result(prompt, output)
-
         if LOG_DANSWER_MODEL_INTERACTIONS and output:
             content = output.content or ""
             if isinstance(output, AIMessage):

backend/danswer/llm/factory.py

@@ -10,7 +10,6 @@ from danswer.llm.exceptions import GenAIDisabledException
 from danswer.llm.interfaces import LLM
 from danswer.llm.override_models import LLMOverride
 from danswer.utils.headers import build_llm_extra_headers
-from danswer.utils.long_term_log import LongTermLogger
 
 
 def get_main_llm_from_tuple(
@@ -23,7 +22,6 @@ def get_llms_for_persona(
     persona: Persona,
     llm_override: LLMOverride | None = None,
     additional_headers: dict[str, str] | None = None,
-    long_term_logger: LongTermLogger | None = None,
 ) -> tuple[LLM, LLM]:
     model_provider_override = llm_override.model_provider if llm_override else None
     model_version_override = llm_override.model_version if llm_override else None
@@ -34,7 +32,6 @@ def get_llms_for_persona(
         return get_default_llms(
             temperature=temperature_override or GEN_AI_TEMPERATURE,
             additional_headers=additional_headers,
-            long_term_logger=long_term_logger,
         )
 
     with get_session_context_manager() as db_session:
@@ -60,7 +57,6 @@ def get_llms_for_persona(
             api_version=llm_provider.api_version,
             custom_config=llm_provider.custom_config,
             additional_headers=additional_headers,
-            long_term_logger=long_term_logger,
         )
 
     return _create_llm(model), _create_llm(fast_model)
@@ -70,7 +66,6 @@ def get_default_llms(
     timeout: int = QA_TIMEOUT,
     temperature: float = GEN_AI_TEMPERATURE,
     additional_headers: dict[str, str] | None = None,
-    long_term_logger: LongTermLogger | None = None,
 ) -> tuple[LLM, LLM]:
     if DISABLE_GENERATIVE_AI:
         raise GenAIDisabledException()
@@ -102,7 +97,6 @@ def get_default_llms(
             timeout=timeout,
             temperature=temperature,
             additional_headers=additional_headers,
-            long_term_logger=long_term_logger,
         )
 
     return _create_llm(model_name), _create_llm(fast_model_name)
@@ -119,7 +113,6 @@ def get_llm(
     temperature: float = GEN_AI_TEMPERATURE,
     timeout: int = QA_TIMEOUT,
     additional_headers: dict[str, str] | None = None,
-    long_term_logger: LongTermLogger | None = None,
 ) -> LLM:
     return DefaultMultiLLM(
         model_provider=provider,
@@ -132,5 +125,4 @@ def get_llm(
         temperature=temperature,
         custom_config=custom_config,
         extra_headers=build_llm_extra_headers(additional_headers),
-        long_term_logger=long_term_logger,
     )

backend/danswer/main.py

@@ -64,9 +64,6 @@ from danswer.server.features.prompt.api import basic_router as prompt_router
 from danswer.server.features.tool.api import admin_router as admin_tool_router
 from danswer.server.features.tool.api import router as tool_router
 from danswer.server.gpts.api import router as gpts_router
-from danswer.server.long_term_logs.long_term_logs_api import (
-    router as long_term_logs_router,
-)
 from danswer.server.manage.administrative import router as admin_router
 from danswer.server.manage.embedding.api import admin_router as embedding_admin_router
 from danswer.server.manage.embedding.api import basic_router as embedding_router
@@ -279,7 +276,6 @@ def get_application() -> FastAPI:
     include_router_with_global_prefix_prepended(
         application, get_full_openai_assistants_api_router()
     )
-    include_router_with_global_prefix_prepended(application, long_term_logs_router)
 
     if AUTH_TYPE == AuthType.DISABLED:
         # Server logs this during auth setup verification step

backend/danswer/natural_language_processing/search_nlp_models.py

@@ -16,7 +16,7 @@ from danswer.configs.model_configs import (
 )
 from danswer.configs.model_configs import DOC_EMBEDDING_CONTEXT_SIZE
 from danswer.db.models import SearchSettings
-from danswer.indexing.indexing_heartbeat import IndexingHeartbeatInterface
+from danswer.indexing.indexing_heartbeat import Heartbeat
 from danswer.natural_language_processing.utils import get_tokenizer
 from danswer.natural_language_processing.utils import tokenizer_trim_content
 from danswer.utils.logger import setup_logger
@@ -99,7 +99,7 @@ class EmbeddingModel:
         api_url: str | None,
         provider_type: EmbeddingProvider | None,
         retrim_content: bool = False,
-        callback: IndexingHeartbeatInterface | None = None,
+        heartbeat: Heartbeat | None = None,
         api_version: str | None = None,
         deployment_name: str | None = None,
     ) -> None:
@@ -116,7 +116,7 @@ class EmbeddingModel:
         self.tokenizer = get_tokenizer(
             model_name=model_name, provider_type=provider_type
         )
-        self.callback = callback
+        self.heartbeat = heartbeat
 
         model_server_url = build_model_server_url(server_host, server_port)
         self.embed_server_endpoint = f"{model_server_url}/encoder/bi-encoder-embed"
@@ -160,10 +160,6 @@ class EmbeddingModel:
 
         embeddings: list[Embedding] = []
         for idx, text_batch in enumerate(text_batches, start=1):
-            if self.callback:
-                if self.callback.should_stop():
-                    raise RuntimeError("_batch_encode_texts detected stop signal")
-
             logger.debug(f"Encoding batch {idx} of {len(text_batches)}")
             embed_request = EmbedRequest(
                 model_name=self.model_name,
@@ -183,8 +179,8 @@ class EmbeddingModel:
             response = self._make_model_server_request(embed_request)
             embeddings.extend(response.embeddings)
 
-            if self.callback:
-                self.callback.progress("_batch_encode_texts", 1)
+            if self.heartbeat:
+                self.heartbeat.heartbeat()
         return embeddings
 
     def encode(

backend/danswer/natural_language_processing/utils.py

@@ -89,70 +89,67 @@ def _check_tokenizer_cache(
     model_provider: EmbeddingProvider | None, model_name: str | None
 ) -> BaseTokenizer:
     global _TOKENIZER_CACHE
+
     id_tuple = (model_provider, model_name)
 
     if id_tuple not in _TOKENIZER_CACHE:
-        tokenizer = None
+        if model_provider in [EmbeddingProvider.OPENAI, EmbeddingProvider.AZURE]:
+            if model_name is None:
+                raise ValueError(
+                    "model_name is required for OPENAI and AZURE embeddings"
+                )
 
-        if model_name:
-            tokenizer = _try_initialize_tokenizer(model_name, model_provider)
+            _TOKENIZER_CACHE[id_tuple] = TiktokenTokenizer(model_name)
+            return _TOKENIZER_CACHE[id_tuple]
 
-        if not tokenizer:
-            logger.info(
+        try:
+            if model_name is None:
+                model_name = DOCUMENT_ENCODER_MODEL
+
+            logger.debug(f"Initializing HuggingFaceTokenizer for: {model_name}")
+            _TOKENIZER_CACHE[id_tuple] = HuggingFaceTokenizer(model_name)
+        except Exception as primary_error:
+            logger.error(
+                f"Error initializing HuggingFaceTokenizer for {model_name}: {primary_error}"
+            )
+            logger.warning(
                 f"Falling back to default embedding model: {DOCUMENT_ENCODER_MODEL}"
             )
-            tokenizer = HuggingFaceTokenizer(DOCUMENT_ENCODER_MODEL)
 
-        _TOKENIZER_CACHE[id_tuple] = tokenizer
+            try:
+                # Cache this tokenizer name to the default so we don't have to try to load it again
+                # and fail again
+                _TOKENIZER_CACHE[id_tuple] = HuggingFaceTokenizer(
+                    DOCUMENT_ENCODER_MODEL
+                )
+            except Exception as fallback_error:
+                logger.error(
+                    f"Error initializing fallback HuggingFaceTokenizer: {fallback_error}"
+                )
+                raise ValueError(
+                    f"Failed to initialize tokenizer for {model_name} and fallback model"
+                ) from fallback_error
 
     return _TOKENIZER_CACHE[id_tuple]
 
 
-def _try_initialize_tokenizer(
-    model_name: str, model_provider: EmbeddingProvider | None
-) -> BaseTokenizer | None:
-    tokenizer: BaseTokenizer | None = None
-
-    if model_provider is not None:
-        # Try using TiktokenTokenizer first if model_provider exists
-        try:
-            tokenizer = TiktokenTokenizer(model_name)
-            logger.info(f"Initialized TiktokenTokenizer for: {model_name}")
-            return tokenizer
-        except Exception as tiktoken_error:
-            logger.debug(
-                f"TiktokenTokenizer not available for model {model_name}: {tiktoken_error}"
-            )
-    else:
-        # If no provider specified, try HuggingFaceTokenizer
-        try:
-            tokenizer = HuggingFaceTokenizer(model_name)
-            logger.info(f"Initialized HuggingFaceTokenizer for: {model_name}")
-            return tokenizer
-        except Exception as hf_error:
-            logger.warning(
-                f"Error initializing HuggingFaceTokenizer for {model_name}: {hf_error}"
-            )
-
-    # If both initializations fail, return None
-    return None
-
-
 _DEFAULT_TOKENIZER: BaseTokenizer = HuggingFaceTokenizer(DOCUMENT_ENCODER_MODEL)
 
 
 def get_tokenizer(
     model_name: str | None, provider_type: EmbeddingProvider | str | None
 ) -> BaseTokenizer:
-    if isinstance(provider_type, str):
-        try:
-            provider_type = EmbeddingProvider(provider_type)
-        except ValueError:
-            logger.debug(
-                f"Invalid provider_type '{provider_type}'. Falling back to default tokenizer."
-            )
-            return _DEFAULT_TOKENIZER
-    return _check_tokenizer_cache(provider_type, model_name)
+    if provider_type is not None:
+        if isinstance(provider_type, str):
+            try:
+                provider_type = EmbeddingProvider(provider_type)
+            except ValueError:
+                logger.debug(
+                    f"Invalid provider_type '{provider_type}'. Falling back to default tokenizer."
+                )
+                return _DEFAULT_TOKENIZER
+        return _check_tokenizer_cache(provider_type, model_name)
+    return _DEFAULT_TOKENIZER
 
 
 def tokenizer_trim_content(

backend/danswer/one_shot_answer/answer_question.py

@@ -64,7 +64,6 @@ from danswer.tools.tool_implementations.search.search_tool import (
 )
 from danswer.tools.tool_runner import ToolCallKickoff
 from danswer.utils.logger import setup_logger
-from danswer.utils.long_term_log import LongTermLogger
 from danswer.utils.timing import log_generator_function_time
 from danswer.utils.variable_functionality import fetch_ee_implementation_or_noop
 
@@ -125,11 +124,6 @@ def stream_answer_objects(
         danswerbot_flow=danswerbot_flow,
     )
 
-    # permanent "log" store, used primarily for debugging
-    long_term_logger = LongTermLogger(
-        metadata={"user_id": str(user_id), "chat_session_id": str(chat_session.id)}
-    )
-
     temporary_persona: Persona | None = None
 
     if query_req.persona_config is not None:
@@ -140,9 +134,7 @@ def stream_answer_objects(
     persona = temporary_persona if temporary_persona else chat_session.persona
 
     try:
-        llm, fast_llm = get_llms_for_persona(
-            persona=persona, long_term_logger=long_term_logger
-        )
+        llm, fast_llm = get_llms_for_persona(persona=persona)
     except ValueError as e:
         logger.error(
             f"Failed to initialize LLMs for persona '{persona.name}': {str(e)}"
@@ -245,9 +237,7 @@ def stream_answer_objects(
         question=query_msg.message,
         answer_style_config=answer_config,
         prompt_config=PromptConfig.from_model(prompt),
-        llm=get_main_llm_from_tuple(
-            get_llms_for_persona(persona=persona, long_term_logger=long_term_logger)
-        ),
+        llm=get_main_llm_from_tuple(get_llms_for_persona(persona=persona)),
         single_message_history=history_str,
         tools=[search_tool] if search_tool else [],
         force_use_tool=(

backend/danswer/redis/redis_connector_credential_pair.py

@@ -1,10 +1,9 @@
 import time
-from typing import cast
 from uuid import uuid4
 
+import redis
 from celery import Celery
 from redis import Redis
-from redis.lock import Lock as RedisLock
 from sqlalchemy.orm import Session
 
 from danswer.configs.constants import CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT
@@ -14,7 +13,6 @@ from danswer.db.connector_credential_pair import get_connector_credential_pair_f
 from danswer.db.document import (
     construct_document_select_for_connector_credential_pair_by_needs_sync,
 )
-from danswer.db.models import Document
 from danswer.redis.redis_object_helper import RedisObjectHelper
 
 
@@ -32,9 +30,6 @@ class RedisConnectorCredentialPair(RedisObjectHelper):
     def __init__(self, tenant_id: str | None, id: int) -> None:
         super().__init__(tenant_id, str(id))
 
-        # documents that should be skipped
-        self.skip_docs: set[str] = set()
-
     @classmethod
     def get_fence_key(cls) -> str:
         return RedisConnectorCredentialPair.FENCE_PREFIX
@@ -50,19 +45,14 @@ class RedisConnectorCredentialPair(RedisObjectHelper):
         # example: connector_taskset
         return f"{self.TASKSET_PREFIX}"
 
-    def set_skip_docs(self, skip_docs: set[str]) -> None:
-        # documents that should be skipped. Note that this classes updates
-        # the list on the fly
-        self.skip_docs = skip_docs
-
     def generate_tasks(
         self,
         celery_app: Celery,
         db_session: Session,
         redis_client: Redis,
-        lock: RedisLock,
+        lock: redis.lock.Lock,
         tenant_id: str | None,
-    ) -> tuple[int, int] | None:
+    ) -> int | None:
         last_lock_time = time.monotonic()
 
         async_results = []
@@ -74,10 +64,7 @@ class RedisConnectorCredentialPair(RedisObjectHelper):
             cc_pair.connector_id, cc_pair.credential_id
         )
 
-        num_docs = 0
-
         for doc in db_session.scalars(stmt).yield_per(1):
-            doc = cast(Document, doc)
             current_time = time.monotonic()
             if current_time - last_lock_time >= (
                 CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT / 4
@@ -85,12 +72,6 @@ class RedisConnectorCredentialPair(RedisObjectHelper):
                 lock.reacquire()
                 last_lock_time = current_time
 
-            num_docs += 1
-
-            # check if we should skip the document (typically because it's already syncing)
-            if doc.id in self.skip_docs:
-                continue
-
             # celery's default task id format is "dd32ded3-00aa-4884-8b21-42f8332e7fac"
             # the key for the result is "celery-task-meta-dd32ded3-00aa-4884-8b21-42f8332e7fac"
             # we prefix the task id so it's easier to keep track of who created the task
@@ -113,6 +94,5 @@ class RedisConnectorCredentialPair(RedisObjectHelper):
             )
 
             async_results.append(result)
-            self.skip_docs.add(doc.id)
 
-        return len(async_results), num_docs
+        return len(async_results)

backend/danswer/redis/redis_connector_delete.py

@@ -6,7 +6,6 @@ from uuid import uuid4
 import redis
 from celery import Celery
 from pydantic import BaseModel
-from redis.lock import Lock as RedisLock
 from sqlalchemy.orm import Session
 
 from danswer.configs.constants import CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT
@@ -14,10 +13,9 @@ from danswer.configs.constants import DanswerCeleryPriority
 from danswer.configs.constants import DanswerCeleryQueues
 from danswer.db.connector_credential_pair import get_connector_credential_pair_from_id
 from danswer.db.document import construct_document_select_for_connector_credential_pair
-from danswer.db.models import Document as DbDocument
 
 
-class RedisConnectorDeletePayload(BaseModel):
+class RedisConnectorDeletionFenceData(BaseModel):
     num_tasks: int | None
     submitted: datetime
 
@@ -54,18 +52,20 @@ class RedisConnectorDelete:
         return False
 
     @property
-    def payload(self) -> RedisConnectorDeletePayload | None:
+    def payload(self) -> RedisConnectorDeletionFenceData | None:
         # read related data and evaluate/print task progress
         fence_bytes = cast(bytes, self.redis.get(self.fence_key))
         if fence_bytes is None:
             return None
 
         fence_str = fence_bytes.decode("utf-8")
-        payload = RedisConnectorDeletePayload.model_validate_json(cast(str, fence_str))
+        payload = RedisConnectorDeletionFenceData.model_validate_json(
+            cast(str, fence_str)
+        )
 
         return payload
 
-    def set_fence(self, payload: RedisConnectorDeletePayload | None) -> None:
+    def set_fence(self, payload: RedisConnectorDeletionFenceData | None) -> None:
         if not payload:
             self.redis.delete(self.fence_key)
             return
@@ -83,7 +83,7 @@ class RedisConnectorDelete:
         self,
         celery_app: Celery,
         db_session: Session,
-        lock: RedisLock,
+        lock: redis.lock.Lock,
     ) -> int | None:
         """Returns None if the cc_pair doesn't exist.
         Otherwise, returns an int with the number of generated tasks."""
@@ -97,8 +97,7 @@ class RedisConnectorDelete:
         stmt = construct_document_select_for_connector_credential_pair(
             cc_pair.connector_id, cc_pair.credential_id
         )
-        for doc_temp in db_session.scalars(stmt).yield_per(1):
-            doc: DbDocument = doc_temp
+        for doc in db_session.scalars(stmt).yield_per(1):
             current_time = time.monotonic()
             if current_time - last_lock_time >= (
                 CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT / 4
@@ -130,10 +129,6 @@ class RedisConnectorDelete:
 
         return len(async_results)
 
-    def reset(self) -> None:
-        self.redis.delete(self.taskset_key)
-        self.redis.delete(self.fence_key)
-
     @staticmethod
     def remove_from_taskset(id: int, task_id: str, r: redis.Redis) -> None:
         taskset_key = f"{RedisConnectorDelete.TASKSET_PREFIX}_{id}"

backend/danswer/redis/redis_connector_doc_perm_sync.py

@@ -6,7 +6,6 @@ from uuid import uuid4
 import redis
 from celery import Celery
 from pydantic import BaseModel
-from redis.lock import Lock as RedisLock
 
 from danswer.access.models import DocExternalAccess
 from danswer.configs.constants import CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT
@@ -128,7 +127,7 @@ class RedisConnectorPermissionSync:
     def generate_tasks(
         self,
         celery_app: Celery,
-        lock: RedisLock | None,
+        lock: redis.lock.Lock | None,
         new_permissions: list[DocExternalAccess],
         source_string: str,
     ) -> int | None:

backend/danswer/redis/redis_connector_ext_group_sync.py

@@ -2,7 +2,6 @@ from typing import cast
 
 import redis
 from celery import Celery
-from redis.lock import Lock as RedisLock
 from sqlalchemy.orm import Session
 
 
@@ -104,7 +103,7 @@ class RedisConnectorExternalGroupSync:
         self,
         celery_app: Celery,
         db_session: Session,
-        lock: RedisLock | None,
+        lock: redis.lock.Lock | None,
     ) -> int | None:
         pass
 

backend/danswer/redis/redis_connector_index.py

@@ -6,7 +6,7 @@ import redis
 from pydantic import BaseModel
 
 
-class RedisConnectorIndexPayload(BaseModel):
+class RedisConnectorIndexingFenceData(BaseModel):
     index_attempt_id: int | None
     started: datetime | None
     submitted: datetime
@@ -71,20 +71,22 @@ class RedisConnectorIndex:
         return False
 
     @property
-    def payload(self) -> RedisConnectorIndexPayload | None:
+    def payload(self) -> RedisConnectorIndexingFenceData | None:
         # read related data and evaluate/print task progress
         fence_bytes = cast(bytes, self.redis.get(self.fence_key))
         if fence_bytes is None:
             return None
 
         fence_str = fence_bytes.decode("utf-8")
-        payload = RedisConnectorIndexPayload.model_validate_json(cast(str, fence_str))
+        payload = RedisConnectorIndexingFenceData.model_validate_json(
+            cast(str, fence_str)
+        )
 
         return payload
 
     def set_fence(
         self,
-        payload: RedisConnectorIndexPayload | None,
+        payload: RedisConnectorIndexingFenceData | None,
     ) -> None:
         if not payload:
             self.redis.delete(self.fence_key)

backend/danswer/redis/redis_connector_prune.py

@@ -4,7 +4,6 @@ from uuid import uuid4
 
 import redis
 from celery import Celery
-from redis.lock import Lock as RedisLock
 from sqlalchemy.orm import Session
 
 from danswer.configs.constants import CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT
@@ -106,7 +105,7 @@ class RedisConnectorPrune:
         documents_to_prune: set[str],
         celery_app: Celery,
         db_session: Session,
-        lock: RedisLock | None,
+        lock: redis.lock.Lock | None,
     ) -> int | None:
         last_lock_time = time.monotonic()
 
@@ -150,12 +149,6 @@ class RedisConnectorPrune:
 
         return len(async_results)
 
-    def reset(self) -> None:
-        self.redis.delete(self.generator_progress_key)
-        self.redis.delete(self.generator_complete_key)
-        self.redis.delete(self.taskset_key)
-        self.redis.delete(self.fence_key)
-
     @staticmethod
     def remove_from_taskset(id: int, task_id: str, r: redis.Redis) -> None:
         taskset_key = f"{RedisConnectorPrune.TASKSET_PREFIX}_{id}"

backend/danswer/redis/redis_document_set.py

@@ -5,7 +5,6 @@ from uuid import uuid4
 import redis
 from celery import Celery
 from redis import Redis
-from redis.lock import Lock as RedisLock
 from sqlalchemy.orm import Session
 
 from danswer.configs.constants import CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT
@@ -51,9 +50,9 @@ class RedisDocumentSet(RedisObjectHelper):
         celery_app: Celery,
         db_session: Session,
         redis_client: Redis,
-        lock: RedisLock,
+        lock: redis.lock.Lock,
         tenant_id: str | None,
-    ) -> tuple[int, int] | None:
+    ) -> int | None:
         last_lock_time = time.monotonic()
 
         async_results = []
@@ -85,7 +84,7 @@ class RedisDocumentSet(RedisObjectHelper):
 
             async_results.append(result)
 
-        return len(async_results), len(async_results)
+        return len(async_results)
 
     def reset(self) -> None:
         self.redis.delete(self.taskset_key)

backend/danswer/redis/redis_object_helper.py

@@ -1,9 +1,9 @@
 from abc import ABC
 from abc import abstractmethod
 
+import redis
 from celery import Celery
 from redis import Redis
-from redis.lock import Lock as RedisLock
 from sqlalchemy.orm import Session
 
 from danswer.redis.redis_pool import get_redis_client
@@ -85,13 +85,7 @@ class RedisObjectHelper(ABC):
         celery_app: Celery,
         db_session: Session,
         redis_client: Redis,
-        lock: RedisLock,
+        lock: redis.lock.Lock,
         tenant_id: str | None,
-    ) -> tuple[int, int] | None:
-        """First element should be the number of actual tasks generated, second should
-        be the number of docs that were candidates to be synced for the cc pair.
-
-        The need for this is when we are syncing stale docs referenced by multiple
-        connectors. In a single pass across multiple cc pairs, we only want a task
-        for be created for a particular document id the first time we see it.
-        The rest can be skipped."""
+    ) -> int | None:
+        pass

backend/danswer/redis/redis_usergroup.py

@@ -5,7 +5,6 @@ from uuid import uuid4
 import redis
 from celery import Celery
 from redis import Redis
-from redis.lock import Lock as RedisLock
 from sqlalchemy.orm import Session
 
 from danswer.configs.constants import CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT
@@ -52,15 +51,15 @@ class RedisUserGroup(RedisObjectHelper):
         celery_app: Celery,
         db_session: Session,
         redis_client: Redis,
-        lock: RedisLock,
+        lock: redis.lock.Lock,
         tenant_id: str | None,
-    ) -> tuple[int, int] | None:
+    ) -> int | None:
         last_lock_time = time.monotonic()
 
         async_results = []
 
         if not global_version.is_ee_version():
-            return 0, 0
+            return 0
 
         try:
             construct_document_select_by_usergroup = fetch_versioned_implementation(
@@ -68,7 +67,7 @@ class RedisUserGroup(RedisObjectHelper):
                 "construct_document_select_by_usergroup",
             )
         except ModuleNotFoundError:
-            return 0, 0
+            return 0
 
         stmt = construct_document_select_by_usergroup(int(self._id))
         for doc in db_session.scalars(stmt).yield_per(1):
@@ -98,7 +97,7 @@ class RedisUserGroup(RedisObjectHelper):
 
             async_results.append(result)
 
-        return len(async_results), len(async_results)
+        return len(async_results)
 
     def reset(self) -> None:
         self.redis.delete(self.taskset_key)

backend/danswer/server/documents/connector.py

@@ -865,31 +865,21 @@ def connector_run_once(
             )
             if attempt_id:
                 logger.info(
-                    f"connector_run_once - try_creating_indexing_task succeeded: "
-                    f"connector={run_info.connector_id} "
-                    f"cc_pair={cc_pair.id} "
-                    f"attempt={attempt_id} "
+                    f"try_creating_indexing_task succeeded: cc_pair={cc_pair.id} attempt_id={attempt_id}"
                 )
                 index_attempt_ids.append(attempt_id)
             else:
-                logger.info(
-                    f"connector_run_once - try_creating_indexing_task failed: "
-                    f"connector={run_info.connector_id} "
-                    f"cc_pair={cc_pair.id}"
-                )
+                logger.info(f"try_creating_indexing_task failed: cc_pair={cc_pair.id}")
 
     if not index_attempt_ids:
-        msg = "No new indexing attempts created, indexing jobs are queued or running."
-        logger.info(msg)
         raise HTTPException(
             status_code=400,
-            detail=msg,
+            detail="No new indexing attempts created, indexing jobs are queued or running.",
         )
 
-    msg = f"Successfully created {len(index_attempt_ids)} index attempts. {index_attempt_ids}"
     return StatusResponse(
         success=True,
-        message=msg,
+        message=f"Successfully created {len(index_attempt_ids)} index attempts",
         data=index_attempt_ids,
     )
 

backend/danswer/server/features/persona/api.py

@@ -18,16 +18,12 @@ from danswer.configs.constants import NotificationType
 from danswer.db.engine import get_session
 from danswer.db.models import User
 from danswer.db.notification import create_notification
-from danswer.db.persona import create_assistant_category
 from danswer.db.persona import create_update_persona
-from danswer.db.persona import delete_persona_category
-from danswer.db.persona import get_assistant_categories
 from danswer.db.persona import get_persona_by_id
 from danswer.db.persona import get_personas
 from danswer.db.persona import mark_persona_as_deleted
 from danswer.db.persona import mark_persona_as_not_deleted
 from danswer.db.persona import update_all_personas_display_priority
-from danswer.db.persona import update_persona_category
 from danswer.db.persona import update_persona_public_status
 from danswer.db.persona import update_persona_shared_users
 from danswer.db.persona import update_persona_visibility
@@ -36,8 +32,6 @@ from danswer.file_store.models import ChatFileType
 from danswer.llm.answering.prompts.utils import build_dummy_prompt
 from danswer.server.features.persona.models import CreatePersonaRequest
 from danswer.server.features.persona.models import ImageGenerationToolStatus
-from danswer.server.features.persona.models import PersonaCategoryCreate
-from danswer.server.features.persona.models import PersonaCategoryResponse
 from danswer.server.features.persona.models import PersonaSharedNotificationData
 from danswer.server.features.persona.models import PersonaSnapshot
 from danswer.server.features.persona.models import PromptTemplateResponse
@@ -45,7 +39,6 @@ from danswer.server.models import DisplayPriorityRequest
 from danswer.tools.utils import is_image_generation_available
 from danswer.utils.logger import setup_logger
 
-
 logger = setup_logger()
 
 
@@ -191,59 +184,6 @@ def update_persona(
     )
 
 
-class PersonaCategoryPatchRequest(BaseModel):
-    category_description: str
-    category_name: str
-
-
-@basic_router.get("/categories")
-def get_categories(
-    db: Session = Depends(get_session),
-    _: User | None = Depends(current_user),
-) -> list[PersonaCategoryResponse]:
-    return [
-        PersonaCategoryResponse.from_model(category)
-        for category in get_assistant_categories(db_session=db)
-    ]
-
-
-@admin_router.post("/categories")
-def create_category(
-    category: PersonaCategoryCreate,
-    db: Session = Depends(get_session),
-    _: User | None = Depends(current_admin_user),
-) -> PersonaCategoryResponse:
-    """Create a new assistant category"""
-    category_model = create_assistant_category(
-        name=category.name, description=category.description, db_session=db
-    )
-    return PersonaCategoryResponse.from_model(category_model)
-
-
-@admin_router.patch("/category/{category_id}")
-def patch_persona_category(
-    category_id: int,
-    persona_category_patch_request: PersonaCategoryPatchRequest,
-    _: User | None = Depends(current_admin_user),
-    db_session: Session = Depends(get_session),
-) -> None:
-    update_persona_category(
-        category_id=category_id,
-        category_description=persona_category_patch_request.category_description,
-        category_name=persona_category_patch_request.category_name,
-        db_session=db_session,
-    )
-
-
-@admin_router.delete("/category/{category_id}")
-def delete_category(
-    category_id: int,
-    _: User | None = Depends(current_admin_user),
-    db_session: Session = Depends(get_session),
-) -> None:
-    delete_persona_category(category_id=category_id, db_session=db_session)
-
-
 class PersonaShareRequest(BaseModel):
     user_ids: list[UUID]
 

backend/danswer/server/features/persona/models.py

@@ -5,7 +5,6 @@ from pydantic import BaseModel
 from pydantic import Field
 
 from danswer.db.models import Persona
-from danswer.db.models import PersonaCategory
 from danswer.db.models import StarterMessage
 from danswer.search.enums import RecencyBiasSetting
 from danswer.server.features.document_set.models import DocumentSet
@@ -42,7 +41,6 @@ class CreatePersonaRequest(BaseModel):
     is_default_persona: bool = False
     display_priority: int | None = None
     search_start_date: datetime | None = None
-    category_id: int | None = None
 
 
 class PersonaSnapshot(BaseModel):
@@ -70,7 +68,6 @@ class PersonaSnapshot(BaseModel):
     uploaded_image_id: str | None = None
     is_default_persona: bool
     search_start_date: datetime | None = None
-    category_id: int | None = None
 
     @classmethod
     def from_model(
@@ -118,7 +115,6 @@ class PersonaSnapshot(BaseModel):
             icon_shape=persona.icon_shape,
             uploaded_image_id=persona.uploaded_image_id,
             search_start_date=persona.search_start_date,
-            category_id=persona.category_id,
         )
 
 
@@ -132,22 +128,3 @@ class PersonaSharedNotificationData(BaseModel):
 
 class ImageGenerationToolStatus(BaseModel):
     is_available: bool
-
-
-class PersonaCategoryCreate(BaseModel):
-    name: str
-    description: str
-
-
-class PersonaCategoryResponse(BaseModel):
-    id: int
-    name: str
-    description: str | None
-
-    @classmethod
-    def from_model(cls, category: PersonaCategory) -> "PersonaCategoryResponse":
-        return PersonaCategoryResponse(
-            id=category.id,
-            name=category.name,
-            description=category.description,
-        )

backend/danswer/server/long_term_logs/long_term_logs_api.py

@@ -1,106 +0,0 @@
-import json
-import shutil
-import tempfile
-import zipfile
-from datetime import datetime
-from pathlib import Path
-
-from fastapi import APIRouter
-from fastapi import Depends
-from fastapi import HTTPException
-from fastapi.responses import FileResponse
-from starlette.background import BackgroundTask
-
-from danswer.auth.users import current_admin_user
-from danswer.db.models import User
-from danswer.utils.long_term_log import LongTermLogger
-
-router = APIRouter(prefix="/admin/long-term-logs")
-
-
-@router.get("/{category}")
-def get_long_term_logs(
-    category: str,
-    start_time: datetime | None = None,
-    end_time: datetime | None = None,
-    _: User | None = Depends(current_admin_user),
-) -> list[dict | list | str]:
-    """Fetch logs for a specific category within an optional time range.
-    Only accessible by admin users."""
-    try:
-        logger = LongTermLogger()
-        return logger.fetch_category(  # type: ignore
-            category=category,
-            start_time=start_time,
-            end_time=end_time,
-        )
-    except Exception as e:
-        raise HTTPException(
-            status_code=500,
-            detail=f"Failed to fetch logs for category '{category}': {str(e)}",
-        )
-
-
-@router.get("/{category}/download")
-def download_long_term_logs_zip(
-    category: str,
-    start_time: datetime | None = None,
-    end_time: datetime | None = None,
-    _: User | None = Depends(current_admin_user),
-) -> FileResponse:
-    """Download logs for a specific category as a ZIP file.
-    Only accessible by admin users."""
-    try:
-        logger = LongTermLogger()
-        logs = logger.fetch_category(
-            category=category,
-            start_time=start_time,
-            end_time=end_time,
-        )
-
-        # Create temporary files without using context manager
-        temp_dir = tempfile.mkdtemp()
-        temp_dir_path = Path(temp_dir)
-
-        # Create JSON file
-        json_path = temp_dir_path / f"{category}-logs.json"
-        with open(json_path, "w") as f:
-            json.dump(logs, f, indent=2, default=str)
-
-        # Create ZIP file
-        zip_path = temp_dir_path / f"{category}-logs.zip"
-        with zipfile.ZipFile(zip_path, "w") as zip_file:
-            zip_file.write(json_path, json_path.name)
-
-        # Let FastAPI handle cleanup by setting background tasks
-        return FileResponse(
-            path=zip_path,
-            filename=f"{category}-logs.zip",
-            media_type="application/zip",
-            background=BackgroundTask(
-                lambda: shutil.rmtree(temp_dir, ignore_errors=True)
-            ),
-        )
-
-    except Exception as e:
-        raise HTTPException(
-            status_code=500,
-            detail=f"Failed to create ZIP file for category '{category}': {str(e)}",
-        )
-
-
-@router.get("")
-def get_available_categories(
-    _: User | None = Depends(current_admin_user),
-) -> list[str]:
-    """Get a list of all available log categories.
-    Only accessible by admin users."""
-    try:
-        logger = LongTermLogger()
-        # Get all subdirectories in the log directory
-        categories = [d.name for d in logger.log_file_path.iterdir() if d.is_dir()]
-        return categories
-    except Exception as e:
-        raise HTTPException(
-            status_code=500, detail=f"Failed to fetch log categories: {str(e)}"
-        )

backend/danswer/server/manage/llm/api.py

@@ -30,6 +30,7 @@ from danswer.utils.threadpool_concurrency import run_functions_tuples_in_paralle
 
 logger = setup_logger()
 
+
 admin_router = APIRouter(prefix="/admin/llm")
 basic_router = APIRouter(prefix="/llm")
 

backend/danswer/server/manage/models.py

@@ -13,9 +13,8 @@ from danswer.configs.constants import AuthType
 from danswer.danswerbot.slack.config import VALID_SLACK_FILTERS
 from danswer.db.models import AllowedAnswerFilters
 from danswer.db.models import ChannelConfig
-from danswer.db.models import SlackBot as SlackAppModel
+from danswer.db.models import SlackBotConfig as SlackBotConfigModel
 from danswer.db.models import SlackBotResponseType
-from danswer.db.models import SlackChannelConfig as SlackChannelConfigModel
 from danswer.db.models import User
 from danswer.search.models import SavedSearchSettings
 from danswer.server.features.persona.models import PersonaSnapshot
@@ -128,32 +127,22 @@ class HiddenUpdateRequest(BaseModel):
     hidden: bool
 
 
-class SlackBotCreationRequest(BaseModel):
-    name: str
-    enabled: bool
-
-    bot_token: str
-    app_token: str
-
-
 class SlackBotTokens(BaseModel):
     bot_token: str
     app_token: str
     model_config = ConfigDict(frozen=True)
 
 
-class SlackChannelConfigCreationRequest(BaseModel):
-    slack_bot_id: int
-    # currently, a persona is created for each Slack channel config
+class SlackBotConfigCreationRequest(BaseModel):
+    # currently, a persona is created for each slack bot config
     # in the future, `document_sets` will probably be replaced
     # by an optional `PersonaSnapshot` object. Keeping it like this
     # for now for simplicity / speed of development
     document_sets: list[int] | None = None
-
-    # NOTE: only one of `document_sets` / `persona_id` should be set
-    persona_id: int | None = None
-
-    channel_name: str
+    persona_id: (
+        int | None
+    ) = None  # NOTE: only one of `document_sets` / `persona_id` should be set
+    channel_names: list[str]
     respond_tag_only: bool = False
     respond_to_bots: bool = False
     enable_auto_filters: bool = False
@@ -176,17 +165,14 @@ class SlackChannelConfigCreationRequest(BaseModel):
         return value
 
     @model_validator(mode="after")
-    def validate_document_sets_and_persona_id(
-        self,
-    ) -> "SlackChannelConfigCreationRequest":
+    def validate_document_sets_and_persona_id(self) -> "SlackBotConfigCreationRequest":
         if self.document_sets and self.persona_id:
             raise ValueError("Only one of `document_sets` / `persona_id` should be set")
 
         return self
 
 
-class SlackChannelConfig(BaseModel):
-    slack_bot_id: int
+class SlackBotConfig(BaseModel):
     id: int
     persona: PersonaSnapshot | None
     channel_config: ChannelConfig
@@ -197,53 +183,25 @@ class SlackChannelConfig(BaseModel):
 
     @classmethod
     def from_model(
-        cls, slack_channel_config_model: SlackChannelConfigModel
-    ) -> "SlackChannelConfig":
+        cls, slack_bot_config_model: SlackBotConfigModel
+    ) -> "SlackBotConfig":
         return cls(
-            id=slack_channel_config_model.id,
-            slack_bot_id=slack_channel_config_model.slack_bot_id,
+            id=slack_bot_config_model.id,
             persona=(
                 PersonaSnapshot.from_model(
-                    slack_channel_config_model.persona, allow_deleted=True
+                    slack_bot_config_model.persona, allow_deleted=True
                 )
-                if slack_channel_config_model.persona
+                if slack_bot_config_model.persona
                 else None
             ),
-            channel_config=slack_channel_config_model.channel_config,
-            response_type=slack_channel_config_model.response_type,
+            channel_config=slack_bot_config_model.channel_config,
+            response_type=slack_bot_config_model.response_type,
             # XXX this is going away soon
             standard_answer_categories=[
                 StandardAnswerCategory.from_model(standard_answer_category_model)
-                for standard_answer_category_model in slack_channel_config_model.standard_answer_categories
+                for standard_answer_category_model in slack_bot_config_model.standard_answer_categories
             ],
-            enable_auto_filters=slack_channel_config_model.enable_auto_filters,
-        )
-
-
-class SlackBot(BaseModel):
-    """
-    This model is identical to the SlackAppModel, but it contains
-    a `configs_count` field to make it easier to fetch the number
-    of SlackChannelConfigs associated with a SlackBot.
-    """
-
-    id: int
-    name: str
-    enabled: bool
-    configs_count: int
-
-    bot_token: str
-    app_token: str
-
-    @classmethod
-    def from_model(cls, slack_bot_model: SlackAppModel) -> "SlackBot":
-        return cls(
-            id=slack_bot_model.id,
-            name=slack_bot_model.name,
-            enabled=slack_bot_model.enabled,
-            bot_token=slack_bot_model.bot_token,
-            app_token=slack_bot_model.app_token,
-            configs_count=len(slack_bot_model.slack_channel_configs),
+            enable_auto_filters=slack_bot_config_model.enable_auto_filters,
         )
 
 

backend/danswer/server/manage/slack_bot.py

@@ -4,57 +4,53 @@ from fastapi import HTTPException
 from sqlalchemy.orm import Session
 
 from danswer.auth.users import current_admin_user
-from danswer.danswerbot.slack.config import validate_channel_name
+from danswer.danswerbot.slack.config import validate_channel_names
+from danswer.danswerbot.slack.tokens import fetch_tokens
+from danswer.danswerbot.slack.tokens import save_tokens
 from danswer.db.constants import SLACK_BOT_PERSONA_PREFIX
 from danswer.db.engine import get_session
 from danswer.db.models import ChannelConfig
 from danswer.db.models import User
 from danswer.db.persona import get_persona_by_id
-from danswer.db.slack_bot import fetch_slack_bot
-from danswer.db.slack_bot import fetch_slack_bots
-from danswer.db.slack_bot import insert_slack_bot
-from danswer.db.slack_bot import remove_slack_bot
-from danswer.db.slack_bot import update_slack_bot
-from danswer.db.slack_channel_config import create_slack_channel_persona
-from danswer.db.slack_channel_config import fetch_slack_channel_config
-from danswer.db.slack_channel_config import fetch_slack_channel_configs
-from danswer.db.slack_channel_config import insert_slack_channel_config
-from danswer.db.slack_channel_config import remove_slack_channel_config
-from danswer.db.slack_channel_config import update_slack_channel_config
-from danswer.server.manage.models import SlackBot
-from danswer.server.manage.models import SlackBotCreationRequest
-from danswer.server.manage.models import SlackChannelConfig
-from danswer.server.manage.models import SlackChannelConfigCreationRequest
+from danswer.db.slack_bot_config import create_slack_bot_persona
+from danswer.db.slack_bot_config import fetch_slack_bot_config
+from danswer.db.slack_bot_config import fetch_slack_bot_configs
+from danswer.db.slack_bot_config import insert_slack_bot_config
+from danswer.db.slack_bot_config import remove_slack_bot_config
+from danswer.db.slack_bot_config import update_slack_bot_config
+from danswer.key_value_store.interface import KvKeyNotFoundError
+from danswer.server.manage.models import SlackBotConfig
+from danswer.server.manage.models import SlackBotConfigCreationRequest
+from danswer.server.manage.models import SlackBotTokens
 
 
 router = APIRouter(prefix="/manage")
 
 
 def _form_channel_config(
+    slack_bot_config_creation_request: SlackBotConfigCreationRequest,
+    current_slack_bot_config_id: int | None,
     db_session: Session,
-    slack_channel_config_creation_request: SlackChannelConfigCreationRequest,
-    current_slack_channel_config_id: int | None,
 ) -> ChannelConfig:
-    raw_channel_name = slack_channel_config_creation_request.channel_name
-    respond_tag_only = slack_channel_config_creation_request.respond_tag_only
+    raw_channel_names = slack_bot_config_creation_request.channel_names
+    respond_tag_only = slack_bot_config_creation_request.respond_tag_only
     respond_member_group_list = (
-        slack_channel_config_creation_request.respond_member_group_list
+        slack_bot_config_creation_request.respond_member_group_list
     )
-    answer_filters = slack_channel_config_creation_request.answer_filters
-    follow_up_tags = slack_channel_config_creation_request.follow_up_tags
+    answer_filters = slack_bot_config_creation_request.answer_filters
+    follow_up_tags = slack_bot_config_creation_request.follow_up_tags
 
-    if not raw_channel_name:
+    if not raw_channel_names:
         raise HTTPException(
             status_code=400,
             detail="Must provide at least one channel name",
         )
 
     try:
-        cleaned_channel_name = validate_channel_name(
+        cleaned_channel_names = validate_channel_names(
+            channel_names=raw_channel_names,
+            current_slack_bot_config_id=current_slack_bot_config_id,
             db_session=db_session,
-            channel_name=raw_channel_name,
-            current_slack_channel_config_id=current_slack_channel_config_id,
-            current_slack_bot_id=slack_channel_config_creation_request.slack_bot_id,
         )
     except ValueError as e:
         raise HTTPException(
@@ -69,7 +65,7 @@ def _form_channel_config(
         )
 
     channel_config: ChannelConfig = {
-        "channel_name": cleaned_channel_name,
+        "channel_names": cleaned_channel_names,
     }
     if respond_tag_only is not None:
         channel_config["respond_tag_only"] = respond_tag_only
@@ -82,73 +78,69 @@ def _form_channel_config(
 
     channel_config[
         "respond_to_bots"
-    ] = slack_channel_config_creation_request.respond_to_bots
+    ] = slack_bot_config_creation_request.respond_to_bots
 
     return channel_config
 
 
-@router.post("/admin/slack-app/channel")
-def create_slack_channel_config(
-    slack_channel_config_creation_request: SlackChannelConfigCreationRequest,
+@router.post("/admin/slack-bot/config")
+def create_slack_bot_config(
+    slack_bot_config_creation_request: SlackBotConfigCreationRequest,
     db_session: Session = Depends(get_session),
     _: User | None = Depends(current_admin_user),
-) -> SlackChannelConfig:
+) -> SlackBotConfig:
     channel_config = _form_channel_config(
-        db_session=db_session,
-        slack_channel_config_creation_request=slack_channel_config_creation_request,
-        current_slack_channel_config_id=None,
+        slack_bot_config_creation_request, None, db_session
     )
 
     persona_id = None
-    if slack_channel_config_creation_request.persona_id is not None:
-        persona_id = slack_channel_config_creation_request.persona_id
-    elif slack_channel_config_creation_request.document_sets:
-        persona_id = create_slack_channel_persona(
+    if slack_bot_config_creation_request.persona_id is not None:
+        persona_id = slack_bot_config_creation_request.persona_id
+    elif slack_bot_config_creation_request.document_sets:
+        persona_id = create_slack_bot_persona(
             db_session=db_session,
-            channel_name=channel_config["channel_name"],
-            document_set_ids=slack_channel_config_creation_request.document_sets,
+            channel_names=channel_config["channel_names"],
+            document_set_ids=slack_bot_config_creation_request.document_sets,
             existing_persona_id=None,
         ).id
 
-    slack_channel_config_model = insert_slack_channel_config(
-        slack_bot_id=slack_channel_config_creation_request.slack_bot_id,
+    slack_bot_config_model = insert_slack_bot_config(
         persona_id=persona_id,
         channel_config=channel_config,
-        response_type=slack_channel_config_creation_request.response_type,
-        standard_answer_category_ids=slack_channel_config_creation_request.standard_answer_categories,
+        response_type=slack_bot_config_creation_request.response_type,
+        # XXX this is going away soon
+        standard_answer_category_ids=slack_bot_config_creation_request.standard_answer_categories,
         db_session=db_session,
-        enable_auto_filters=slack_channel_config_creation_request.enable_auto_filters,
+        enable_auto_filters=slack_bot_config_creation_request.enable_auto_filters,
     )
-    return SlackChannelConfig.from_model(slack_channel_config_model)
+    return SlackBotConfig.from_model(slack_bot_config_model)
 
 
-@router.patch("/admin/slack-app/channel/{slack_channel_config_id}")
-def patch_slack_channel_config(
-    slack_channel_config_id: int,
-    slack_channel_config_creation_request: SlackChannelConfigCreationRequest,
+@router.patch("/admin/slack-bot/config/{slack_bot_config_id}")
+def patch_slack_bot_config(
+    slack_bot_config_id: int,
+    slack_bot_config_creation_request: SlackBotConfigCreationRequest,
     db_session: Session = Depends(get_session),
     _: User | None = Depends(current_admin_user),
-) -> SlackChannelConfig:
+) -> SlackBotConfig:
     channel_config = _form_channel_config(
-        db_session=db_session,
-        slack_channel_config_creation_request=slack_channel_config_creation_request,
-        current_slack_channel_config_id=slack_channel_config_id,
+        slack_bot_config_creation_request, slack_bot_config_id, db_session
     )
 
     persona_id = None
-    if slack_channel_config_creation_request.persona_id is not None:
-        persona_id = slack_channel_config_creation_request.persona_id
-    elif slack_channel_config_creation_request.document_sets:
-        existing_slack_channel_config = fetch_slack_channel_config(
-            db_session=db_session, slack_channel_config_id=slack_channel_config_id
+    if slack_bot_config_creation_request.persona_id is not None:
+        persona_id = slack_bot_config_creation_request.persona_id
+    elif slack_bot_config_creation_request.document_sets:
+        existing_slack_bot_config = fetch_slack_bot_config(
+            db_session=db_session, slack_bot_config_id=slack_bot_config_id
         )
-        if existing_slack_channel_config is None:
+        if existing_slack_bot_config is None:
             raise HTTPException(
                 status_code=404,
-                detail="Slack channel config not found",
+                detail="Slack bot config not found",
             )
 
-        existing_persona_id = existing_slack_channel_config.persona_id
+        existing_persona_id = existing_slack_bot_config.persona_id
         if existing_persona_id is not None:
             persona = get_persona_by_id(
                 persona_id=existing_persona_id,
@@ -163,133 +155,62 @@ def patch_slack_channel_config(
                 # for this DanswerBot config
                 existing_persona_id = None
             else:
-                existing_persona_id = existing_slack_channel_config.persona_id
+                existing_persona_id = existing_slack_bot_config.persona_id
 
-        persona_id = create_slack_channel_persona(
+        persona_id = create_slack_bot_persona(
             db_session=db_session,
-            channel_name=channel_config["channel_name"],
-            document_set_ids=slack_channel_config_creation_request.document_sets,
+            channel_names=channel_config["channel_names"],
+            document_set_ids=slack_bot_config_creation_request.document_sets,
             existing_persona_id=existing_persona_id,
-            enable_auto_filters=slack_channel_config_creation_request.enable_auto_filters,
+            enable_auto_filters=slack_bot_config_creation_request.enable_auto_filters,
         ).id
 
-    slack_channel_config_model = update_slack_channel_config(
-        db_session=db_session,
-        slack_channel_config_id=slack_channel_config_id,
+    slack_bot_config_model = update_slack_bot_config(
+        slack_bot_config_id=slack_bot_config_id,
         persona_id=persona_id,
         channel_config=channel_config,
-        response_type=slack_channel_config_creation_request.response_type,
-        standard_answer_category_ids=slack_channel_config_creation_request.standard_answer_categories,
-        enable_auto_filters=slack_channel_config_creation_request.enable_auto_filters,
+        response_type=slack_bot_config_creation_request.response_type,
+        standard_answer_category_ids=slack_bot_config_creation_request.standard_answer_categories,
+        db_session=db_session,
+        enable_auto_filters=slack_bot_config_creation_request.enable_auto_filters,
     )
-    return SlackChannelConfig.from_model(slack_channel_config_model)
+    return SlackBotConfig.from_model(slack_bot_config_model)
 
 
-@router.delete("/admin/slack-app/channel/{slack_channel_config_id}")
-def delete_slack_channel_config(
-    slack_channel_config_id: int,
+@router.delete("/admin/slack-bot/config/{slack_bot_config_id}")
+def delete_slack_bot_config(
+    slack_bot_config_id: int,
     db_session: Session = Depends(get_session),
     user: User | None = Depends(current_admin_user),
 ) -> None:
-    remove_slack_channel_config(
-        db_session=db_session,
-        slack_channel_config_id=slack_channel_config_id,
-        user=user,
+    remove_slack_bot_config(
+        slack_bot_config_id=slack_bot_config_id, user=user, db_session=db_session
     )
 
 
-@router.get("/admin/slack-app/channel")
-def list_slack_channel_configs(
+@router.get("/admin/slack-bot/config")
+def list_slack_bot_configs(
     db_session: Session = Depends(get_session),
     _: User | None = Depends(current_admin_user),
-) -> list[SlackChannelConfig]:
-    slack_channel_config_models = fetch_slack_channel_configs(db_session=db_session)
+) -> list[SlackBotConfig]:
+    slack_bot_config_models = fetch_slack_bot_configs(db_session=db_session)
     return [
-        SlackChannelConfig.from_model(slack_channel_config_model)
-        for slack_channel_config_model in slack_channel_config_models
-    ]
-
-
-@router.post("/admin/slack-app/bots")
-def create_bot(
-    slack_bot_creation_request: SlackBotCreationRequest,
-    db_session: Session = Depends(get_session),
-    _: User | None = Depends(current_admin_user),
-) -> SlackBot:
-    slack_bot_model = insert_slack_bot(
-        db_session=db_session,
-        name=slack_bot_creation_request.name,
-        enabled=slack_bot_creation_request.enabled,
-        bot_token=slack_bot_creation_request.bot_token,
-        app_token=slack_bot_creation_request.app_token,
-    )
-    return SlackBot.from_model(slack_bot_model)
-
-
-@router.patch("/admin/slack-app/bots/{slack_bot_id}")
-def patch_bot(
-    slack_bot_id: int,
-    slack_bot_creation_request: SlackBotCreationRequest,
-    db_session: Session = Depends(get_session),
-    _: User | None = Depends(current_admin_user),
-) -> SlackBot:
-    slack_bot_model = update_slack_bot(
-        db_session=db_session,
-        slack_bot_id=slack_bot_id,
-        name=slack_bot_creation_request.name,
-        enabled=slack_bot_creation_request.enabled,
-        bot_token=slack_bot_creation_request.bot_token,
-        app_token=slack_bot_creation_request.app_token,
-    )
-    return SlackBot.from_model(slack_bot_model)
-
-
-@router.delete("/admin/slack-app/bots/{slack_bot_id}")
-def delete_bot(
-    slack_bot_id: int,
-    db_session: Session = Depends(get_session),
-    _: User | None = Depends(current_admin_user),
-) -> None:
-    remove_slack_bot(
-        db_session=db_session,
-        slack_bot_id=slack_bot_id,
-    )
-
-
-@router.get("/admin/slack-app/bots/{slack_bot_id}")
-def get_bot_by_id(
-    slack_bot_id: int,
-    db_session: Session = Depends(get_session),
-    _: User | None = Depends(current_admin_user),
-) -> SlackBot:
-    slack_bot_model = fetch_slack_bot(
-        db_session=db_session,
-        slack_bot_id=slack_bot_id,
-    )
-    return SlackBot.from_model(slack_bot_model)
-
-
-@router.get("/admin/slack-app/bots")
-def list_bots(
-    db_session: Session = Depends(get_session),
-    _: User | None = Depends(current_admin_user),
-) -> list[SlackBot]:
-    slack_bot_models = fetch_slack_bots(db_session=db_session)
-    return [
-        SlackBot.from_model(slack_bot_model) for slack_bot_model in slack_bot_models
-    ]
-
-
-@router.get("/admin/slack-app/bots/{bot_id}/config")
-def list_bot_configs(
-    bot_id: int,
-    db_session: Session = Depends(get_session),
-    _: User | None = Depends(current_admin_user),
-) -> list[SlackChannelConfig]:
-    slack_bot_config_models = fetch_slack_channel_configs(
-        db_session=db_session, slack_bot_id=bot_id
-    )
-    return [
-        SlackChannelConfig.from_model(slack_bot_config_model)
+        SlackBotConfig.from_model(slack_bot_config_model)
         for slack_bot_config_model in slack_bot_config_models
     ]
+
+
+@router.put("/admin/slack-bot/tokens")
+def put_tokens(
+    tokens: SlackBotTokens,
+    _: User | None = Depends(current_admin_user),
+) -> None:
+    save_tokens(tokens=tokens)
+
+
+@router.get("/admin/slack-bot/tokens")
+def get_tokens(_: User | None = Depends(current_admin_user)) -> SlackBotTokens:
+    try:
+        return fetch_tokens()
+    except KvKeyNotFoundError:
+        raise HTTPException(status_code=404, detail="No tokens found")

backend/danswer/server/manage/users.py

@@ -49,7 +49,6 @@ from danswer.db.models import User
 from danswer.db.models import User__UserGroup
 from danswer.db.users import get_user_by_email
 from danswer.db.users import list_users
-from danswer.db.users import validate_user_role_update
 from danswer.key_value_store.factory import get_kv_store
 from danswer.server.manage.models import AllUsersResponse
 from danswer.server.manage.models import UserByEmail
@@ -85,31 +84,28 @@ def set_user_role(
     if not user_to_update:
         raise HTTPException(status_code=404, detail="User not found")
 
-    current_role = user_to_update.role
-    requested_role = user_role_update_request.new_role
-    if requested_role == current_role:
+    if user_role_update_request.new_role == UserRole.CURATOR:
+        raise HTTPException(
+            status_code=400,
+            detail="Curator role must be set via the User Group Menu",
+        )
+
+    if user_to_update.role == user_role_update_request.new_role:
         return
 
-    # This will raise an exception if the role update is invalid
-    validate_user_role_update(
-        requested_role=requested_role,
-        current_role=current_role,
-    )
-
-    if user_to_update.id == current_user.id:
+    if current_user.id == user_to_update.id:
         raise HTTPException(
             status_code=400,
             detail="An admin cannot demote themselves from admin role!",
         )
 
-    if requested_role == UserRole.CURATOR:
-        # Remove all curator db relationships before changing role
+    if user_to_update.role == UserRole.CURATOR:
         fetch_ee_implementation_or_noop(
             "danswer.db.user_group",
             "remove_curator_status__no_commit",
         )(db_session, user_to_update)
 
-    user_to_update.role = user_role_update_request.new_role
+    user_to_update.role = user_role_update_request.new_role.value
 
     db_session.commit()
 
@@ -127,7 +123,7 @@ def list_all_users(
 
     users = [
         user
-        for user in list_users(db_session, email_filter_string=q)
+        for user in list_users(db_session, email_filter_string=q, user=user)
         if not is_api_key_email_address(user.email)
     ]
     accepted_emails = {user.email for user in users}

backend/danswer/tools/tool.py

@@ -3,9 +3,9 @@ from collections.abc import Generator
 from typing import Any
 from typing import TYPE_CHECKING
 
+from danswer.key_value_store.interface import JSON_ro
 from danswer.llm.answering.models import PreviousMessage
 from danswer.llm.interfaces import LLM
-from danswer.utils.special_types import JSON_ro
 
 
 if TYPE_CHECKING:

backend/danswer/tools/tool_implementations/custom/custom_tool.py

@@ -13,13 +13,13 @@ import requests
 from langchain_core.messages import HumanMessage
 from langchain_core.messages import SystemMessage
 from pydantic import BaseModel
-from requests import JSONDecodeError
 
 from danswer.configs.constants import FileOrigin
 from danswer.db.engine import get_session_with_default_tenant
 from danswer.file_store.file_store import get_default_file_store
 from danswer.file_store.models import ChatFileType
 from danswer.file_store.models import InMemoryChatFile
+from danswer.key_value_store.interface import JSON_ro
 from danswer.llm.answering.models import PreviousMessage
 from danswer.llm.answering.prompts.build import AnswerPromptBuilder
 from danswer.llm.interfaces import LLM
@@ -57,7 +57,6 @@ from danswer.tools.tool_implementations.custom.prompt import (
 from danswer.utils.headers import header_list_to_header_dict
 from danswer.utils.headers import HeaderItemDict
 from danswer.utils.logger import setup_logger
-from danswer.utils.special_types import JSON_ro
 
 logger = setup_logger()
 
@@ -242,8 +241,6 @@ class CustomTool(BaseTool):
         )
         content_type = response.headers.get("Content-Type", "")
 
-        tool_result: Any
-        response_type: str
         if "text/csv" in content_type:
             file_ids = self._save_and_get_file_references(
                 response.content, content_type
@@ -259,15 +256,8 @@ class CustomTool(BaseTool):
             response_type = "image"
 
         else:
-            try:
-                tool_result = response.json()
-                response_type = "json"
-            except JSONDecodeError:
-                logger.exception(
-                    f"Failed to parse response as JSON for tool '{self._name}'"
-                )
-                tool_result = response.text
-                response_type = "text"
+            tool_result = response.json()
+            response_type = "json"
 
         logger.info(
             f"Returning tool response for {self._name} with type {response_type}"

backend/danswer/tools/tool_implementations/images/image_generation_tool.py

@@ -9,6 +9,7 @@ from pydantic import BaseModel
 
 from danswer.chat.chat_utils import combine_message_chain
 from danswer.configs.model_configs import GEN_AI_HISTORY_CUTOFF
+from danswer.key_value_store.interface import JSON_ro
 from danswer.llm.answering.models import PreviousMessage
 from danswer.llm.answering.prompts.build import AnswerPromptBuilder
 from danswer.llm.interfaces import LLM
@@ -23,7 +24,6 @@ from danswer.tools.tool_implementations.images.prompt import (
 )
 from danswer.utils.headers import build_llm_extra_headers
 from danswer.utils.logger import setup_logger
-from danswer.utils.special_types import JSON_ro
 from danswer.utils.threadpool_concurrency import run_functions_tuples_in_parallel
 
 

backend/danswer/tools/tool_implementations/internet_search/internet_search_tool.py

@@ -10,6 +10,7 @@ from danswer.chat.chat_utils import combine_message_chain
 from danswer.chat.models import LlmDoc
 from danswer.configs.constants import DocumentSource
 from danswer.configs.model_configs import GEN_AI_HISTORY_CUTOFF
+from danswer.key_value_store.interface import JSON_ro
 from danswer.llm.answering.models import AnswerStyleConfig
 from danswer.llm.answering.models import PreviousMessage
 from danswer.llm.answering.models import PromptConfig
@@ -36,7 +37,6 @@ from danswer.tools.tool_implementations.search_like_tool_utils import (
     FINAL_CONTEXT_DOCUMENTS_ID,
 )
 from danswer.utils.logger import setup_logger
-from danswer.utils.special_types import JSON_ro
 
 logger = setup_logger()
 

backend/danswer/tools/tool_implementations/search/search_tool.py

@@ -16,6 +16,7 @@ from danswer.configs.chat_configs import CONTEXT_CHUNKS_BELOW
 from danswer.configs.model_configs import GEN_AI_MODEL_FALLBACK_MAX_TOKENS
 from danswer.db.models import Persona
 from danswer.db.models import User
+from danswer.key_value_store.interface import JSON_ro
 from danswer.llm.answering.llm_response_handler import LLMCall
 from danswer.llm.answering.models import AnswerStyleConfig
 from danswer.llm.answering.models import ContextualPruningConfig
@@ -48,7 +49,6 @@ from danswer.tools.tool_implementations.search_like_tool_utils import (
     FINAL_CONTEXT_DOCUMENTS_ID,
 )
 from danswer.utils.logger import setup_logger
-from danswer.utils.special_types import JSON_ro
 
 logger = setup_logger()
 

backend/danswer/tools/tool_implementations/search_like_tool_utils.py

@@ -55,12 +55,9 @@ def build_next_prompt_for_search_like_tool(
             )
         )
     elif answer_style_config.quotes_config:
-        # For Quotes, the system prompt is included in the user prompt
-        prompt_builder.update_system_prompt(None)
-
         prompt_builder.update_user_prompt(
             build_quotes_user_message(
-                message=prompt_builder.raw_user_message,
+                message=prompt_builder.user_message_and_token_cnt[0],
                 context_docs=final_context_documents,
                 history_str=prompt_builder.single_message_history or "",
                 prompt=prompt_config,

backend/danswer/utils/long_term_log.py

@@ -1,115 +0,0 @@
-import json
-import os
-import threading
-from datetime import datetime
-from pathlib import Path
-from typing import Any
-
-from danswer.utils.logger import setup_logger
-from danswer.utils.special_types import JSON_ro
-
-logger = setup_logger()
-
-_LOG_FILE_NAME_TIMESTAMP_FORMAT = "%Y-%m-%d_%H-%M-%S-%f"
-
-
-class LongTermLogger:
-    """NOTE: should support a LOT of data AND should be extremely fast,
-    ideally done in a background thread."""
-
-    def __init__(
-        self,
-        metadata: dict[str, str] | None = None,
-        log_file_path: str = "/tmp/long_term_log",
-        max_files_per_category: int = 1000,
-    ):
-        self.metadata = metadata
-        self.log_file_path = Path(log_file_path)
-        self.max_files_per_category = max_files_per_category
-        try:
-            # Create directory if it doesn't exist
-            os.makedirs(os.path.dirname(log_file_path), exist_ok=True)
-        except Exception as e:
-            logger.error(f"Error creating directory for long-term logs: {e}")
-
-    def _cleanup_old_files(self, category_path: Path) -> None:
-        try:
-            files = sorted(
-                category_path.glob("*.json"),
-                key=lambda x: x.stat().st_mtime,  # Sort by modification time
-                reverse=True,
-            )
-            # Delete oldest files that exceed the limit
-            for file in files[self.max_files_per_category :]:
-                try:
-                    file.unlink()
-                except Exception as e:
-                    logger.error(f"Error deleting old log file {file}: {e}")
-        except Exception as e:
-            logger.error(f"Error during log rotation cleanup: {e}")
-
-    def _record(self, message: Any, category: str) -> None:
-        category_path = self.log_file_path / category
-        try:
-            # Create directory if it doesn't exist
-            os.makedirs(category_path, exist_ok=True)
-
-            # Perform cleanup before writing new file
-            self._cleanup_old_files(category_path)
-
-            final_record = {
-                "metadata": self.metadata,
-                "record": message,
-            }
-
-            file_path = (
-                category_path
-                / f"{datetime.now().strftime(_LOG_FILE_NAME_TIMESTAMP_FORMAT)}.json"
-            )
-            with open(file_path, "w+") as f:
-                # default allows us to "ignore" unserializable objects
-                json.dump(final_record, f, default=lambda x: str(x))
-        except Exception as e:
-            logger.error(f"Error recording log: {e}")
-
-    def record(self, message: JSON_ro, category: str = "default") -> None:
-        try:
-            # Run in separate thread to have minimal overhead in main flows
-            thread = threading.Thread(
-                target=self._record, args=(message, category), daemon=True
-            )
-            thread.start()
-        except Exception:
-            # Should never interfere with normal functions of Danswer
-            pass
-
-    def fetch_category(
-        self,
-        category: str,
-        start_time: datetime | None = None,
-        end_time: datetime | None = None,
-        limit: int = 100,
-    ) -> list[JSON_ro]:
-        category_path = self.log_file_path / category
-        files = list(category_path.glob("*.json"))
-
-        results: list[JSON_ro] = []
-        for file in files:
-            # Parse timestamp from filename (YYYY-MM-DD_HH-MM-SS.json)
-            try:
-                file_time = datetime.strptime(
-                    file.stem, _LOG_FILE_NAME_TIMESTAMP_FORMAT
-                )
-
-                # Skip if outside time range
-                if start_time and file_time < start_time:
-                    continue
-                if end_time and file_time > end_time:
-                    continue
-
-                results.append(json.loads(file.read_text()))
-            except ValueError:
-                # Skip files that don't match expected format
-                continue
-
-        return results

backend/danswer/utils/special_types.py

@@ -1,7 +0,0 @@
-from collections.abc import Mapping
-from collections.abc import Sequence
-from typing import TypeAlias
-
-JSON_ro: TypeAlias = (
-    Mapping[str, "JSON_ro"] | Sequence["JSON_ro"] | str | int | float | bool | None
-)

backend/danswer/utils/text_processing.py

@@ -4,10 +4,6 @@ import re
 import string
 from urllib.parse import quote
 
-from danswer.utils.logger import setup_logger
-
-
-logger = setup_logger(__name__)
 
 ESCAPE_SEQUENCE_RE = re.compile(
     r"""
@@ -81,8 +77,7 @@ def extract_embedded_json(s: str) -> dict:
     last_brace_index = s.rfind("}")
 
     if first_brace_index == -1 or last_brace_index == -1:
-        logger.warning("No valid json found, assuming answer is entire string")
-        return {"answer": s, "quotes": []}
+        raise ValueError("No valid json found")
 
     json_str = s[first_brace_index : last_brace_index + 1]
     try:

backend/ee/danswer/danswerbot/slack/handlers/handle_standard_answers.py

@@ -19,7 +19,7 @@ from danswer.db.chat import get_chat_messages_by_sessions
 from danswer.db.chat import get_chat_sessions_by_slack_thread_id
 from danswer.db.chat import get_or_create_root_message
 from danswer.db.models import Prompt
-from danswer.db.models import SlackChannelConfig
+from danswer.db.models import SlackBotConfig
 from danswer.db.models import StandardAnswer as StandardAnswerModel
 from danswer.utils.logger import DanswerLoggingAdapter
 from danswer.utils.logger import setup_logger
@@ -80,7 +80,7 @@ def oneoff_standard_answers(
 def _handle_standard_answers(
     message_info: SlackMessageInfo,
     receiver_ids: list[str] | None,
-    slack_channel_config: SlackChannelConfig | None,
+    slack_bot_config: SlackBotConfig | None,
     prompt: Prompt | None,
     logger: DanswerLoggingAdapter,
     client: WebClient,
@@ -95,12 +95,12 @@ def _handle_standard_answers(
     we still need to respond to the users.
     """
     # if no channel config, then no standard answers are configured
-    if not slack_channel_config:
+    if not slack_bot_config:
         return False
 
     slack_thread_id = message_info.thread_to_respond
     configured_standard_answer_categories = (
-        slack_channel_config.standard_answer_categories if slack_channel_config else []
+        slack_bot_config.standard_answer_categories if slack_bot_config else []
     )
     configured_standard_answers = set(
         [
@@ -150,9 +150,7 @@ def _handle_standard_answers(
             db_session=db_session,
             description="",
             user_id=None,
-            persona_id=slack_channel_config.persona.id
-            if slack_channel_config.persona
-            else 0,
+            persona_id=slack_bot_config.persona.id if slack_bot_config.persona else 0,
             danswerbot_flow=True,
             slack_thread_id=slack_thread_id,
             one_shot=True,

backend/ee/danswer/db/external_perm.py

@@ -9,7 +9,7 @@ from sqlalchemy.orm import Session
 from danswer.access.utils import prefix_group_w_source
 from danswer.configs.constants import DocumentSource
 from danswer.db.models import User__ExternalUserGroupId
-from danswer.db.users import batch_add_ext_perm_user_if_not_exists
+from danswer.db.users import batch_add_non_web_user_if_not_exists__no_commit
 
 
 class ExternalUserGroup(BaseModel):
@@ -49,6 +49,10 @@ def replace_user__ext_group_for_cc_pair(
     This function clears all existing external user group relations for a given cc_pair_id
     and replaces them with the new group definitions and commits the changes.
     """
+    delete_user__ext_group_for_cc_pair__no_commit(
+        db_session=db_session,
+        cc_pair_id=cc_pair_id,
+    )
 
     # collect all emails from all groups to batch add all users at once for efficiency
     all_group_member_emails = set()
@@ -57,15 +61,10 @@ def replace_user__ext_group_for_cc_pair(
             all_group_member_emails.add(user_email)
 
     # batch add users if they don't exist and get their ids
-    all_group_members = batch_add_ext_perm_user_if_not_exists(
+    all_group_members = batch_add_non_web_user_if_not_exists__no_commit(
         db_session=db_session, emails=list(all_group_member_emails)
     )
 
-    delete_user__ext_group_for_cc_pair__no_commit(
-        db_session=db_session,
-        cc_pair_id=cc_pair_id,
-    )
-
     # map emails to ids
     email_id_map = {user.email: user.id for user in all_group_members}
 

backend/ee/danswer/db/user_group.py

@@ -411,8 +411,6 @@ def _validate_curator_status__no_commit(
             .all()
         )
 
-        # if the user is a curator in any of their groups, set their role to CURATOR
-        # otherwise, set their role to BASIC
         if curator_relationships:
             user.role = UserRole.CURATOR
         elif user.role == UserRole.CURATOR:
@@ -438,15 +436,6 @@ def update_user_curator_relationship(
     user = fetch_user_by_id(db_session, set_curator_request.user_id)
     if not user:
         raise ValueError(f"User with id '{set_curator_request.user_id}' not found")
-
-    if user.role == UserRole.ADMIN:
-        raise ValueError(
-            f"User '{user.email}' is an admin and therefore has all permissions "
-            "of a curator. If you'd like this user to only have curator permissions, "
-            "you must update their role to BASIC then assign them to be CURATOR in the "
-            "appropriate groups."
-        )
-
     requested_user_groups = fetch_user_groups_for_user(
         db_session=db_session,
         user_id=set_curator_request.user_id,

backend/ee/danswer/external_permissions/confluence/group_sync.py

@@ -1,5 +1,7 @@
-from danswer.connectors.confluence.onyx_confluence import build_confluence_client
+from typing import Any
+
 from danswer.connectors.confluence.onyx_confluence import OnyxConfluence
+from danswer.connectors.confluence.utils import build_confluence_client
 from danswer.connectors.confluence.utils import get_user_email_from_username__server
 from danswer.db.models import ConnectorCredentialPair
 from danswer.utils.logger import setup_logger
@@ -9,30 +11,26 @@ from ee.danswer.db.external_perm import ExternalUserGroup
 logger = setup_logger()
 
 
-def _build_group_member_email_map(
+def _get_group_members_email_paginated(
     confluence_client: OnyxConfluence,
-) -> dict[str, set[str]]:
-    group_member_emails: dict[str, set[str]] = {}
-    for user_result in confluence_client.paginated_cql_user_retrieval():
-        user = user_result["user"]
-        email = user.get("email")
+    group_name: str,
+) -> set[str]:
+    members: list[dict[str, Any]] = []
+    for member_batch in confluence_client.paginated_group_members_retrieval(group_name):
+        members.extend(member_batch)
+
+    group_member_emails: set[str] = set()
+    for member in members:
+        email = member.get("email")
         if not email:
-            # This field is only present in Confluence Server
-            user_name = user.get("username")
-            # If it is present, try to get the email using a Server-specific method
+            user_name = member.get("username")
             if user_name:
                 email = get_user_email_from_username__server(
                     confluence_client=confluence_client,
                     user_name=user_name,
                 )
-        if not email:
-            # If we still don't have an email, skip this user
-            continue
-
-        for group in confluence_client.paginated_groups_by_user_retrieval(user):
-            # group name uniqueness is enforced by Confluence, so we can use it as a group ID
-            group_id = group["name"]
-            group_member_emails.setdefault(group_id, set()).add(email)
+        if email:
+            group_member_emails.add(email)
 
     return group_member_emails
 
@@ -40,20 +38,31 @@ def _build_group_member_email_map(
 def confluence_group_sync(
     cc_pair: ConnectorCredentialPair,
 ) -> list[ExternalUserGroup]:
+    is_cloud = cc_pair.connector.connector_specific_config.get("is_cloud", False)
     confluence_client = build_confluence_client(
-        credentials=cc_pair.credential.credential_json,
-        is_cloud=cc_pair.connector.connector_specific_config.get("is_cloud", False),
+        credentials_json=cc_pair.credential.credential_json,
+        is_cloud=is_cloud,
         wiki_base=cc_pair.connector.connector_specific_config["wiki_base"],
     )
 
-    group_member_email_map = _build_group_member_email_map(
-        confluence_client=confluence_client,
-    )
+    # Get all group names
+    group_names: list[str] = []
+    for group_batch in confluence_client.paginated_groups_retrieval():
+        for group in group_batch:
+            if group_name := group.get("name"):
+                group_names.append(group_name)
+
+    # For each group name, get all members and create a danswer group
     danswer_groups: list[ExternalUserGroup] = []
-    for group_id, group_member_emails in group_member_email_map.items():
+    for group_name in group_names:
+        group_member_emails = _get_group_members_email_paginated(
+            confluence_client, group_name
+        )
+        if not group_member_emails:
+            continue
         danswer_groups.append(
             ExternalUserGroup(
-                id=group_id,
+                id=group_name,
                 user_emails=list(group_member_emails),
             )
         )