update

.github/CODEOWNERS

@@ -1 +0,0 @@
-* @onyx-dot-app/onyx-core-team

.github/workflows/docker-build-push-model-server-container-on-tag.yml

@@ -12,40 +12,29 @@ env:
   BUILDKIT_PROGRESS: plain
 
 jobs:
-
-#   Bypassing this for now as the idea of not building is glitching
-#   releases and builds that depends on everything being tagged in docker
-#   1) Preliminary job to check if the changed files are relevant
-#   check_model_server_changes:
-#     runs-on: ubuntu-latest
-#     outputs:
-#       changed: ${{ steps.check.outputs.changed }}
-#     steps:
-#       - name: Checkout code
-#         uses: actions/checkout@v4
-# 
-#       - name: Check if relevant files changed
-#         id: check
-#         run: |
-#           # Default to "false"
-#           echo "changed=false" >> $GITHUB_OUTPUT
-# 
-#           # Compare the previous commit (github.event.before) to the current one (github.sha)
-#           # If any file in backend/model_server/** or backend/Dockerfile.model_server is changed,
-#           # set changed=true
-#           if git diff --name-only ${{ github.event.before }} ${{ github.sha }} \
-#              | grep -E '^backend/model_server/|^backend/Dockerfile.model_server'; then
-#             echo "changed=true" >> $GITHUB_OUTPUT
-#           fi
-
+  # 1) Preliminary job to check if the changed files are relevant
   check_model_server_changes:
     runs-on: ubuntu-latest
     outputs:
-      changed: "true"
+      changed: ${{ steps.check.outputs.changed }}
     steps:
-      - name: Bypass check and set output
-        run: echo "changed=true" >> $GITHUB_OUTPUT
-        
+      - name: Checkout code
+        uses: actions/checkout@v4
+
+      - name: Check if relevant files changed
+        id: check
+        run: |
+          # Default to "false"
+          echo "changed=false" >> $GITHUB_OUTPUT
+
+          # Compare the previous commit (github.event.before) to the current one (github.sha)
+          # If any file in backend/model_server/** or backend/Dockerfile.model_server is changed,
+          # set changed=true
+          if git diff --name-only ${{ github.event.before }} ${{ github.sha }} \
+             | grep -E '^backend/model_server/|^backend/Dockerfile.model_server'; then
+            echo "changed=true" >> $GITHUB_OUTPUT
+          fi
+
   build-amd64:
     needs: [check_model_server_changes]
     if: needs.check_model_server_changes.outputs.changed == 'true'

.github/workflows/nightly-scan-licenses.yml

@@ -53,90 +53,24 @@ jobs:
           exclude: '(?i)^(pylint|aio[-_]*).*'
           
       - name: Print report
-        if: always()
+        if: ${{ always() }}
         run: echo "${{ steps.license_check_report.outputs.report }}"
       
       - name: Install npm dependencies
         working-directory: ./web
         run: npm ci
+        
+      - name: Run Trivy vulnerability scanner in repo mode
+        uses: aquasecurity/trivy-action@0.28.0
+        with:
+          scan-type: fs
+          scanners: license
+          format: table
+#           format: sarif
+#           output: trivy-results.sarif
+          severity: HIGH,CRITICAL
 
-        # be careful enabling the sarif and upload as it may spam the security tab
-        # with a huge amount of items. Work out the issues before enabling upload.       
-#       - name: Run Trivy vulnerability scanner in repo mode
-#         if: always()
-#         uses: aquasecurity/trivy-action@0.29.0
+#       - name: Upload Trivy scan results to GitHub Security tab
+#         uses: github/codeql-action/upload-sarif@v3
 #         with:
-#           scan-type: fs
-#           scan-ref: .
-#           scanners: license
-#           format: table
-#           severity: HIGH,CRITICAL
-# #           format: sarif
-# #           output: trivy-results.sarif
-# 
-# #       - name: Upload Trivy scan results to GitHub Security tab
-# #         uses: github/codeql-action/upload-sarif@v3
-# #         with:
-# #           sarif_file: trivy-results.sarif
-
-  scan-trivy:
-    # See https://runs-on.com/runners/linux/
-    runs-on: [runs-on,runner=2cpu-linux-x64,"run-id=${{ github.run_id }}"]
-      
-    steps:
-    - 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 }}
-
-    # Backend
-    - name: Pull backend docker image
-      run: docker pull onyxdotapp/onyx-backend:latest
-
-    - name: Run Trivy vulnerability scanner on backend
-      uses: aquasecurity/trivy-action@0.29.0
-      env:
-        TRIVY_DB_REPOSITORY: 'public.ecr.aws/aquasecurity/trivy-db:2'
-        TRIVY_JAVA_DB_REPOSITORY: 'public.ecr.aws/aquasecurity/trivy-java-db:1'
-      with:
-        image-ref: onyxdotapp/onyx-backend:latest
-        scanners: license
-        severity: HIGH,CRITICAL
-        vuln-type: library
-        exit-code: 0  # Set to 1 if we want a failed scan to fail the workflow
-
-    # Web server
-    - name: Pull web server docker image
-      run: docker pull onyxdotapp/onyx-web-server:latest
-          
-    - name: Run Trivy vulnerability scanner on web server
-      uses: aquasecurity/trivy-action@0.29.0
-      env:
-        TRIVY_DB_REPOSITORY: 'public.ecr.aws/aquasecurity/trivy-db:2'
-        TRIVY_JAVA_DB_REPOSITORY: 'public.ecr.aws/aquasecurity/trivy-java-db:1'
-      with:
-        image-ref: onyxdotapp/onyx-web-server:latest
-        scanners: license
-        severity: HIGH,CRITICAL
-        vuln-type: library
-        exit-code: 0
-
-    # Model server
-    - name: Pull model server docker image
-      run: docker pull onyxdotapp/onyx-model-server:latest
-
-    - name: Run Trivy vulnerability scanner
-      uses: aquasecurity/trivy-action@0.29.0
-      env:
-        TRIVY_DB_REPOSITORY: 'public.ecr.aws/aquasecurity/trivy-db:2'
-        TRIVY_JAVA_DB_REPOSITORY: 'public.ecr.aws/aquasecurity/trivy-java-db:1'
-      with:
-        image-ref: onyxdotapp/onyx-model-server:latest
-        scanners: license
-        severity: HIGH,CRITICAL
-        vuln-type: library
-        exit-code: 0
+#           sarif_file: trivy-results.sarif

.github/workflows/pr-mit-integration-tests.yml

@@ -1,209 +0,0 @@
-name: Run MIT Integration Tests v2
-concurrency:
-  group: Run-MIT-Integration-Tests-${{ github.workflow }}-${{ github.head_ref || github.event.workflow_run.head_branch || github.run_id }}
-  cancel-in-progress: true
-
-on:
-  merge_group:
-  pull_request:
-    branches:
-      - main
-      - "release/**"
-
-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: ${{ secrets.CONFLUENCE_ACCESS_TOKEN }}
-
-jobs:
-  integration-tests-mit:
-    # See https://runs-on.com/runners/linux/
-    runs-on: [runs-on, runner=32cpu-linux-x64, "run-id=${{ github.run_id }}"]
-    steps:
-      - name: Checkout code
-        uses: actions/checkout@v4
-
-      - 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: ${{ secrets.DOCKER_TOKEN }}
-
-      # tag every docker image with "test" so that we can spin up the correct set
-      # of images during testing
-
-      # We don't need to build the Web Docker image since it's not yet used
-      # in the integration tests. We have a separate action to verify that it builds
-      # successfully.
-      - name: Pull Web Docker image
-        run: |
-          docker pull onyxdotapp/onyx-web-server:latest
-          docker tag onyxdotapp/onyx-web-server:latest onyxdotapp/onyx-web-server:test
-
-      # we use the runs-on cache for docker builds
-      # in conjunction with runs-on runners, it has better speed and unlimited caching
-      # https://runs-on.com/caching/s3-cache-for-github-actions/
-      # https://runs-on.com/caching/docker/
-      # https://github.com/moby/buildkit#s3-cache-experimental
-
-      # images are built and run locally for testing purposes. Not pushed.
-      - name: Build Backend Docker image
-        uses: ./.github/actions/custom-build-and-push
-        with:
-          context: ./backend
-          file: ./backend/Dockerfile
-          platforms: linux/amd64
-          tags: onyxdotapp/onyx-backend:test
-          push: false
-          load: true
-          cache-from: type=s3,prefix=cache/${{ 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=${{ env.RUNS_ON_S3_BUCKET_CACHE }},mode=max
-
-      - name: Build Model Server Docker image
-        uses: ./.github/actions/custom-build-and-push
-        with:
-          context: ./backend
-          file: ./backend/Dockerfile.model_server
-          platforms: linux/amd64
-          tags: onyxdotapp/onyx-model-server:test
-          push: false
-          load: true
-          cache-from: type=s3,prefix=cache/${{ 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=${{ env.RUNS_ON_S3_BUCKET_CACHE }},mode=max
-
-      - name: Build integration test Docker image
-        uses: ./.github/actions/custom-build-and-push
-        with:
-          context: ./backend
-          file: ./backend/tests/integration/Dockerfile
-          platforms: linux/amd64
-          tags: onyxdotapp/onyx-integration:test
-          push: false
-          load: true
-          cache-from: type=s3,prefix=cache/${{ github.repository }}/integration-tests/integration/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }}
-          cache-to: type=s3,prefix=cache/${{ github.repository }}/integration-tests/integration/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }},mode=max
-
-      # NOTE: Use pre-ping/null pool to reduce flakiness due to dropped connections
-      - name: Start Docker containers
-        run: |
-          cd deployment/docker_compose
-          AUTH_TYPE=basic \
-          POSTGRES_POOL_PRE_PING=true \
-          POSTGRES_USE_NULL_POOL=true \
-          REQUIRE_EMAIL_VERIFICATION=false \
-          DISABLE_TELEMETRY=true \
-          IMAGE_TAG=test \
-          INTEGRATION_TESTS_MODE=true \
-          docker compose -f docker-compose.dev.yml -p onyx-stack up -d
-        id: start_docker
-
-      - name: Wait for service to be ready
-        run: |
-          echo "Starting wait-for-service script..."
-
-          docker logs -f onyx-stack-api_server-1 &
-
-          start_time=$(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 $response). Retrying in 5 seconds..."
-            fi
-            
-            sleep 5
-          done
-          echo "Finished waiting for service."
-
-      - name: Start Mock Services
-        run: |
-          cd backend/tests/integration/mock_services
-          docker compose -f docker-compose.mock-it-services.yml \
-            -p mock-it-services-stack up -d
-
-      # NOTE: Use pre-ping/null to reduce flakiness due to dropped connections
-      - name: Run Standard Integration Tests
-        run: |
-          echo "Running integration tests..."
-          docker run --rm --network onyx-stack_default \
-            --name test-runner \
-            -e POSTGRES_HOST=relational_db \
-            -e POSTGRES_USER=postgres \
-            -e POSTGRES_PASSWORD=password \
-            -e POSTGRES_DB=postgres \
-            -e POSTGRES_POOL_PRE_PING=true \
-            -e POSTGRES_USE_NULL_POOL=true \
-            -e VESPA_HOST=index \
-            -e REDIS_HOST=cache \
-            -e API_SERVER_HOST=api_server \
-            -e OPENAI_API_KEY=${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 \
-            -e MOCK_CONNECTOR_SERVER_HOST=mock_connector_server \
-            -e MOCK_CONNECTOR_SERVER_PORT=8001 \
-            onyxdotapp/onyx-integration:test \
-            /app/tests/integration/tests \
-            /app/tests/integration/connector_job_tests
-        continue-on-error: true
-        id: run_tests
-
-      - name: Check test results
-        run: |
-          if [ ${{ steps.run_tests.outcome }} == 'failure' ]; then
-            echo "Integration tests failed. Exiting with error."
-            exit 1
-          else
-            echo "All integration tests passed successfully."
-          fi
-
-      # ------------------------------------------------------------
-      # Always gather logs BEFORE "down":
-      - name: Dump API server logs
-        if: always()
-        run: |
-          cd deployment/docker_compose
-          docker compose -f docker-compose.dev.yml -p onyx-stack logs --no-color api_server > $GITHUB_WORKSPACE/api_server.log || true
-
-      - name: Dump all-container logs (optional)
-        if: always()
-        run: |
-          cd deployment/docker_compose
-          docker compose -f docker-compose.dev.yml -p onyx-stack logs --no-color > $GITHUB_WORKSPACE/docker-compose.log || true
-
-      - name: Upload logs
-        if: always()
-        uses: actions/upload-artifact@v4
-        with:
-          name: docker-all-logs
-          path: ${{ github.workspace }}/docker-compose.log
-      # ------------------------------------------------------------
-
-      - name: Stop Docker containers
-        if: always()
-        run: |
-          cd deployment/docker_compose
-          docker compose -f docker-compose.dev.yml -p onyx-stack down -v

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

@@ -1,7 +1,6 @@
 name: Connector Tests
 
 on:
-  merge_group:
   pull_request:
     branches: [main]
   schedule:
@@ -9,10 +8,6 @@ on:
     - cron: "0 16 * * *"
 
 env:
-  # AWS
-  AWS_ACCESS_KEY_ID_DAILY_CONNECTOR_TESTS: ${{ secrets.AWS_ACCESS_KEY_ID_DAILY_CONNECTOR_TESTS }}
-  AWS_SECRET_ACCESS_KEY_DAILY_CONNECTOR_TESTS: ${{ secrets.AWS_SECRET_ACCESS_KEY_DAILY_CONNECTOR_TESTS }}
-  
   # Confluence
   CONFLUENCE_TEST_SPACE_URL: ${{ secrets.CONFLUENCE_TEST_SPACE_URL }}
   CONFLUENCE_TEST_SPACE: ${{ secrets.CONFLUENCE_TEST_SPACE }}
@@ -49,21 +44,14 @@ env:
   SHAREPOINT_CLIENT_SECRET: ${{ secrets.SHAREPOINT_CLIENT_SECRET }}
   SHAREPOINT_CLIENT_DIRECTORY_ID: ${{ secrets.SHAREPOINT_CLIENT_DIRECTORY_ID }}
   SHAREPOINT_SITE: ${{ secrets.SHAREPOINT_SITE }}
-  # Github
-  ACCESS_TOKEN_GITHUB: ${{ secrets.ACCESS_TOKEN_GITHUB }}
   # Gitbook
   GITBOOK_SPACE_ID: ${{ secrets.GITBOOK_SPACE_ID }}
   GITBOOK_API_KEY: ${{ secrets.GITBOOK_API_KEY }}
-  # Notion
-  NOTION_INTEGRATION_TOKEN: ${{ secrets.NOTION_INTEGRATION_TOKEN }}
-  # Highspot
-  HIGHSPOT_KEY: ${{ secrets.HIGHSPOT_KEY }}
-  HIGHSPOT_SECRET: ${{ secrets.HIGHSPOT_SECRET }}
 
 jobs:
   connectors-check:
     # See https://runs-on.com/runners/linux/
-    runs-on: [runs-on, runner=8cpu-linux-x64, "run-id=${{ github.run_id }}"]
+    runs-on: [runs-on,runner=8cpu-linux-x64,"run-id=${{ github.run_id }}"]
 
     env:
       PYTHONPATH: ./backend
@@ -88,7 +76,7 @@ jobs:
           pip install --retries 5 --timeout 30 -r backend/requirements/dev.txt
           playwright install chromium
           playwright install-deps chromium
-
+          
       - name: Run Tests
         shell: script -q -e -c "bash --noprofile --norc -eo pipefail {0}"
         run: py.test -o junit_family=xunit2 -xv --ff backend/tests/daily/connectors

.vscode/launch.template.jsonc

@@ -6,419 +6,396 @@
     // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
     "version": "0.2.0",
     "compounds": [
-      {
-        // Dummy entry used to label the group
-        "name": "--- Compound ---",
-        "configurations": ["--- Individual ---"],
-        "presentation": {
-          "group": "1"
-        }
-      },
-      {
-        "name": "Run All Onyx Services",
-        "configurations": [
-          "Web Server",
-          "Model Server",
-          "API Server",
-          "Slack Bot",
-          "Celery primary",
-          "Celery light",
-          "Celery heavy",
-          "Celery indexing",
-          "Celery user files indexing",
-          "Celery beat",
-          "Celery monitoring"
-        ],
-        "presentation": {
-          "group": "1"
-        }
-      },
-      {
-        "name": "Web / Model / API",
-        "configurations": ["Web Server", "Model Server", "API Server"],
-        "presentation": {
-          "group": "1"
-        }
-      },
-      {
-        "name": "Celery (all)",
-        "configurations": [
-          "Celery primary",
-          "Celery light",
-          "Celery heavy",
-          "Celery indexing",
-          "Celery user files indexing",
-          "Celery beat",
-          "Celery monitoring"
-        ],
-        "presentation": {
-          "group": "1"
-        }
-      }
+		{
+			// Dummy entry used to label the group
+			"name": "--- Compound ---",
+		  	"configurations": [
+		  		"--- Individual ---"
+		  	],
+			"presentation": {
+				 "group": "1",
+			 }
+		},
+        {
+            "name": "Run All Onyx Services",
+            "configurations": [
+                "Web Server",
+                "Model Server",
+                "API Server",
+                "Slack Bot",
+		  		"Celery primary", 
+		  		"Celery light", 
+		  		"Celery heavy", 
+		  		"Celery indexing", 
+		  		"Celery beat",
+                "Celery monitoring",
+            ],
+			"presentation": {
+				 "group": "1",
+			 }
+        },
+		{
+			"name": "Web / Model / API",
+		  	"configurations": [
+		  		"Web Server", 
+		  		"Model Server", 
+		  		"API Server",
+		  	],
+			"presentation": {
+				 "group": "1",
+			 }
+		},
+		{
+			"name": "Celery (all)",
+		  	"configurations": [
+		  		"Celery primary", 
+		  		"Celery light", 
+		  		"Celery heavy", 
+		  		"Celery indexing", 
+		  		"Celery beat",
+                "Celery monitoring",
+		  	],
+			"presentation": {
+				 "group": "1",
+			 }
+		}
     ],
     "configurations": [
-      {
-        // Dummy entry used to label the group
-        "name": "--- Individual ---",
-        "type": "node",
-        "request": "launch",
-        "presentation": {
-          "group": "2",
-          "order": 0
-        }
-      },
-      {
-        "name": "Web Server",
-        "type": "node",
-        "request": "launch",
-        "cwd": "${workspaceRoot}/web",
-        "runtimeExecutable": "npm",
-        "envFile": "${workspaceFolder}/.vscode/.env",
-        "runtimeArgs": ["run", "dev"],
-        "presentation": {
-          "group": "2"
+		{
+	    	// Dummy entry used to label the group
+			"name": "--- Individual ---",
+			"type": "node",
+			"request": "launch",
+			"presentation": {
+				"group": "2",
+				"order": 0
+			}
+		},
+        {
+            "name": "Web Server",
+            "type": "node",
+            "request": "launch",
+            "cwd": "${workspaceRoot}/web",
+            "runtimeExecutable": "npm",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "runtimeArgs": [
+                "run", "dev"
+            ],
+            "presentation": {
+				 "group": "2",
+			 },
+            "console": "integratedTerminal",
+            "consoleTitle": "Web Server Console"
         },
-        "console": "integratedTerminal",
-        "consoleTitle": "Web Server Console"
-      },
-      {
-        "name": "Model Server",
-        "consoleName": "Model Server",
-        "type": "debugpy",
-        "request": "launch",
-        "module": "uvicorn",
-        "cwd": "${workspaceFolder}/backend",
-        "envFile": "${workspaceFolder}/.vscode/.env",
-        "env": {
-          "LOG_LEVEL": "DEBUG",
-          "PYTHONUNBUFFERED": "1"
+        {
+            "name": "Model Server",
+            "consoleName": "Model Server",
+            "type": "debugpy",
+            "request": "launch",
+            "module": "uvicorn",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {
+                "LOG_LEVEL": "DEBUG",
+                "PYTHONUNBUFFERED": "1"
+            },
+            "args": [
+                "model_server.main:app",
+                "--reload",
+                "--port",
+                "9000"
+            ],
+            "presentation": {
+				 "group": "2",
+			},
+            "consoleTitle": "Model Server Console"
         },
-        "args": ["model_server.main:app", "--reload", "--port", "9000"],
-        "presentation": {
-          "group": "2"
+        {
+            "name": "API Server",
+            "consoleName": "API Server",
+            "type": "debugpy",
+            "request": "launch",
+            "module": "uvicorn",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {
+                "LOG_DANSWER_MODEL_INTERACTIONS": "True",
+                "LOG_LEVEL": "DEBUG",
+                "PYTHONUNBUFFERED": "1"
+            },
+            "args": [
+                "onyx.main:app",
+                "--reload",
+                "--port",
+                "8080"
+            ],
+            "presentation": {
+				 "group": "2",
+			 },
+            "consoleTitle": "API Server Console"
         },
-        "consoleTitle": "Model Server Console"
-      },
-      {
-        "name": "API Server",
-        "consoleName": "API Server",
-        "type": "debugpy",
-        "request": "launch",
-        "module": "uvicorn",
-        "cwd": "${workspaceFolder}/backend",
-        "envFile": "${workspaceFolder}/.vscode/.env",
-        "env": {
-          "LOG_DANSWER_MODEL_INTERACTIONS": "True",
-          "LOG_LEVEL": "DEBUG",
-          "PYTHONUNBUFFERED": "1"
+        // For the listener to access the Slack API,
+        // DANSWER_BOT_SLACK_APP_TOKEN & DANSWER_BOT_SLACK_BOT_TOKEN need to be set in .env file located in the root of the project
+        {
+            "name": "Slack Bot",
+            "consoleName": "Slack Bot",
+            "type": "debugpy",
+            "request": "launch",
+            "program": "onyx/onyxbot/slack/listener.py",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {
+                "LOG_LEVEL": "DEBUG",
+                "PYTHONUNBUFFERED": "1",
+                "PYTHONPATH": "."
+            },
+            "presentation": {
+				 "group": "2",
+			},
+            "consoleTitle": "Slack Bot Console"
         },
-        "args": ["onyx.main:app", "--reload", "--port", "8080"],
-        "presentation": {
-          "group": "2"
+        {
+            "name": "Celery primary",
+            "type": "debugpy",
+            "request": "launch",
+            "module": "celery",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {
+                "LOG_LEVEL": "INFO",
+                "PYTHONUNBUFFERED": "1",
+                "PYTHONPATH": "."
+            },
+            "args": [
+                "-A",
+                "onyx.background.celery.versioned_apps.primary",
+                "worker",
+                "--pool=threads",
+                "--concurrency=4",
+                "--prefetch-multiplier=1",
+                "--loglevel=INFO",
+                "--hostname=primary@%n",
+                "-Q",
+                "celery",
+            ],
+            "presentation": {
+				 "group": "2",
+			 },
+            "consoleTitle": "Celery primary Console"
         },
-        "consoleTitle": "API Server Console"
-      },
-      // For the listener to access the Slack API,
-      // DANSWER_BOT_SLACK_APP_TOKEN & DANSWER_BOT_SLACK_BOT_TOKEN need to be set in .env file located in the root of the project
-      {
-        "name": "Slack Bot",
-        "consoleName": "Slack Bot",
-        "type": "debugpy",
-        "request": "launch",
-        "program": "onyx/onyxbot/slack/listener.py",
-        "cwd": "${workspaceFolder}/backend",
-        "envFile": "${workspaceFolder}/.vscode/.env",
-        "env": {
-          "LOG_LEVEL": "DEBUG",
-          "PYTHONUNBUFFERED": "1",
-          "PYTHONPATH": "."
+        {
+            "name": "Celery light",
+            "type": "debugpy",
+            "request": "launch",
+            "module": "celery",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {
+                "LOG_LEVEL": "INFO",
+                "PYTHONUNBUFFERED": "1",
+                "PYTHONPATH": "."
+            },
+            "args": [
+                "-A",
+                "onyx.background.celery.versioned_apps.light",
+                "worker",
+                "--pool=threads",
+                "--concurrency=64",
+                "--prefetch-multiplier=8",
+                "--loglevel=INFO",
+                "--hostname=light@%n",
+                "-Q",
+                "vespa_metadata_sync,connector_deletion,doc_permissions_upsert,checkpoint_cleanup",
+            ],
+            "presentation": {
+				 "group": "2",
+			 },
+            "consoleTitle": "Celery light Console"
         },
-        "presentation": {
-          "group": "2"
+        {
+            "name": "Celery heavy",
+            "type": "debugpy",
+            "request": "launch",
+            "module": "celery",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {
+                "LOG_LEVEL": "INFO",
+                "PYTHONUNBUFFERED": "1",
+                "PYTHONPATH": "."
+            },
+            "args": [
+                "-A",
+                "onyx.background.celery.versioned_apps.heavy",
+                "worker",
+                "--pool=threads",
+                "--concurrency=4",
+                "--prefetch-multiplier=1",
+                "--loglevel=INFO",
+                "--hostname=heavy@%n",
+                "-Q",
+                "connector_pruning,connector_doc_permissions_sync,connector_external_group_sync",
+            ],
+            "presentation": {
+				 "group": "2",
+			 },
+            "consoleTitle": "Celery heavy Console"
         },
-        "consoleTitle": "Slack Bot Console"
-      },
-      {
-        "name": "Celery primary",
-        "type": "debugpy",
-        "request": "launch",
-        "module": "celery",
-        "cwd": "${workspaceFolder}/backend",
-        "envFile": "${workspaceFolder}/.vscode/.env",
-        "env": {
-          "LOG_LEVEL": "INFO",
-          "PYTHONUNBUFFERED": "1",
-          "PYTHONPATH": "."
+        {
+            "name": "Celery indexing",
+            "type": "debugpy",
+            "request": "launch",
+            "module": "celery",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {
+                "ENABLE_MULTIPASS_INDEXING": "false",
+                "LOG_LEVEL": "DEBUG",
+                "PYTHONUNBUFFERED": "1",
+                "PYTHONPATH": "."
+            },
+            "args": [
+                "-A",
+                "onyx.background.celery.versioned_apps.indexing",
+                "worker",
+                "--pool=threads",
+                "--concurrency=1",
+                "--prefetch-multiplier=1",
+                "--loglevel=INFO",
+                "--hostname=indexing@%n",
+                "-Q",
+                "connector_indexing",
+            ],
+            "presentation": {
+				 "group": "2",
+			 },
+            "consoleTitle": "Celery indexing Console"
         },
-        "args": [
-          "-A",
-          "onyx.background.celery.versioned_apps.primary",
-          "worker",
-          "--pool=threads",
-          "--concurrency=4",
-          "--prefetch-multiplier=1",
-          "--loglevel=INFO",
-          "--hostname=primary@%n",
-          "-Q",
-          "celery"
-        ],
-        "presentation": {
-          "group": "2"
+        {
+            "name": "Celery monitoring",
+            "type": "debugpy",
+            "request": "launch",
+            "module": "celery",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {},
+            "args": [
+                "-A",
+                "onyx.background.celery.versioned_apps.monitoring",
+                "worker",
+                "--pool=solo",
+                "--concurrency=1",
+                "--prefetch-multiplier=1",
+                "--loglevel=INFO",
+                "--hostname=monitoring@%n",
+                "-Q",
+                "monitoring",
+            ],
+            "presentation": {
+				 "group": "2",
+			 },
+            "consoleTitle": "Celery monitoring Console"
         },
-        "consoleTitle": "Celery primary Console"
-      },
-      {
-        "name": "Celery light",
-        "type": "debugpy",
-        "request": "launch",
-        "module": "celery",
-        "cwd": "${workspaceFolder}/backend",
-        "envFile": "${workspaceFolder}/.vscode/.env",
-        "env": {
-          "LOG_LEVEL": "INFO",
-          "PYTHONUNBUFFERED": "1",
-          "PYTHONPATH": "."
+        {
+            "name": "Celery beat",
+            "type": "debugpy",
+            "request": "launch",
+            "module": "celery",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {
+                "LOG_LEVEL": "DEBUG",
+                "PYTHONUNBUFFERED": "1",
+                "PYTHONPATH": "."
+            },
+            "args": [
+                "-A",
+                "onyx.background.celery.versioned_apps.beat",
+                "beat",
+                "--loglevel=INFO",
+            ],
+            "presentation": {
+				 "group": "2",
+			 },
+            "consoleTitle": "Celery beat Console"
         },
-        "args": [
-          "-A",
-          "onyx.background.celery.versioned_apps.light",
-          "worker",
-          "--pool=threads",
-          "--concurrency=64",
-          "--prefetch-multiplier=8",
-          "--loglevel=INFO",
-          "--hostname=light@%n",
-          "-Q",
-          "vespa_metadata_sync,connector_deletion,doc_permissions_upsert"
-        ],
-        "presentation": {
-          "group": "2"
+        {
+            "name": "Pytest",
+            "consoleName": "Pytest",
+            "type": "debugpy",
+            "request": "launch",
+            "module": "pytest",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {
+                "LOG_LEVEL": "DEBUG",
+                "PYTHONUNBUFFERED": "1",
+                "PYTHONPATH": "."
+            },
+            "args": [
+                "-v"
+                // Specify a sepcific module/test to run or provide nothing to run all tests
+                //"tests/unit/onyx/llm/answering/test_prune_and_merge.py"
+            ],
+            "presentation": {
+				 "group": "2",
+			 },
+            "consoleTitle": "Pytest Console"
         },
-        "consoleTitle": "Celery light Console"
-      },
-      {
-        "name": "Celery heavy",
-        "type": "debugpy",
-        "request": "launch",
-        "module": "celery",
-        "cwd": "${workspaceFolder}/backend",
-        "envFile": "${workspaceFolder}/.vscode/.env",
-        "env": {
-          "LOG_LEVEL": "INFO",
-          "PYTHONUNBUFFERED": "1",
-          "PYTHONPATH": "."
+		{
+	    	// Dummy entry used to label the group
+			"name": "--- Tasks ---",
+			"type": "node",
+			"request": "launch",
+			"presentation": {
+				"group": "3",
+				"order": 0
+			}
+		},
+        {
+            "name": "Clear and Restart External Volumes and Containers",
+            "type": "node",
+            "request": "launch",
+            "runtimeExecutable": "bash",
+            "runtimeArgs": ["${workspaceFolder}/backend/scripts/restart_containers.sh"],
+            "cwd": "${workspaceFolder}",
+            "console": "integratedTerminal",
+            "stopOnEntry": true,
+            "presentation": {
+				 "group": "3",
+			 },
         },
-        "args": [
-          "-A",
-          "onyx.background.celery.versioned_apps.heavy",
-          "worker",
-          "--pool=threads",
-          "--concurrency=4",
-          "--prefetch-multiplier=1",
-          "--loglevel=INFO",
-          "--hostname=heavy@%n",
-          "-Q",
-          "connector_pruning,connector_doc_permissions_sync,connector_external_group_sync"
-        ],
-        "presentation": {
-          "group": "2"
+        {
+	        // Celery jobs launched through a single background script (legacy)
+	        // Recommend using the "Celery (all)" compound launch instead.
+            "name": "Background Jobs",
+            "consoleName": "Background Jobs",
+            "type": "debugpy",
+            "request": "launch",
+            "program": "scripts/dev_run_background_jobs.py",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {
+                "LOG_DANSWER_MODEL_INTERACTIONS": "True",
+                "LOG_LEVEL": "DEBUG",
+                "PYTHONUNBUFFERED": "1",
+                "PYTHONPATH": "."
+            },
         },
-        "consoleTitle": "Celery heavy Console"
-      },
-      {
-        "name": "Celery indexing",
-        "type": "debugpy",
-        "request": "launch",
-        "module": "celery",
-        "cwd": "${workspaceFolder}/backend",
-        "envFile": "${workspaceFolder}/.vscode/.env",
-        "env": {
-          "ENABLE_MULTIPASS_INDEXING": "false",
-          "LOG_LEVEL": "DEBUG",
-          "PYTHONUNBUFFERED": "1",
-          "PYTHONPATH": "."
+        {
+            "name": "Install Python Requirements",
+            "type": "node",
+            "request": "launch",
+            "runtimeExecutable": "bash",
+            "runtimeArgs": [
+                "-c",
+                "pip install -r backend/requirements/default.txt && pip install -r backend/requirements/dev.txt && pip install -r backend/requirements/ee.txt && pip install -r backend/requirements/model_server.txt"
+            ],
+            "cwd": "${workspaceFolder}",
+            "console": "integratedTerminal",
+            "presentation": {
+                 "group": "3"
+            }
         },
-        "args": [
-          "-A",
-          "onyx.background.celery.versioned_apps.indexing",
-          "worker",
-          "--pool=threads",
-          "--concurrency=1",
-          "--prefetch-multiplier=1",
-          "--loglevel=INFO",
-          "--hostname=indexing@%n",
-          "-Q",
-          "connector_indexing"
-        ],
-        "presentation": {
-          "group": "2"
-        },
-        "consoleTitle": "Celery indexing Console"
-      },
-      {
-        "name": "Celery monitoring",
-        "type": "debugpy",
-        "request": "launch",
-        "module": "celery",
-        "cwd": "${workspaceFolder}/backend",
-        "envFile": "${workspaceFolder}/.vscode/.env",
-        "env": {},
-        "args": [
-          "-A",
-          "onyx.background.celery.versioned_apps.monitoring",
-          "worker",
-          "--pool=solo",
-          "--concurrency=1",
-          "--prefetch-multiplier=1",
-          "--loglevel=INFO",
-          "--hostname=monitoring@%n",
-          "-Q",
-          "monitoring"
-        ],
-        "presentation": {
-          "group": "2"
-        },
-        "consoleTitle": "Celery monitoring Console"
-      },
-      {
-        "name": "Celery beat",
-        "type": "debugpy",
-        "request": "launch",
-        "module": "celery",
-        "cwd": "${workspaceFolder}/backend",
-        "envFile": "${workspaceFolder}/.vscode/.env",
-        "env": {
-          "LOG_LEVEL": "DEBUG",
-          "PYTHONUNBUFFERED": "1",
-          "PYTHONPATH": "."
-        },
-        "args": [
-          "-A",
-          "onyx.background.celery.versioned_apps.beat",
-          "beat",
-          "--loglevel=INFO"
-        ],
-        "presentation": {
-          "group": "2"
-        },
-        "consoleTitle": "Celery beat Console"
-      },
-      {
-        "name": "Celery user files indexing",
-        "type": "debugpy",
-        "request": "launch",
-        "module": "celery",
-        "cwd": "${workspaceFolder}/backend",
-        "envFile": "${workspaceFolder}/.vscode/.env",
-        "env": {
-          "LOG_LEVEL": "DEBUG",
-          "PYTHONUNBUFFERED": "1",
-          "PYTHONPATH": "."
-        },
-        "args": [
-          "-A",
-          "onyx.background.celery.versioned_apps.indexing",
-          "worker",
-          "--pool=threads",
-          "--concurrency=1",
-          "--prefetch-multiplier=1",
-          "--loglevel=INFO",
-          "--hostname=user_files_indexing@%n",
-          "-Q",
-          "user_files_indexing"
-        ],
-        "presentation": {
-          "group": "2"
-        },
-        "consoleTitle": "Celery user files indexing Console"
-      },
-      {
-        "name": "Pytest",
-        "consoleName": "Pytest",
-        "type": "debugpy",
-        "request": "launch",
-        "module": "pytest",
-        "cwd": "${workspaceFolder}/backend",
-        "envFile": "${workspaceFolder}/.vscode/.env",
-        "env": {
-          "LOG_LEVEL": "DEBUG",
-          "PYTHONUNBUFFERED": "1",
-          "PYTHONPATH": "."
-        },
-        "args": [
-          "-v"
-          // Specify a sepcific module/test to run or provide nothing to run all tests
-          //"tests/unit/onyx/llm/answering/test_prune_and_merge.py"
-        ],
-        "presentation": {
-          "group": "2"
-        },
-        "consoleTitle": "Pytest Console"
-      },
-      {
-        // Dummy entry used to label the group
-        "name": "--- Tasks ---",
-        "type": "node",
-        "request": "launch",
-        "presentation": {
-          "group": "3",
-          "order": 0
-        }
-      },
-      {
-        "name": "Clear and Restart External Volumes and Containers",
-        "type": "node",
-        "request": "launch",
-        "runtimeExecutable": "bash",
-        "runtimeArgs": [
-          "${workspaceFolder}/backend/scripts/restart_containers.sh"
-        ],
-        "cwd": "${workspaceFolder}",
-        "console": "integratedTerminal",
-        "stopOnEntry": true,
-        "presentation": {
-          "group": "3"
-        }
-      },
-      {
-        // Celery jobs launched through a single background script (legacy)
-        // Recommend using the "Celery (all)" compound launch instead.
-        "name": "Background Jobs",
-        "consoleName": "Background Jobs",
-        "type": "debugpy",
-        "request": "launch",
-        "program": "scripts/dev_run_background_jobs.py",
-        "cwd": "${workspaceFolder}/backend",
-        "envFile": "${workspaceFolder}/.vscode/.env",
-        "env": {
-          "LOG_DANSWER_MODEL_INTERACTIONS": "True",
-          "LOG_LEVEL": "DEBUG",
-          "PYTHONUNBUFFERED": "1",
-          "PYTHONPATH": "."
-        }
-      },
-      {
-        "name": "Install Python Requirements",
-        "type": "node",
-        "request": "launch",
-        "runtimeExecutable": "bash",
-        "runtimeArgs": [
-          "-c",
-          "pip install -r backend/requirements/default.txt && pip install -r backend/requirements/dev.txt && pip install -r backend/requirements/ee.txt && pip install -r backend/requirements/model_server.txt"
-        ],
-        "cwd": "${workspaceFolder}",
-        "console": "integratedTerminal",
-        "presentation": {
-          "group": "3"
-        }
-      },
-      {
-        "name": "Debug React Web App in Chrome",
-        "type": "chrome",
-        "request": "launch",
-        "url": "http://localhost:3000",
-        "webRoot": "${workspaceFolder}/web"
-      }
     ]
-  }
-  
+}

README.md

@@ -114,4 +114,3 @@ To try the Onyx Enterprise Edition:
 
 ## 💡 Contributing
 Looking to contribute? Please check out the [Contribution Guide](CONTRIBUTING.md) for more details.
-

backend/Dockerfile

@@ -8,7 +8,7 @@ Edition features outside of personal development or testing purposes. Please rea
 founders@onyx.app for more information. Please visit https://github.com/onyx-dot-app/onyx"
 
 # Default ONYX_VERSION, typically overriden during builds by GitHub Actions.
-ARG ONYX_VERSION=0.0.0-dev
+ARG ONYX_VERSION=0.8-dev
 # DO_NOT_TRACK is used to disable telemetry for Unstructured
 ENV ONYX_VERSION=${ONYX_VERSION} \
     DANSWER_RUNNING_IN_DOCKER="true" \
@@ -102,7 +102,6 @@ COPY ./alembic /app/alembic
 COPY ./alembic_tenants /app/alembic_tenants
 COPY ./alembic.ini /app/alembic.ini
 COPY supervisord.conf /usr/etc/supervisord.conf
-COPY ./static /app/static
 
 # Escape hatch scripts
 COPY ./scripts/debugging /app/scripts/debugging

backend/Dockerfile.model_server

@@ -7,7 +7,7 @@ You can find it at https://hub.docker.com/r/onyx/onyx-model-server. For more det
 visit https://github.com/onyx-dot-app/onyx."
 
 # Default ONYX_VERSION, typically overriden during builds by GitHub Actions.
-ARG ONYX_VERSION=0.0.0-dev
+ARG ONYX_VERSION=0.8-dev
 ENV ONYX_VERSION=${ONYX_VERSION} \
     DANSWER_RUNNING_IN_DOCKER="true"
 
@@ -31,8 +31,7 @@ RUN python -c "from transformers import AutoTokenizer; \
 AutoTokenizer.from_pretrained('distilbert-base-uncased'); \
 AutoTokenizer.from_pretrained('mixedbread-ai/mxbai-rerank-xsmall-v1'); \
 from huggingface_hub import snapshot_download; \
-snapshot_download(repo_id='onyx-dot-app/hybrid-intent-token-classifier'); \
-snapshot_download(repo_id='onyx-dot-app/information-content-model'); \
+snapshot_download(repo_id='danswer/hybrid-intent-token-classifier', revision='v1.0.3'); \
 snapshot_download('nomic-ai/nomic-embed-text-v1'); \
 snapshot_download('mixedbread-ai/mxbai-rerank-xsmall-v1'); \
 from sentence_transformers import SentenceTransformer; \
@@ -46,7 +45,6 @@ WORKDIR /app
 
 # Utils used by model server
 COPY ./onyx/utils/logger.py /app/onyx/utils/logger.py
-COPY ./onyx/utils/middleware.py /app/onyx/utils/middleware.py
 
 # Place to fetch version information
 COPY ./onyx/__init__.py /app/onyx/__init__.py

backend/alembic.ini

@@ -84,7 +84,7 @@ keys = console
 keys = generic
 
 [logger_root]
-level = INFO
+level = WARN
 handlers = console
 qualname =
 

backend/alembic/env.py

@@ -25,9 +25,6 @@ from shared_configs.configs import MULTI_TENANT, POSTGRES_DEFAULT_SCHEMA
 from onyx.db.models import Base
 from celery.backends.database.session import ResultModelBase  # type: ignore
 
-# Make sure in alembic.ini [logger_root] level=INFO is set or most logging will be
-# hidden! (defaults to level=WARN)
-
 # Alembic Config object
 config = context.config
 
@@ -39,7 +36,6 @@ if config.config_file_name is not None and config.attributes.get(
 target_metadata = [Base.metadata, ResultModelBase.metadata]
 
 EXCLUDE_TABLES = {"kombu_queue", "kombu_message"}
-
 logger = logging.getLogger(__name__)
 
 ssl_context: ssl.SSLContext | None = None
@@ -68,7 +64,7 @@ def include_object(
     return True
 
 
-def get_schema_options() -> tuple[str, bool, bool, bool]:
+def get_schema_options() -> tuple[str, bool, bool]:
     x_args_raw = context.get_x_argument()
     x_args = {}
     for arg in x_args_raw:
@@ -80,10 +76,6 @@ def get_schema_options() -> tuple[str, bool, bool, bool]:
     create_schema = x_args.get("create_schema", "true").lower() == "true"
     upgrade_all_tenants = x_args.get("upgrade_all_tenants", "false").lower() == "true"
 
-    # continue on error with individual tenant
-    # only applies to online migrations
-    continue_on_error = x_args.get("continue", "false").lower() == "true"
-
     if (
         MULTI_TENANT
         and schema_name == POSTGRES_DEFAULT_SCHEMA
@@ -94,12 +86,14 @@ def get_schema_options() -> tuple[str, bool, bool, bool]:
             "Please specify a tenant-specific schema."
         )
 
-    return schema_name, create_schema, upgrade_all_tenants, continue_on_error
+    return schema_name, create_schema, upgrade_all_tenants
 
 
 def do_run_migrations(
     connection: Connection, schema_name: str, create_schema: bool
 ) -> None:
+    logger.info(f"About to migrate schema: {schema_name}")
+
     if create_schema:
         connection.execute(text(f'CREATE SCHEMA IF NOT EXISTS "{schema_name}"'))
         connection.execute(text("COMMIT"))
@@ -140,12 +134,7 @@ def provide_iam_token_for_alembic(
 
 
 async def run_async_migrations() -> None:
-    (
-        schema_name,
-        create_schema,
-        upgrade_all_tenants,
-        continue_on_error,
-    ) = get_schema_options()
+    schema_name, create_schema, upgrade_all_tenants = get_schema_options()
 
     engine = create_async_engine(
         build_connection_string(),
@@ -162,15 +151,9 @@ async def run_async_migrations() -> None:
 
     if upgrade_all_tenants:
         tenant_schemas = get_all_tenant_ids()
-
-        i_tenant = 0
-        num_tenants = len(tenant_schemas)
         for schema in tenant_schemas:
-            i_tenant += 1
-            logger.info(
-                f"Migrating schema: index={i_tenant} num_tenants={num_tenants} schema={schema}"
-            )
             try:
+                logger.info(f"Migrating schema: {schema}")
                 async with engine.connect() as connection:
                     await connection.run_sync(
                         do_run_migrations,
@@ -179,12 +162,7 @@ async def run_async_migrations() -> None:
                     )
             except Exception as e:
                 logger.error(f"Error migrating schema {schema}: {e}")
-                if not continue_on_error:
-                    logger.error("--continue is not set, raising exception!")
-                    raise
-
-                logger.warning("--continue is set, continuing to next schema.")
-
+                raise
     else:
         try:
             logger.info(f"Migrating schema: {schema_name}")
@@ -202,11 +180,7 @@ async def run_async_migrations() -> None:
 
 
 def run_migrations_offline() -> None:
-    """This doesn't really get used when we migrate in the cloud."""
-
-    logger.info("run_migrations_offline starting.")
-
-    schema_name, _, upgrade_all_tenants, continue_on_error = get_schema_options()
+    schema_name, _, upgrade_all_tenants = get_schema_options()
     url = build_connection_string()
 
     if upgrade_all_tenants:
@@ -256,7 +230,6 @@ def run_migrations_offline() -> None:
 
 
 def run_migrations_online() -> None:
-    logger.info("run_migrations_online starting.")
     asyncio.run(run_async_migrations())
 
 

backend/alembic/versions/3781a5eb12cb_add_chunk_stats_table.py

@@ -1,51 +0,0 @@
-"""add chunk stats table
-
-Revision ID: 3781a5eb12cb
-Revises: df46c75b714e
-Create Date: 2025-03-10 10:02:30.586666
-
-"""
-from alembic import op
-import sqlalchemy as sa
-
-# revision identifiers, used by Alembic.
-revision = "3781a5eb12cb"
-down_revision = "df46c75b714e"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    op.create_table(
-        "chunk_stats",
-        sa.Column("id", sa.String(), primary_key=True, index=True),
-        sa.Column(
-            "document_id",
-            sa.String(),
-            sa.ForeignKey("document.id"),
-            nullable=False,
-            index=True,
-        ),
-        sa.Column("chunk_in_doc_id", sa.Integer(), nullable=False),
-        sa.Column("information_content_boost", sa.Float(), nullable=True),
-        sa.Column(
-            "last_modified",
-            sa.DateTime(timezone=True),
-            nullable=False,
-            index=True,
-            server_default=sa.func.now(),
-        ),
-        sa.Column("last_synced", sa.DateTime(timezone=True), nullable=True, index=True),
-        sa.UniqueConstraint(
-            "document_id", "chunk_in_doc_id", name="uq_chunk_stats_doc_chunk"
-        ),
-    )
-
-    op.create_index(
-        "ix_chunk_sync_status", "chunk_stats", ["last_modified", "last_synced"]
-    )
-
-
-def downgrade() -> None:
-    op.drop_index("ix_chunk_sync_status", table_name="chunk_stats")
-    op.drop_table("chunk_stats")

backend/alembic/versions/3934b1bc7b62_update_github_connector_repo_name_to_.py

@@ -1,125 +0,0 @@
-"""Update GitHub connector repo_name to repositories
-
-Revision ID: 3934b1bc7b62
-Revises: b7c2b63c4a03
-Create Date: 2025-03-05 10:50:30.516962
-
-"""
-from alembic import op
-import sqlalchemy as sa
-import json
-import logging
-
-# revision identifiers, used by Alembic.
-revision = "3934b1bc7b62"
-down_revision = "b7c2b63c4a03"
-branch_labels = None
-depends_on = None
-
-logger = logging.getLogger("alembic.runtime.migration")
-
-
-def upgrade() -> None:
-    # Get all GitHub connectors
-    conn = op.get_bind()
-
-    # First get all GitHub connectors
-    github_connectors = conn.execute(
-        sa.text(
-            """
-            SELECT id, connector_specific_config
-            FROM connector
-            WHERE source = 'GITHUB'
-            """
-        )
-    ).fetchall()
-
-    # Update each connector's config
-    updated_count = 0
-    for connector_id, config in github_connectors:
-        try:
-            if not config:
-                logger.warning(f"Connector {connector_id} has no config, skipping")
-                continue
-
-            # Parse the config if it's a string
-            if isinstance(config, str):
-                config = json.loads(config)
-
-            if "repo_name" not in config:
-                continue
-
-            # Create new config with repositories instead of repo_name
-            new_config = dict(config)
-            repo_name_value = new_config.pop("repo_name")
-            new_config["repositories"] = repo_name_value
-
-            # Update the connector with the new config
-            conn.execute(
-                sa.text(
-                    """
-                    UPDATE connector
-                    SET connector_specific_config = :new_config
-                    WHERE id = :connector_id
-                    """
-                ),
-                {"connector_id": connector_id, "new_config": json.dumps(new_config)},
-            )
-            updated_count += 1
-        except Exception as e:
-            logger.error(f"Error updating connector {connector_id}: {str(e)}")
-
-
-def downgrade() -> None:
-    # Get all GitHub connectors
-    conn = op.get_bind()
-
-    logger.debug(
-        "Starting rollback of GitHub connectors from repositories to repo_name"
-    )
-
-    github_connectors = conn.execute(
-        sa.text(
-            """
-            SELECT id, connector_specific_config
-            FROM connector
-            WHERE source = 'GITHUB'
-            """
-        )
-    ).fetchall()
-
-    logger.debug(f"Found {len(github_connectors)} GitHub connectors to rollback")
-
-    # Revert each GitHub connector to use repo_name instead of repositories
-    reverted_count = 0
-    for connector_id, config in github_connectors:
-        try:
-            if not config:
-                continue
-
-            # Parse the config if it's a string
-            if isinstance(config, str):
-                config = json.loads(config)
-
-            if "repositories" not in config:
-                continue
-
-            # Create new config with repo_name instead of repositories
-            new_config = dict(config)
-            repositories_value = new_config.pop("repositories")
-            new_config["repo_name"] = repositories_value
-
-            # Update the connector with the new config
-            conn.execute(
-                sa.text(
-                    """
-                    UPDATE connector
-                    SET connector_specific_config = :new_config
-                    WHERE id = :connector_id
-                    """
-                ),
-                {"new_config": json.dumps(new_config), "connector_id": connector_id},
-            )
-            reverted_count += 1
-        except Exception as e:
-            logger.error(f"Error reverting connector {connector_id}: {str(e)}")

backend/alembic/versions/3bd4c84fe72f_improved_index.py

@@ -1,98 +0,0 @@
-"""improved index
-
-Revision ID: 3bd4c84fe72f
-Revises: 8f43500ee275
-Create Date: 2025-02-26 13:07:56.217791
-
-"""
-from alembic import op
-
-
-# revision identifiers, used by Alembic.
-revision = "3bd4c84fe72f"
-down_revision = "8f43500ee275"
-branch_labels = None
-depends_on = None
-
-
-# NOTE:
-# This migration addresses issues with the previous migration (8f43500ee275) which caused
-# an outage by creating an index without using CONCURRENTLY. This migration:
-#
-# 1. Creates more efficient full-text search capabilities using tsvector columns and GIN indexes
-# 2. Uses CONCURRENTLY for all index creation to prevent table locking
-# 3. Explicitly manages transactions with COMMIT statements to allow CONCURRENTLY to work
-# (see: https://www.postgresql.org/docs/9.4/sql-createindex.html#SQL-CREATEINDEX-CONCURRENTLY)
-# (see: https://github.com/sqlalchemy/alembic/issues/277)
-# 4. Adds indexes to both chat_message and chat_session tables for comprehensive search
-
-
-def upgrade() -> None:
-    # First, drop any existing indexes to avoid conflicts
-    op.execute("COMMIT")
-    op.execute("DROP INDEX CONCURRENTLY IF EXISTS idx_chat_message_tsv;")
-
-    op.execute("COMMIT")
-    op.execute("DROP INDEX CONCURRENTLY IF EXISTS idx_chat_session_desc_tsv;")
-
-    op.execute("COMMIT")
-    op.execute("DROP INDEX IF EXISTS idx_chat_message_message_lower;")
-
-    # Drop existing columns if they exist
-    op.execute("ALTER TABLE chat_message DROP COLUMN IF EXISTS message_tsv;")
-    op.execute("ALTER TABLE chat_session DROP COLUMN IF EXISTS description_tsv;")
-
-    # Create a GIN index for full-text search on chat_message.message
-    op.execute(
-        """
-        ALTER TABLE chat_message
-        ADD COLUMN message_tsv tsvector
-        GENERATED ALWAYS AS (to_tsvector('english', message)) STORED;
-        """
-    )
-
-    # Commit the current transaction before creating concurrent indexes
-    op.execute("COMMIT")
-
-    op.execute(
-        """
-        CREATE INDEX CONCURRENTLY IF NOT EXISTS idx_chat_message_tsv
-        ON chat_message
-        USING GIN (message_tsv)
-        """
-    )
-
-    # Also add a stored tsvector column for chat_session.description
-    op.execute(
-        """
-        ALTER TABLE chat_session
-        ADD COLUMN description_tsv tsvector
-        GENERATED ALWAYS AS (to_tsvector('english', coalesce(description, ''))) STORED;
-        """
-    )
-
-    # Commit again before creating the second concurrent index
-    op.execute("COMMIT")
-
-    op.execute(
-        """
-        CREATE INDEX CONCURRENTLY IF NOT EXISTS idx_chat_session_desc_tsv
-        ON chat_session
-        USING GIN (description_tsv)
-        """
-    )
-
-
-def downgrade() -> None:
-    # Drop the indexes first (use CONCURRENTLY for dropping too)
-    op.execute("COMMIT")
-    op.execute("DROP INDEX CONCURRENTLY IF EXISTS idx_chat_message_tsv;")
-
-    op.execute("COMMIT")
-    op.execute("DROP INDEX CONCURRENTLY IF EXISTS idx_chat_session_desc_tsv;")
-
-    # Then drop the columns
-    op.execute("ALTER TABLE chat_message DROP COLUMN IF EXISTS message_tsv;")
-    op.execute("ALTER TABLE chat_session DROP COLUMN IF EXISTS description_tsv;")
-
-    op.execute("DROP INDEX IF EXISTS idx_chat_message_message_lower;")

backend/alembic/versions/4794bc13e484_update_prompt_length.py

@@ -1,50 +0,0 @@
-"""update prompt length
-
-Revision ID: 4794bc13e484
-Revises: f7505c5b0284
-Create Date: 2025-04-02 11:26:36.180328
-
-"""
-from alembic import op
-import sqlalchemy as sa
-
-
-# revision identifiers, used by Alembic.
-revision = "4794bc13e484"
-down_revision = "f7505c5b0284"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    op.alter_column(
-        "prompt",
-        "system_prompt",
-        existing_type=sa.TEXT(),
-        type_=sa.String(length=5000000),
-        existing_nullable=False,
-    )
-    op.alter_column(
-        "prompt",
-        "task_prompt",
-        existing_type=sa.TEXT(),
-        type_=sa.String(length=5000000),
-        existing_nullable=False,
-    )
-
-
-def downgrade() -> None:
-    op.alter_column(
-        "prompt",
-        "system_prompt",
-        existing_type=sa.String(length=5000000),
-        type_=sa.TEXT(),
-        existing_nullable=False,
-    )
-    op.alter_column(
-        "prompt",
-        "task_prompt",
-        existing_type=sa.String(length=5000000),
-        type_=sa.TEXT(),
-        existing_nullable=False,
-    )

backend/alembic/versions/4d58345da04a_lowercase_user_emails.py

@@ -5,10 +5,7 @@ Revises: f1ca58b2f2ec
 Create Date: 2025-01-29 07:48:46.784041
 
 """
-import logging
-from typing import cast
 from alembic import op
-from sqlalchemy.exc import IntegrityError
 from sqlalchemy.sql import text
 
 
@@ -18,45 +15,21 @@ down_revision = "f1ca58b2f2ec"
 branch_labels = None
 depends_on = None
 
-logger = logging.getLogger("alembic.runtime.migration")
-
 
 def upgrade() -> None:
-    """Conflicts on lowercasing will result in the uppercased email getting a
-    unique integer suffix when converted to lowercase."""
-
+    # Get database connection
     connection = op.get_bind()
 
-    # Fetch all user emails that are not already lowercase
-    user_emails = connection.execute(
-        text('SELECT id, email FROM "user" WHERE email != LOWER(email)')
-    ).fetchall()
-
-    for user_id, email in user_emails:
-        email = cast(str, email)
-        username, domain = email.rsplit("@", 1)
-        new_email = f"{username.lower()}@{domain.lower()}"
-        attempt = 1
-
-        while True:
-            try:
-                # Try updating the email
-                connection.execute(
-                    text('UPDATE "user" SET email = :new_email WHERE id = :user_id'),
-                    {"new_email": new_email, "user_id": user_id},
-                )
-                break  # Success, exit loop
-            except IntegrityError:
-                next_email = f"{username.lower()}_{attempt}@{domain.lower()}"
-                # Email conflict occurred, append `_1`, `_2`, etc., to the username
-                logger.warning(
-                    f"Conflict while lowercasing email: "
-                    f"old_email={email} "
-                    f"conflicting_email={new_email} "
-                    f"next_email={next_email}"
-                )
-                new_email = next_email
-                attempt += 1
+    # Update all user emails to lowercase
+    connection.execute(
+        text(
+            """
+            UPDATE "user"
+            SET email = LOWER(email)
+            WHERE email != LOWER(email)
+            """
+        )
+    )
 
 
 def downgrade() -> None:

backend/alembic/versions/6a804aeb4830_duplicated_no_harm_user_file_migration.py

@@ -1,117 +0,0 @@
-"""duplicated no-harm user file migration
-
-Revision ID: 6a804aeb4830
-Revises: 8e1ac4f39a9f
-Create Date: 2025-04-01 07:26:10.539362
-
-"""
-from alembic import op
-import sqlalchemy as sa
-from sqlalchemy import inspect
-import datetime
-
-
-# revision identifiers, used by Alembic.
-revision = "6a804aeb4830"
-down_revision = "8e1ac4f39a9f"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    # Check if user_file table already exists
-    conn = op.get_bind()
-    inspector = inspect(conn)
-
-    if not inspector.has_table("user_file"):
-        # Create user_folder table without parent_id
-        op.create_table(
-            "user_folder",
-            sa.Column("id", sa.Integer(), primary_key=True, autoincrement=True),
-            sa.Column("user_id", sa.UUID(), sa.ForeignKey("user.id"), nullable=True),
-            sa.Column("name", sa.String(length=255), nullable=True),
-            sa.Column("description", sa.String(length=255), nullable=True),
-            sa.Column("display_priority", sa.Integer(), nullable=True, default=0),
-            sa.Column(
-                "created_at", sa.DateTime(timezone=True), server_default=sa.func.now()
-            ),
-        )
-
-        # Create user_file table with folder_id instead of parent_folder_id
-        op.create_table(
-            "user_file",
-            sa.Column("id", sa.Integer(), primary_key=True, autoincrement=True),
-            sa.Column("user_id", sa.UUID(), sa.ForeignKey("user.id"), nullable=True),
-            sa.Column(
-                "folder_id",
-                sa.Integer(),
-                sa.ForeignKey("user_folder.id"),
-                nullable=True,
-            ),
-            sa.Column("link_url", sa.String(), nullable=True),
-            sa.Column("token_count", sa.Integer(), nullable=True),
-            sa.Column("file_type", sa.String(), nullable=True),
-            sa.Column("file_id", sa.String(length=255), nullable=False),
-            sa.Column("document_id", sa.String(length=255), nullable=False),
-            sa.Column("name", sa.String(length=255), nullable=False),
-            sa.Column(
-                "created_at",
-                sa.DateTime(),
-                default=datetime.datetime.utcnow,
-            ),
-            sa.Column(
-                "cc_pair_id",
-                sa.Integer(),
-                sa.ForeignKey("connector_credential_pair.id"),
-                nullable=True,
-                unique=True,
-            ),
-        )
-
-        # Create persona__user_file table
-        op.create_table(
-            "persona__user_file",
-            sa.Column(
-                "persona_id",
-                sa.Integer(),
-                sa.ForeignKey("persona.id"),
-                primary_key=True,
-            ),
-            sa.Column(
-                "user_file_id",
-                sa.Integer(),
-                sa.ForeignKey("user_file.id"),
-                primary_key=True,
-            ),
-        )
-
-        # Create persona__user_folder table
-        op.create_table(
-            "persona__user_folder",
-            sa.Column(
-                "persona_id",
-                sa.Integer(),
-                sa.ForeignKey("persona.id"),
-                primary_key=True,
-            ),
-            sa.Column(
-                "user_folder_id",
-                sa.Integer(),
-                sa.ForeignKey("user_folder.id"),
-                primary_key=True,
-            ),
-        )
-
-        op.add_column(
-            "connector_credential_pair",
-            sa.Column("is_user_file", sa.Boolean(), nullable=True, default=False),
-        )
-
-        # Update existing records to have is_user_file=False instead of NULL
-        op.execute(
-            "UPDATE connector_credential_pair SET is_user_file = FALSE WHERE is_user_file IS NULL"
-        )
-
-
-def downgrade() -> None:
-    pass

backend/alembic/versions/8e1ac4f39a9f_enable_contextual_retrieval.py

@@ -1,50 +0,0 @@
-"""enable contextual retrieval
-
-Revision ID: 8e1ac4f39a9f
-Revises: 9aadf32dfeb4
-Create Date: 2024-12-20 13:29:09.918661
-
-"""
-from alembic import op
-import sqlalchemy as sa
-
-
-# revision identifiers, used by Alembic.
-revision = "8e1ac4f39a9f"
-down_revision = "9aadf32dfeb4"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    op.add_column(
-        "search_settings",
-        sa.Column(
-            "enable_contextual_rag",
-            sa.Boolean(),
-            nullable=False,
-            server_default="false",
-        ),
-    )
-    op.add_column(
-        "search_settings",
-        sa.Column(
-            "contextual_rag_llm_name",
-            sa.String(),
-            nullable=True,
-        ),
-    )
-    op.add_column(
-        "search_settings",
-        sa.Column(
-            "contextual_rag_llm_provider",
-            sa.String(),
-            nullable=True,
-        ),
-    )
-
-
-def downgrade() -> None:
-    op.drop_column("search_settings", "enable_contextual_rag")
-    op.drop_column("search_settings", "contextual_rag_llm_name")
-    op.drop_column("search_settings", "contextual_rag_llm_provider")

backend/alembic/versions/8f43500ee275_add_index.py

@@ -18,13 +18,12 @@ depends_on = None
 def upgrade() -> None:
     # Create a basic index on the lowercase message column for direct text matching
     # Limit to 1500 characters to stay well under the 2856 byte limit of btree version 4
-    # op.execute(
-    #     """
-    #     CREATE INDEX idx_chat_message_message_lower
-    #     ON chat_message (LOWER(substring(message, 1, 1500)))
-    #     """
-    # )
-    pass
+    op.execute(
+        """
+        CREATE INDEX idx_chat_message_message_lower
+        ON chat_message (LOWER(substring(message, 1, 1500)))
+        """
+    )
 
 
 def downgrade() -> None:

backend/alembic/versions/9aadf32dfeb4_add_user_files.py

@@ -1,113 +0,0 @@
-"""add user files
-
-Revision ID: 9aadf32dfeb4
-Revises: 3781a5eb12cb
-Create Date: 2025-01-26 16:08:21.551022
-
-"""
-import sqlalchemy as sa
-import datetime
-from alembic import op
-
-
-# revision identifiers, used by Alembic.
-revision = "9aadf32dfeb4"
-down_revision = "3781a5eb12cb"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    # Create user_folder table without parent_id
-    op.create_table(
-        "user_folder",
-        sa.Column("id", sa.Integer(), primary_key=True, autoincrement=True),
-        sa.Column("user_id", sa.UUID(), sa.ForeignKey("user.id"), nullable=True),
-        sa.Column("name", sa.String(length=255), nullable=True),
-        sa.Column("description", sa.String(length=255), nullable=True),
-        sa.Column("display_priority", sa.Integer(), nullable=True, default=0),
-        sa.Column(
-            "created_at", sa.DateTime(timezone=True), server_default=sa.func.now()
-        ),
-    )
-
-    # Create user_file table with folder_id instead of parent_folder_id
-    op.create_table(
-        "user_file",
-        sa.Column("id", sa.Integer(), primary_key=True, autoincrement=True),
-        sa.Column("user_id", sa.UUID(), sa.ForeignKey("user.id"), nullable=True),
-        sa.Column(
-            "folder_id",
-            sa.Integer(),
-            sa.ForeignKey("user_folder.id"),
-            nullable=True,
-        ),
-        sa.Column("link_url", sa.String(), nullable=True),
-        sa.Column("token_count", sa.Integer(), nullable=True),
-        sa.Column("file_type", sa.String(), nullable=True),
-        sa.Column("file_id", sa.String(length=255), nullable=False),
-        sa.Column("document_id", sa.String(length=255), nullable=False),
-        sa.Column("name", sa.String(length=255), nullable=False),
-        sa.Column(
-            "created_at",
-            sa.DateTime(),
-            default=datetime.datetime.utcnow,
-        ),
-        sa.Column(
-            "cc_pair_id",
-            sa.Integer(),
-            sa.ForeignKey("connector_credential_pair.id"),
-            nullable=True,
-            unique=True,
-        ),
-    )
-
-    # Create persona__user_file table
-    op.create_table(
-        "persona__user_file",
-        sa.Column(
-            "persona_id", sa.Integer(), sa.ForeignKey("persona.id"), primary_key=True
-        ),
-        sa.Column(
-            "user_file_id",
-            sa.Integer(),
-            sa.ForeignKey("user_file.id"),
-            primary_key=True,
-        ),
-    )
-
-    # Create persona__user_folder table
-    op.create_table(
-        "persona__user_folder",
-        sa.Column(
-            "persona_id", sa.Integer(), sa.ForeignKey("persona.id"), primary_key=True
-        ),
-        sa.Column(
-            "user_folder_id",
-            sa.Integer(),
-            sa.ForeignKey("user_folder.id"),
-            primary_key=True,
-        ),
-    )
-
-    op.add_column(
-        "connector_credential_pair",
-        sa.Column("is_user_file", sa.Boolean(), nullable=True, default=False),
-    )
-
-    # Update existing records to have is_user_file=False instead of NULL
-    op.execute(
-        "UPDATE connector_credential_pair SET is_user_file = FALSE WHERE is_user_file IS NULL"
-    )
-
-
-def downgrade() -> None:
-    # Drop the persona__user_folder table
-    op.drop_table("persona__user_folder")
-    # Drop the persona__user_file table
-    op.drop_table("persona__user_file")
-    # Drop the user_file table
-    op.drop_table("user_file")
-    # Drop the user_folder table
-    op.drop_table("user_folder")
-    op.drop_column("connector_credential_pair", "is_user_file")

backend/alembic/versions/b7c2b63c4a03_add_background_reindex_enabled_field.py

@@ -1,55 +0,0 @@
-"""add background_reindex_enabled field
-
-Revision ID: b7c2b63c4a03
-Revises: f11b408e39d3
-Create Date: 2024-03-26 12:34:56.789012
-
-"""
-from alembic import op
-import sqlalchemy as sa
-
-from onyx.db.enums import EmbeddingPrecision
-
-
-# revision identifiers, used by Alembic.
-revision = "b7c2b63c4a03"
-down_revision = "f11b408e39d3"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    # Add background_reindex_enabled column with default value of True
-    op.add_column(
-        "search_settings",
-        sa.Column(
-            "background_reindex_enabled",
-            sa.Boolean(),
-            nullable=False,
-            server_default="true",
-        ),
-    )
-
-    # Add embedding_precision column with default value of FLOAT
-    op.add_column(
-        "search_settings",
-        sa.Column(
-            "embedding_precision",
-            sa.Enum(EmbeddingPrecision, native_enum=False),
-            nullable=False,
-            server_default=EmbeddingPrecision.FLOAT.name,
-        ),
-    )
-
-    # Add reduced_dimension column with default value of None
-    op.add_column(
-        "search_settings",
-        sa.Column("reduced_dimension", sa.Integer(), nullable=True),
-    )
-
-
-def downgrade() -> None:
-    # Remove the background_reindex_enabled column
-    op.drop_column("search_settings", "background_reindex_enabled")
-    op.drop_column("search_settings", "embedding_precision")
-    op.drop_column("search_settings", "reduced_dimension")

backend/alembic/versions/df46c75b714e_add_default_vision_provider_to_llm_.py

@@ -1,36 +0,0 @@
-"""add_default_vision_provider_to_llm_provider
-
-Revision ID: df46c75b714e
-Revises: 3934b1bc7b62
-Create Date: 2025-03-11 16:20:19.038945
-
-"""
-from alembic import op
-import sqlalchemy as sa
-
-
-# revision identifiers, used by Alembic.
-revision = "df46c75b714e"
-down_revision = "3934b1bc7b62"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    op.add_column(
-        "llm_provider",
-        sa.Column(
-            "is_default_vision_provider",
-            sa.Boolean(),
-            nullable=True,
-            server_default=sa.false(),
-        ),
-    )
-    op.add_column(
-        "llm_provider", sa.Column("default_vision_model", sa.String(), nullable=True)
-    )
-
-
-def downgrade() -> None:
-    op.drop_column("llm_provider", "default_vision_model")
-    op.drop_column("llm_provider", "is_default_vision_provider")

backend/alembic/versions/f11b408e39d3_force_lowercase_all_users.py

@@ -1,36 +0,0 @@
-"""force lowercase all users
-
-Revision ID: f11b408e39d3
-Revises: 3bd4c84fe72f
-Create Date: 2025-02-26 17:04:55.683500
-
-"""
-
-
-# revision identifiers, used by Alembic.
-revision = "f11b408e39d3"
-down_revision = "3bd4c84fe72f"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    # 1) Convert all existing user emails to lowercase
-    from alembic import op
-
-    op.execute(
-        """
-        UPDATE "user"
-        SET email = LOWER(email)
-        """
-    )
-
-    # 2) Add a check constraint to ensure emails are always lowercase
-    op.create_check_constraint("ensure_lowercase_email", "user", "email = LOWER(email)")
-
-
-def downgrade() -> None:
-    # Drop the check constraint
-    from alembic import op
-
-    op.drop_constraint("ensure_lowercase_email", "user", type_="check")

backend/alembic/versions/f71470ba9274_add_prompt_length_limit.py

@@ -1,50 +0,0 @@
-"""add prompt length limit
-
-Revision ID: f71470ba9274
-Revises: 6a804aeb4830
-Create Date: 2025-04-01 15:07:14.977435
-
-"""
-
-
-# revision identifiers, used by Alembic.
-revision = "f71470ba9274"
-down_revision = "6a804aeb4830"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    # op.alter_column(
-    #     "prompt",
-    #     "system_prompt",
-    #     existing_type=sa.TEXT(),
-    #     type_=sa.String(length=8000),
-    #     existing_nullable=False,
-    # )
-    # op.alter_column(
-    #     "prompt",
-    #     "task_prompt",
-    #     existing_type=sa.TEXT(),
-    #     type_=sa.String(length=8000),
-    #     existing_nullable=False,
-    # )
-    pass
-
-
-def downgrade() -> None:
-    # op.alter_column(
-    #     "prompt",
-    #     "system_prompt",
-    #     existing_type=sa.String(length=8000),
-    #     type_=sa.TEXT(),
-    #     existing_nullable=False,
-    # )
-    # op.alter_column(
-    #     "prompt",
-    #     "task_prompt",
-    #     existing_type=sa.String(length=8000),
-    #     type_=sa.TEXT(),
-    #     existing_nullable=False,
-    # )
-    pass

backend/alembic/versions/f7505c5b0284_updated_constraints_for_ccpairs.py

@@ -1,77 +0,0 @@
-"""updated constraints for ccpairs
-
-Revision ID: f7505c5b0284
-Revises: f71470ba9274
-Create Date: 2025-04-01 17:50:42.504818
-
-"""
-from alembic import op
-
-
-# revision identifiers, used by Alembic.
-revision = "f7505c5b0284"
-down_revision = "f71470ba9274"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    # 1) Drop the old foreign-key constraints
-    op.drop_constraint(
-        "document_by_connector_credential_pair_connector_id_fkey",
-        "document_by_connector_credential_pair",
-        type_="foreignkey",
-    )
-    op.drop_constraint(
-        "document_by_connector_credential_pair_credential_id_fkey",
-        "document_by_connector_credential_pair",
-        type_="foreignkey",
-    )
-
-    # 2) Re-add them with ondelete='CASCADE'
-    op.create_foreign_key(
-        "document_by_connector_credential_pair_connector_id_fkey",
-        source_table="document_by_connector_credential_pair",
-        referent_table="connector",
-        local_cols=["connector_id"],
-        remote_cols=["id"],
-        ondelete="CASCADE",
-    )
-    op.create_foreign_key(
-        "document_by_connector_credential_pair_credential_id_fkey",
-        source_table="document_by_connector_credential_pair",
-        referent_table="credential",
-        local_cols=["credential_id"],
-        remote_cols=["id"],
-        ondelete="CASCADE",
-    )
-
-
-def downgrade() -> None:
-    # Reverse the changes for rollback
-    op.drop_constraint(
-        "document_by_connector_credential_pair_connector_id_fkey",
-        "document_by_connector_credential_pair",
-        type_="foreignkey",
-    )
-    op.drop_constraint(
-        "document_by_connector_credential_pair_credential_id_fkey",
-        "document_by_connector_credential_pair",
-        type_="foreignkey",
-    )
-
-    # Recreate without CASCADE
-    op.create_foreign_key(
-        "document_by_connector_credential_pair_connector_id_fkey",
-        "document_by_connector_credential_pair",
-        "connector",
-        ["connector_id"],
-        ["id"],
-    )
-    op.create_foreign_key(
-        "document_by_connector_credential_pair_credential_id_fkey",
-        "document_by_connector_credential_pair",
-        "credential",
-        ["credential_id"],
-        ["id"],
-    )

backend/alembic_tenants/versions/34e3630c7f32_lowercase_multi_tenant_user_auth.py

@@ -1,42 +0,0 @@
-"""lowercase multi-tenant user auth
-
-Revision ID: 34e3630c7f32
-Revises: a4f6ee863c47
-Create Date: 2025-02-26 15:03:01.211894
-
-"""
-from alembic import op
-
-
-# revision identifiers, used by Alembic.
-revision = "34e3630c7f32"
-down_revision = "a4f6ee863c47"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    # 1) Convert all existing rows to lowercase
-    op.execute(
-        """
-        UPDATE user_tenant_mapping
-        SET email = LOWER(email)
-        """
-    )
-    # 2) Add a check constraint so that emails cannot be written in uppercase
-    op.create_check_constraint(
-        "ensure_lowercase_email",
-        "user_tenant_mapping",
-        "email = LOWER(email)",
-        schema="public",
-    )
-
-
-def downgrade() -> None:
-    # Drop the check constraint
-    op.drop_constraint(
-        "ensure_lowercase_email",
-        "user_tenant_mapping",
-        schema="public",
-        type_="check",
-    )

backend/alembic_tenants/versions/3b45e0018bf1_add_new_available_tenant_table.py

@@ -1,33 +0,0 @@
-"""add new available tenant table
-
-Revision ID: 3b45e0018bf1
-Revises: ac842f85f932
-Create Date: 2025-03-06 09:55:18.229910
-
-"""
-import sqlalchemy as sa
-
-from alembic import op
-
-
-# revision identifiers, used by Alembic.
-revision = "3b45e0018bf1"
-down_revision = "ac842f85f932"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    # Create new_available_tenant table
-    op.create_table(
-        "available_tenant",
-        sa.Column("tenant_id", sa.String(), nullable=False),
-        sa.Column("alembic_version", sa.String(), nullable=False),
-        sa.Column("date_created", sa.DateTime(), nullable=False),
-        sa.PrimaryKeyConstraint("tenant_id"),
-    )
-
-
-def downgrade() -> None:
-    # Drop new_available_tenant table
-    op.drop_table("available_tenant")

backend/alembic_tenants/versions/ac842f85f932_new_column_user_tenant_mapping.py

@@ -1,51 +0,0 @@
-"""new column user tenant mapping
-
-Revision ID: ac842f85f932
-Revises: 34e3630c7f32
-Create Date: 2025-03-03 13:30:14.802874
-
-"""
-import sqlalchemy as sa
-
-from alembic import op
-
-
-# revision identifiers, used by Alembic.
-revision = "ac842f85f932"
-down_revision = "34e3630c7f32"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    # Add active column with default value of True
-    op.add_column(
-        "user_tenant_mapping",
-        sa.Column(
-            "active",
-            sa.Boolean(),
-            nullable=False,
-            server_default="true",
-        ),
-        schema="public",
-    )
-
-    op.drop_constraint("uq_email", "user_tenant_mapping", schema="public")
-
-    # Create a unique index for active=true records
-    # This ensures a user can only be active in one tenant at a time
-    op.execute(
-        "CREATE UNIQUE INDEX uq_user_active_email_idx ON public.user_tenant_mapping (email) WHERE active = true"
-    )
-
-
-def downgrade() -> None:
-    # Drop the unique index for active=true records
-    op.execute("DROP INDEX IF EXISTS uq_user_active_email_idx")
-
-    op.create_unique_constraint(
-        "uq_email", "user_tenant_mapping", ["email"], schema="public"
-    )
-
-    # Remove the active column
-    op.drop_column("user_tenant_mapping", "active", schema="public")

backend/ee/onyx/access/access.py

@@ -93,12 +93,12 @@ def _get_access_for_documents(
         )
 
         # To avoid collisions of group namings between connectors, they need to be prefixed
-        access_map[document_id] = DocumentAccess.build(
-            user_emails=list(non_ee_access.user_emails),
-            user_groups=user_group_info.get(document_id, []),
+        access_map[document_id] = DocumentAccess(
+            user_emails=non_ee_access.user_emails,
+            user_groups=set(user_group_info.get(document_id, [])),
             is_public=is_public_anywhere,
-            external_user_emails=list(ext_u_emails),
-            external_user_group_ids=list(ext_u_groups),
+            external_user_emails=ext_u_emails,
+            external_user_group_ids=ext_u_groups,
         )
     return access_map
 

backend/ee/onyx/background/celery/apps/primary.py

@@ -4,11 +4,12 @@ from ee.onyx.server.reporting.usage_export_generation import create_new_usage_re
 from onyx.background.celery.apps.primary import celery_app
 from onyx.background.task_utils import build_celery_task_wrapper
 from onyx.configs.app_configs import JOB_TIMEOUT
-from onyx.db.chat import delete_chat_session
-from onyx.db.chat import get_chat_sessions_older_than
-from onyx.db.engine import get_session_with_current_tenant
+from onyx.db.chat import delete_chat_sessions_older_than
+from onyx.db.engine import get_session_with_tenant
 from onyx.server.settings.store import load_settings
 from onyx.utils.logger import setup_logger
+from shared_configs.configs import MULTI_TENANT
+from shared_configs.contextvars import CURRENT_TENANT_ID_CONTEXTVAR
 
 logger = setup_logger()
 
@@ -17,28 +18,11 @@ logger = setup_logger()
 
 @build_celery_task_wrapper(name_chat_ttl_task)
 @celery_app.task(soft_time_limit=JOB_TIMEOUT)
-def perform_ttl_management_task(retention_limit_days: int, *, tenant_id: str) -> None:
-    with get_session_with_current_tenant() as db_session:
-        old_chat_sessions = get_chat_sessions_older_than(
-            retention_limit_days, db_session
-        )
-
-    for user_id, session_id in old_chat_sessions:
-        # one session per delete so that we don't blow up if a deletion fails.
-        with get_session_with_current_tenant() as db_session:
-            try:
-                delete_chat_session(
-                    user_id,
-                    session_id,
-                    db_session,
-                    include_deleted=True,
-                    hard_delete=True,
-                )
-            except Exception:
-                logger.exception(
-                    "delete_chat_session exceptioned. "
-                    f"user_id={user_id} session_id={session_id}"
-                )
+def perform_ttl_management_task(
+    retention_limit_days: int, *, tenant_id: str | None
+) -> None:
+    with get_session_with_tenant(tenant_id=tenant_id) as db_session:
+        delete_chat_sessions_older_than(retention_limit_days, db_session)
 
 
 #####
@@ -51,19 +35,24 @@ def perform_ttl_management_task(retention_limit_days: int, *, tenant_id: str) ->
     ignore_result=True,
     soft_time_limit=JOB_TIMEOUT,
 )
-def check_ttl_management_task(*, tenant_id: str) -> None:
+def check_ttl_management_task(*, tenant_id: str | None) -> None:
     """Runs periodically to check if any ttl tasks should be run and adds them
     to the queue"""
+    token = None
+    if MULTI_TENANT and tenant_id is not None:
+        token = CURRENT_TENANT_ID_CONTEXTVAR.set(tenant_id)
 
     settings = load_settings()
     retention_limit_days = settings.maximum_chat_retention_days
-    with get_session_with_current_tenant() as db_session:
+    with get_session_with_tenant(tenant_id=tenant_id) as db_session:
         if should_perform_chat_ttl_check(retention_limit_days, db_session):
             perform_ttl_management_task.apply_async(
                 kwargs=dict(
                     retention_limit_days=retention_limit_days, tenant_id=tenant_id
                 ),
             )
+    if token is not None:
+        CURRENT_TENANT_ID_CONTEXTVAR.reset(token)
 
 
 @celery_app.task(
@@ -71,9 +60,9 @@ def check_ttl_management_task(*, tenant_id: str) -> None:
     ignore_result=True,
     soft_time_limit=JOB_TIMEOUT,
 )
-def autogenerate_usage_report_task(*, tenant_id: str) -> None:
+def autogenerate_usage_report_task(*, tenant_id: str | None) -> None:
     """This generates usage report under the /admin/generate-usage/report endpoint"""
-    with get_session_with_current_tenant() as db_session:
+    with get_session_with_tenant(tenant_id=tenant_id) as db_session:
         create_new_usage_report(
             db_session=db_session,
             user_id=None,

backend/ee/onyx/background/celery/tasks/vespa/tasks.py

@@ -18,7 +18,7 @@ logger = setup_logger()
 
 
 def monitor_usergroup_taskset(
-    tenant_id: str, key_bytes: bytes, r: Redis, db_session: Session
+    tenant_id: str | None, key_bytes: bytes, r: Redis, db_session: Session
 ) -> None:
     """This function is likely to move in the worker refactor happening next."""
     fence_key = key_bytes.decode("utf-8")

backend/ee/onyx/chat/process_message.py

@@ -2,6 +2,7 @@ from ee.onyx.server.query_and_chat.models import OneShotQAResponse
 from onyx.chat.models import AllCitations
 from onyx.chat.models import LLMRelevanceFilterResponse
 from onyx.chat.models import OnyxAnswerPiece
+from onyx.chat.models import OnyxContexts
 from onyx.chat.models import QADocsResponse
 from onyx.chat.models import StreamingError
 from onyx.chat.process_message import ChatPacketStream
@@ -31,6 +32,8 @@ def gather_stream_for_answer_api(
             response.llm_selected_doc_indices = packet.llm_selected_doc_indices
         elif isinstance(packet, AllCitations):
             response.citations = packet.citations
+        elif isinstance(packet, OnyxContexts):
+            response.contexts = packet
 
     if answer:
         response.answer = answer

backend/ee/onyx/configs/app_configs.py

@@ -25,10 +25,6 @@ SAML_CONF_DIR = os.environ.get("SAML_CONF_DIR") or "/app/ee/onyx/configs/saml_co
 #####
 # Auto Permission Sync
 #####
-DEFAULT_PERMISSION_DOC_SYNC_FREQUENCY = int(
-    os.environ.get("DEFAULT_PERMISSION_DOC_SYNC_FREQUENCY") or 5 * 60
-)
-
 # In seconds, default is 5 minutes
 CONFLUENCE_PERMISSION_GROUP_SYNC_FREQUENCY = int(
     os.environ.get("CONFLUENCE_PERMISSION_GROUP_SYNC_FREQUENCY") or 5 * 60
@@ -43,7 +39,6 @@ CONFLUENCE_ANONYMOUS_ACCESS_IS_PUBLIC = (
 CONFLUENCE_PERMISSION_DOC_SYNC_FREQUENCY = int(
     os.environ.get("CONFLUENCE_PERMISSION_DOC_SYNC_FREQUENCY") or 5 * 60
 )
-
 NUM_PERMISSION_WORKERS = int(os.environ.get("NUM_PERMISSION_WORKERS") or 2)
 
 
@@ -64,26 +59,15 @@ SUPER_CLOUD_API_KEY = os.environ.get("SUPER_CLOUD_API_KEY", "api_key")
 
 OAUTH_SLACK_CLIENT_ID = os.environ.get("OAUTH_SLACK_CLIENT_ID", "")
 OAUTH_SLACK_CLIENT_SECRET = os.environ.get("OAUTH_SLACK_CLIENT_SECRET", "")
-OAUTH_CONFLUENCE_CLOUD_CLIENT_ID = os.environ.get(
-    "OAUTH_CONFLUENCE_CLOUD_CLIENT_ID", ""
-)
-OAUTH_CONFLUENCE_CLOUD_CLIENT_SECRET = os.environ.get(
-    "OAUTH_CONFLUENCE_CLOUD_CLIENT_SECRET", ""
-)
-OAUTH_JIRA_CLOUD_CLIENT_ID = os.environ.get("OAUTH_JIRA_CLOUD_CLIENT_ID", "")
-OAUTH_JIRA_CLOUD_CLIENT_SECRET = os.environ.get("OAUTH_JIRA_CLOUD_CLIENT_SECRET", "")
+OAUTH_CONFLUENCE_CLIENT_ID = os.environ.get("OAUTH_CONFLUENCE_CLIENT_ID", "")
+OAUTH_CONFLUENCE_CLIENT_SECRET = os.environ.get("OAUTH_CONFLUENCE_CLIENT_SECRET", "")
+OAUTH_JIRA_CLIENT_ID = os.environ.get("OAUTH_JIRA_CLIENT_ID", "")
+OAUTH_JIRA_CLIENT_SECRET = os.environ.get("OAUTH_JIRA_CLIENT_SECRET", "")
 OAUTH_GOOGLE_DRIVE_CLIENT_ID = os.environ.get("OAUTH_GOOGLE_DRIVE_CLIENT_ID", "")
 OAUTH_GOOGLE_DRIVE_CLIENT_SECRET = os.environ.get(
     "OAUTH_GOOGLE_DRIVE_CLIENT_SECRET", ""
 )
 
-GOOGLE_DRIVE_PERMISSION_GROUP_SYNC_FREQUENCY = int(
-    os.environ.get("GOOGLE_DRIVE_PERMISSION_GROUP_SYNC_FREQUENCY") or 5 * 60
-)
-
-SLACK_PERMISSION_DOC_SYNC_FREQUENCY = int(
-    os.environ.get("SLACK_PERMISSION_DOC_SYNC_FREQUENCY") or 5 * 60
-)
 
 # The posthog client does not accept empty API keys or hosts however it fails silently
 # when the capture is called. These defaults prevent Posthog issues from breaking the Onyx app

backend/ee/onyx/db/query_history.py

@@ -134,9 +134,7 @@ def fetch_chat_sessions_eagerly_by_time(
     limit: int | None = 500,
     initial_time: datetime | None = None,
 ) -> list[ChatSession]:
-    """Sorted by oldest to newest, then by message id"""
-
-    asc_time_order: UnaryExpression = asc(ChatSession.time_created)
+    time_order: UnaryExpression = desc(ChatSession.time_created)
     message_order: UnaryExpression = asc(ChatMessage.id)
 
     filters: list[ColumnElement | BinaryExpression] = [
@@ -149,7 +147,8 @@ def fetch_chat_sessions_eagerly_by_time(
     subquery = (
         db_session.query(ChatSession.id, ChatSession.time_created)
         .filter(*filters)
-        .order_by(asc_time_order)
+        .order_by(ChatSession.id, time_order)
+        .distinct(ChatSession.id)
         .limit(limit)
         .subquery()
     )
@@ -165,7 +164,7 @@ def fetch_chat_sessions_eagerly_by_time(
                 ChatMessage.chat_message_feedbacks
             ),
         )
-        .order_by(asc_time_order, message_order)
+        .order_by(time_order, message_order)
     )
 
     chat_sessions = query.all()

backend/ee/onyx/db/usage_export.py

@@ -16,20 +16,13 @@ from onyx.db.models import UsageReport
 from onyx.file_store.file_store import get_default_file_store
 
 
-# Gets skeletons of all messages in the given range
+# Gets skeletons of all message
 def get_empty_chat_messages_entries__paginated(
     db_session: Session,
     period: tuple[datetime, datetime],
     limit: int | None = 500,
     initial_time: datetime | None = None,
 ) -> tuple[Optional[datetime], list[ChatMessageSkeleton]]:
-    """Returns a tuple where:
-    first element is the most recent timestamp out of the sessions iterated
-    - this timestamp can be used to paginate forward in time
-    second element is a list of messages belonging to all the sessions iterated
-
-    Only messages of type USER are returned
-    """
     chat_sessions = fetch_chat_sessions_eagerly_by_time(
         start=period[0],
         end=period[1],
@@ -59,17 +52,18 @@ def get_empty_chat_messages_entries__paginated(
     if len(chat_sessions) == 0:
         return None, []
 
-    return chat_sessions[-1].time_created, message_skeletons
+    return chat_sessions[0].time_created, message_skeletons
 
 
 def get_all_empty_chat_message_entries(
     db_session: Session,
     period: tuple[datetime, datetime],
 ) -> Generator[list[ChatMessageSkeleton], None, None]:
-    """period is the range of time over which to fetch messages."""
     initial_time: Optional[datetime] = period[0]
+    ind = 0
     while True:
-        # iterate from oldest to newest
+        ind += 1
+
         time_created, message_skeletons = get_empty_chat_messages_entries__paginated(
             db_session,
             period,

backend/ee/onyx/db/user_group.py

@@ -424,7 +424,7 @@ def _validate_curator_status__no_commit(
         )
 
         # if the user is a curator in any of their groups, set their role to CURATOR
-        # otherwise, set their role to BASIC only if they were previously a CURATOR
+        # otherwise, set their role to BASIC
         if curator_relationships:
             user.role = UserRole.CURATOR
         elif user.role == UserRole.CURATOR:
@@ -631,16 +631,7 @@ def update_user_group(
     removed_users = db_session.scalars(
         select(User).where(User.id.in_(removed_user_ids))  # type: ignore
     ).unique()
-
-    # Filter out admin and global curator users before validating curator status
-    users_to_validate = [
-        user
-        for user in removed_users
-        if user.role not in [UserRole.ADMIN, UserRole.GLOBAL_CURATOR]
-    ]
-
-    if users_to_validate:
-        _validate_curator_status__no_commit(db_session, users_to_validate)
+    _validate_curator_status__no_commit(db_session, list(removed_users))
 
     # update "time_updated" to now
     db_user_group.time_last_modified_by_user = func.now()

backend/ee/onyx/external_permissions/confluence/doc_sync.py

@@ -2,7 +2,6 @@
 Rules defined here:
 https://confluence.atlassian.com/conf85/check-who-can-view-a-page-1283360557.html
 """
-from collections.abc import Generator
 from typing import Any
 
 from ee.onyx.configs.app_configs import CONFLUENCE_ANONYMOUS_ACCESS_IS_PUBLIC
@@ -10,16 +9,12 @@ from ee.onyx.external_permissions.confluence.constants import ALL_CONF_EMAILS_GR
 from onyx.access.models import DocExternalAccess
 from onyx.access.models import ExternalAccess
 from onyx.connectors.confluence.connector import ConfluenceConnector
-from onyx.connectors.confluence.onyx_confluence import (
-    get_user_email_from_username__server,
-)
 from onyx.connectors.confluence.onyx_confluence import OnyxConfluence
-from onyx.connectors.credentials_provider import OnyxDBCredentialsProvider
+from onyx.connectors.confluence.utils import get_user_email_from_username__server
 from onyx.connectors.models import SlimDocument
 from onyx.db.models import ConnectorCredentialPair
 from onyx.indexing.indexing_heartbeat import IndexingHeartbeatInterface
 from onyx.utils.logger import setup_logger
-from shared_configs.contextvars import get_current_tenant_id
 
 logger = setup_logger()
 
@@ -159,9 +154,6 @@ def _get_space_permissions(
 
         # Stores the permissions for each space
         space_permissions_by_space_key[space_key] = space_permissions
-        logger.info(
-            f"Found space permissions for space '{space_key}': {space_permissions}"
-        )
 
     return space_permissions_by_space_key
 
@@ -267,11 +259,13 @@ def _fetch_all_page_restrictions(
     space_permissions_by_space_key: dict[str, ExternalAccess],
     is_cloud: bool,
     callback: IndexingHeartbeatInterface | None,
-) -> Generator[DocExternalAccess, None, None]:
+) -> list[DocExternalAccess]:
     """
     For all pages, if a page has restrictions, then use those restrictions.
     Otherwise, use the space's restrictions.
     """
+    document_restrictions: list[DocExternalAccess] = []
+
     for slim_doc in slim_docs:
         if callback:
             if callback.should_stop():
@@ -288,9 +282,11 @@ def _fetch_all_page_restrictions(
             confluence_client=confluence_client,
             perm_sync_data=slim_doc.perm_sync_data,
         ):
-            yield DocExternalAccess(
-                doc_id=slim_doc.id,
-                external_access=restrictions,
+            document_restrictions.append(
+                DocExternalAccess(
+                    doc_id=slim_doc.id,
+                    external_access=restrictions,
+                )
             )
             # If there are restrictions, then we don't need to use the space's restrictions
             continue
@@ -324,9 +320,11 @@ def _fetch_all_page_restrictions(
             continue
 
         # If there are no restrictions, then use the space's restrictions
-        yield DocExternalAccess(
-            doc_id=slim_doc.id,
-            external_access=space_permissions,
+        document_restrictions.append(
+            DocExternalAccess(
+                doc_id=slim_doc.id,
+                external_access=space_permissions,
+            )
         )
         if (
             not space_permissions.is_public
@@ -340,12 +338,12 @@ def _fetch_all_page_restrictions(
             )
 
     logger.debug("Finished fetching all page restrictions for space")
+    return document_restrictions
 
 
 def confluence_doc_sync(
-    cc_pair: ConnectorCredentialPair,
-    callback: IndexingHeartbeatInterface | None,
-) -> Generator[DocExternalAccess, None, None]:
+    cc_pair: ConnectorCredentialPair, callback: IndexingHeartbeatInterface | None
+) -> list[DocExternalAccess]:
     """
     Adds the external permissions to the documents in postgres
     if the document doesn't already exists in postgres, we create
@@ -356,11 +354,7 @@ def confluence_doc_sync(
     confluence_connector = ConfluenceConnector(
         **cc_pair.connector.connector_specific_config
     )
-
-    provider = OnyxDBCredentialsProvider(
-        get_current_tenant_id(), "confluence", cc_pair.credential_id
-    )
-    confluence_connector.set_credentials_provider(provider)
+    confluence_connector.load_credentials(cc_pair.credential.credential_json)
 
     is_cloud = cc_pair.connector.connector_specific_config.get("is_cloud", False)
 
@@ -384,7 +378,7 @@ def confluence_doc_sync(
         slim_docs.extend(doc_batch)
 
     logger.debug("Fetching all page restrictions for space")
-    yield from _fetch_all_page_restrictions(
+    return _fetch_all_page_restrictions(
         confluence_client=confluence_connector.confluence_client,
         slim_docs=slim_docs,
         space_permissions_by_space_key=space_permissions_by_space_key,

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

@@ -1,11 +1,9 @@
 from ee.onyx.db.external_perm import ExternalUserGroup
 from ee.onyx.external_permissions.confluence.constants import ALL_CONF_EMAILS_GROUP_NAME
 from onyx.background.error_logging import emit_background_error
-from onyx.connectors.confluence.onyx_confluence import (
-    get_user_email_from_username__server,
-)
+from onyx.connectors.confluence.onyx_confluence import build_confluence_client
 from onyx.connectors.confluence.onyx_confluence import OnyxConfluence
-from onyx.connectors.credentials_provider import OnyxDBCredentialsProvider
+from onyx.connectors.confluence.utils import get_user_email_from_username__server
 from onyx.db.models import ConnectorCredentialPair
 from onyx.utils.logger import setup_logger
 
@@ -63,27 +61,13 @@ def _build_group_member_email_map(
 
 
 def confluence_group_sync(
-    tenant_id: str,
     cc_pair: ConnectorCredentialPair,
 ) -> list[ExternalUserGroup]:
-    provider = OnyxDBCredentialsProvider(tenant_id, "confluence", cc_pair.credential_id)
-    is_cloud = cc_pair.connector.connector_specific_config.get("is_cloud", False)
-    wiki_base: str = cc_pair.connector.connector_specific_config["wiki_base"]
-    url = wiki_base.rstrip("/")
-
-    probe_kwargs = {
-        "max_backoff_retries": 6,
-        "max_backoff_seconds": 10,
-    }
-
-    final_kwargs = {
-        "max_backoff_retries": 10,
-        "max_backoff_seconds": 60,
-    }
-
-    confluence_client = OnyxConfluence(is_cloud, url, provider)
-    confluence_client._probe_connection(**probe_kwargs)
-    confluence_client._initialize_connection(**final_kwargs)
+    confluence_client = build_confluence_client(
+        credentials=cc_pair.credential.credential_json,
+        is_cloud=cc_pair.connector.connector_specific_config.get("is_cloud", False),
+        wiki_base=cc_pair.connector.connector_specific_config["wiki_base"],
+    )
 
     group_member_email_map = _build_group_member_email_map(
         confluence_client=confluence_client,

backend/ee/onyx/external_permissions/gmail/doc_sync.py

@@ -1,4 +1,3 @@
-from collections.abc import Generator
 from datetime import datetime
 from datetime import timezone
 
@@ -33,9 +32,8 @@ def _get_slim_doc_generator(
 
 
 def gmail_doc_sync(
-    cc_pair: ConnectorCredentialPair,
-    callback: IndexingHeartbeatInterface | None,
-) -> Generator[DocExternalAccess, None, None]:
+    cc_pair: ConnectorCredentialPair, callback: IndexingHeartbeatInterface | None
+) -> list[DocExternalAccess]:
     """
     Adds the external permissions to the documents in postgres
     if the document doesn't already exists in postgres, we create
@@ -49,6 +47,7 @@ def gmail_doc_sync(
         cc_pair, gmail_connector, callback=callback
     )
 
+    document_external_access: list[DocExternalAccess] = []
     for slim_doc_batch in slim_doc_generator:
         for slim_doc in slim_doc_batch:
             if callback:
@@ -60,14 +59,17 @@ def gmail_doc_sync(
             if slim_doc.perm_sync_data is None:
                 logger.warning(f"No permissions found for document {slim_doc.id}")
                 continue
-
             if user_email := slim_doc.perm_sync_data.get("user_email"):
                 ext_access = ExternalAccess(
                     external_user_emails=set([user_email]),
                     external_user_group_ids=set(),
                     is_public=False,
                 )
-                yield DocExternalAccess(
-                    doc_id=slim_doc.id,
-                    external_access=ext_access,
+                document_external_access.append(
+                    DocExternalAccess(
+                        doc_id=slim_doc.id,
+                        external_access=ext_access,
+                    )
                 )
+
+    return document_external_access

backend/ee/onyx/external_permissions/google_drive/doc_sync.py

@@ -1,4 +1,3 @@
-from collections.abc import Generator
 from datetime import datetime
 from datetime import timezone
 from typing import Any
@@ -146,9 +145,8 @@ def _get_permissions_from_slim_doc(
 
 
 def gdrive_doc_sync(
-    cc_pair: ConnectorCredentialPair,
-    callback: IndexingHeartbeatInterface | None,
-) -> Generator[DocExternalAccess, None, None]:
+    cc_pair: ConnectorCredentialPair, callback: IndexingHeartbeatInterface | None
+) -> list[DocExternalAccess]:
     """
     Adds the external permissions to the documents in postgres
     if the document doesn't already exists in postgres, we create
@@ -162,6 +160,7 @@ def gdrive_doc_sync(
 
     slim_doc_generator = _get_slim_doc_generator(cc_pair, google_drive_connector)
 
+    document_external_accesses = []
     for slim_doc_batch in slim_doc_generator:
         for slim_doc in slim_doc_batch:
             if callback:
@@ -174,7 +173,10 @@ def gdrive_doc_sync(
                 google_drive_connector=google_drive_connector,
                 slim_doc=slim_doc,
             )
-            yield DocExternalAccess(
-                external_access=ext_access,
-                doc_id=slim_doc.id,
+            document_external_accesses.append(
+                DocExternalAccess(
+                    external_access=ext_access,
+                    doc_id=slim_doc.id,
+                )
             )
+    return document_external_accesses

backend/ee/onyx/external_permissions/google_drive/group_sync.py

@@ -119,7 +119,6 @@ def _build_onyx_groups(
 
 
 def gdrive_group_sync(
-    tenant_id: str,
     cc_pair: ConnectorCredentialPair,
 ) -> list[ExternalUserGroup]:
     # Initialize connector and build credential/service objects

backend/ee/onyx/external_permissions/post_query_censoring.py

@@ -55,7 +55,7 @@ def _post_query_chunk_censoring(
         # if user is None, permissions are not enforced
         return chunks
 
-    final_chunk_dict: dict[str, InferenceChunk] = {}
+    chunks_to_keep = []
     chunks_to_process: dict[DocumentSource, list[InferenceChunk]] = {}
 
     sources_to_censor = _get_all_censoring_enabled_sources()
@@ -64,7 +64,7 @@ def _post_query_chunk_censoring(
         if chunk.source_type in sources_to_censor:
             chunks_to_process.setdefault(chunk.source_type, []).append(chunk)
         else:
-            final_chunk_dict[chunk.unique_id] = chunk
+            chunks_to_keep.append(chunk)
 
     # For each source, filter out the chunks using the permission
     # check function for that source
@@ -79,16 +79,6 @@ def _post_query_chunk_censoring(
                 f" chunks for this source and continuing: {e}"
             )
             continue
+        chunks_to_keep.extend(censored_chunks)
 
-        for censored_chunk in censored_chunks:
-            final_chunk_dict[censored_chunk.unique_id] = censored_chunk
-
-    # IMPORTANT: make sure to retain the same ordering as the original `chunks` passed in
-    final_chunk_list: list[InferenceChunk] = []
-    for chunk in chunks:
-        # only if the chunk is in the final censored chunks, add it to the final list
-        # if it is missing, that means it was intentionally left out
-        if chunk.unique_id in final_chunk_dict:
-            final_chunk_list.append(final_chunk_dict[chunk.unique_id])
-
-    return final_chunk_list
+    return chunks_to_keep

backend/ee/onyx/external_permissions/salesforce/postprocessing.py

@@ -58,7 +58,6 @@ def _get_objects_access_for_user_email_from_salesforce(
         f"Time taken to get Salesforce user ID: {end_time - start_time} seconds"
     )
     if user_id is None:
-        logger.warning(f"User '{user_email}' not found in Salesforce")
         return None
 
     # This is the only query that is not cached in the function
@@ -66,7 +65,6 @@ def _get_objects_access_for_user_email_from_salesforce(
     object_id_to_access = get_objects_access_for_user_id(
         salesforce_client, user_id, list(object_ids)
     )
-    logger.debug(f"Object ID to access: {object_id_to_access}")
     return object_id_to_access
 
 

backend/ee/onyx/external_permissions/salesforce/utils.py

@@ -42,18 +42,11 @@ def get_any_salesforce_client_for_doc_id(
 
 
 def _query_salesforce_user_id(sf_client: Salesforce, user_email: str) -> str | None:
-    query = f"SELECT Id FROM User WHERE Username = '{user_email}' AND IsActive = true"
+    query = f"SELECT Id FROM User WHERE Email = '{user_email}'"
     result = sf_client.query(query)
-    if len(result["records"]) > 0:
-        return result["records"][0]["Id"]
-
-    # try emails
-    query = f"SELECT Id FROM User WHERE Email = '{user_email}' AND IsActive = true"
-    result = sf_client.query(query)
-    if len(result["records"]) > 0:
-        return result["records"][0]["Id"]
-
-    return None
+    if len(result["records"]) == 0:
+        return None
+    return result["records"][0]["Id"]
 
 
 # This contains only the user_ids that we have found in Salesforce.

backend/ee/onyx/external_permissions/slack/doc_sync.py

@@ -1,5 +1,3 @@
-from collections.abc import Generator
-
 from slack_sdk import WebClient
 
 from ee.onyx.external_permissions.slack.utils import fetch_user_id_to_email_map
@@ -16,6 +14,35 @@ from onyx.utils.logger import setup_logger
 logger = setup_logger()
 
 
+def _get_slack_document_ids_and_channels(
+    cc_pair: ConnectorCredentialPair, callback: IndexingHeartbeatInterface | None
+) -> dict[str, list[str]]:
+    slack_connector = SlackConnector(**cc_pair.connector.connector_specific_config)
+    slack_connector.load_credentials(cc_pair.credential.credential_json)
+
+    slim_doc_generator = slack_connector.retrieve_all_slim_documents(callback=callback)
+
+    channel_doc_map: dict[str, list[str]] = {}
+    for doc_metadata_batch in slim_doc_generator:
+        for doc_metadata in doc_metadata_batch:
+            if doc_metadata.perm_sync_data is None:
+                continue
+            channel_id = doc_metadata.perm_sync_data["channel_id"]
+            if channel_id not in channel_doc_map:
+                channel_doc_map[channel_id] = []
+            channel_doc_map[channel_id].append(doc_metadata.id)
+
+        if callback:
+            if callback.should_stop():
+                raise RuntimeError(
+                    "_get_slack_document_ids_and_channels: Stop signal detected"
+                )
+
+            callback.progress("_get_slack_document_ids_and_channels", 1)
+
+    return channel_doc_map
+
+
 def _fetch_workspace_permissions(
     user_id_to_email_map: dict[str, str],
 ) -> ExternalAccess:
@@ -95,37 +122,9 @@ def _fetch_channel_permissions(
     return channel_permissions
 
 
-def _get_slack_document_access(
-    cc_pair: ConnectorCredentialPair,
-    channel_permissions: dict[str, ExternalAccess],
-    callback: IndexingHeartbeatInterface | None,
-) -> Generator[DocExternalAccess, None, None]:
-    slack_connector = SlackConnector(**cc_pair.connector.connector_specific_config)
-    slack_connector.load_credentials(cc_pair.credential.credential_json)
-
-    slim_doc_generator = slack_connector.retrieve_all_slim_documents(callback=callback)
-
-    for doc_metadata_batch in slim_doc_generator:
-        for doc_metadata in doc_metadata_batch:
-            if doc_metadata.perm_sync_data is None:
-                continue
-            channel_id = doc_metadata.perm_sync_data["channel_id"]
-            yield DocExternalAccess(
-                external_access=channel_permissions[channel_id],
-                doc_id=doc_metadata.id,
-            )
-
-        if callback:
-            if callback.should_stop():
-                raise RuntimeError("_get_slack_document_access: Stop signal detected")
-
-            callback.progress("_get_slack_document_access", 1)
-
-
 def slack_doc_sync(
-    cc_pair: ConnectorCredentialPair,
-    callback: IndexingHeartbeatInterface | None,
-) -> Generator[DocExternalAccess, None, None]:
+    cc_pair: ConnectorCredentialPair, callback: IndexingHeartbeatInterface | None
+) -> list[DocExternalAccess]:
     """
     Adds the external permissions to the documents in postgres
     if the document doesn't already exists in postgres, we create
@@ -136,12 +135,9 @@ def slack_doc_sync(
         token=cc_pair.credential.credential_json["slack_bot_token"]
     )
     user_id_to_email_map = fetch_user_id_to_email_map(slack_client)
-    if not user_id_to_email_map:
-        raise ValueError(
-            "No user id to email map found. Please check to make sure that "
-            "your Slack bot token has the `users:read.email` scope"
-        )
-
+    channel_doc_map = _get_slack_document_ids_and_channels(
+        cc_pair=cc_pair, callback=callback
+    )
     workspace_permissions = _fetch_workspace_permissions(
         user_id_to_email_map=user_id_to_email_map,
     )
@@ -151,8 +147,18 @@ def slack_doc_sync(
         user_id_to_email_map=user_id_to_email_map,
     )
 
-    yield from _get_slack_document_access(
-        cc_pair=cc_pair,
-        channel_permissions=channel_permissions,
-        callback=callback,
-    )
+    document_external_accesses = []
+    for channel_id, ext_access in channel_permissions.items():
+        doc_ids = channel_doc_map.get(channel_id)
+        if not doc_ids:
+            # No documents found for channel the channel_id
+            continue
+
+        for doc_id in doc_ids:
+            document_external_accesses.append(
+                DocExternalAccess(
+                    external_access=ext_access,
+                    doc_id=doc_id,
+                )
+            )
+    return document_external_accesses

backend/ee/onyx/external_permissions/sync_params.py

@@ -1,10 +1,7 @@
 from collections.abc import Callable
-from collections.abc import Generator
 
 from ee.onyx.configs.app_configs import CONFLUENCE_PERMISSION_DOC_SYNC_FREQUENCY
 from ee.onyx.configs.app_configs import CONFLUENCE_PERMISSION_GROUP_SYNC_FREQUENCY
-from ee.onyx.configs.app_configs import GOOGLE_DRIVE_PERMISSION_GROUP_SYNC_FREQUENCY
-from ee.onyx.configs.app_configs import SLACK_PERMISSION_DOC_SYNC_FREQUENCY
 from ee.onyx.db.external_perm import ExternalUserGroup
 from ee.onyx.external_permissions.confluence.doc_sync import confluence_doc_sync
 from ee.onyx.external_permissions.confluence.group_sync import confluence_group_sync
@@ -26,12 +23,11 @@ DocSyncFuncType = Callable[
         ConnectorCredentialPair,
         IndexingHeartbeatInterface | None,
     ],
-    Generator[DocExternalAccess, None, None],
+    list[DocExternalAccess],
 ]
 
 GroupSyncFuncType = Callable[
     [
-        str,
         ConnectorCredentialPair,
     ],
     list[ExternalUserGroup],
@@ -68,13 +64,13 @@ GROUP_PERMISSIONS_IS_CC_PAIR_AGNOSTIC: set[DocumentSource] = {
 DOC_PERMISSION_SYNC_PERIODS: dict[DocumentSource, int] = {
     # Polling is not supported so we fetch all doc permissions every 5 minutes
     DocumentSource.CONFLUENCE: CONFLUENCE_PERMISSION_DOC_SYNC_FREQUENCY,
-    DocumentSource.SLACK: SLACK_PERMISSION_DOC_SYNC_FREQUENCY,
+    DocumentSource.SLACK: 5 * 60,
 }
 
 # If nothing is specified here, we run the doc_sync every time the celery beat runs
 EXTERNAL_GROUP_SYNC_PERIODS: dict[DocumentSource, int] = {
     # Polling is not supported so we fetch all group permissions every 30 minutes
-    DocumentSource.GOOGLE_DRIVE: GOOGLE_DRIVE_PERMISSION_GROUP_SYNC_FREQUENCY,
+    DocumentSource.GOOGLE_DRIVE: 5 * 60,
     DocumentSource.CONFLUENCE: CONFLUENCE_PERMISSION_GROUP_SYNC_FREQUENCY,
 }
 

backend/ee/onyx/main.py

@@ -15,7 +15,7 @@ from ee.onyx.server.enterprise_settings.api import (
 )
 from ee.onyx.server.manage.standard_answer import router as standard_answer_router
 from ee.onyx.server.middleware.tenant_tracking import add_tenant_id_middleware
-from ee.onyx.server.oauth.api import router as ee_oauth_router
+from ee.onyx.server.oauth import router as oauth_router
 from ee.onyx.server.query_and_chat.chat_backend import (
     router as chat_router,
 )
@@ -64,15 +64,7 @@ def get_application() -> FastAPI:
         add_tenant_id_middleware(application, logger)
 
     if AUTH_TYPE == AuthType.CLOUD:
-        # For Google OAuth, refresh tokens are requested by:
-        # 1. Adding the right scopes
-        # 2. Properly configuring OAuth in Google Cloud Console to allow offline access
-        oauth_client = GoogleOAuth2(
-            OAUTH_CLIENT_ID,
-            OAUTH_CLIENT_SECRET,
-            # Use standard scopes that include profile and email
-            scopes=["openid", "email", "profile"],
-        )
+        oauth_client = GoogleOAuth2(OAUTH_CLIENT_ID, OAUTH_CLIENT_SECRET)
         include_auth_router_with_prefix(
             application,
             create_onyx_oauth_router(
@@ -95,16 +87,6 @@ def get_application() -> FastAPI:
         )
 
     if AUTH_TYPE == AuthType.OIDC:
-        # Ensure we request offline_access for refresh tokens
-        try:
-            oidc_scopes = list(OIDC_SCOPE_OVERRIDE or BASE_SCOPES)
-            if "offline_access" not in oidc_scopes:
-                oidc_scopes.append("offline_access")
-        except Exception as e:
-            logger.warning(f"Error configuring OIDC scopes: {e}")
-            # Fall back to default scopes if there's an error
-            oidc_scopes = BASE_SCOPES
-
         include_auth_router_with_prefix(
             application,
             create_onyx_oauth_router(
@@ -112,8 +94,8 @@ def get_application() -> FastAPI:
                     OAUTH_CLIENT_ID,
                     OAUTH_CLIENT_SECRET,
                     OPENID_CONFIG_URL,
-                    # Use the configured scopes
-                    base_scopes=oidc_scopes,
+                    # BASE_SCOPES is the same as not setting this
+                    base_scopes=OIDC_SCOPE_OVERRIDE or BASE_SCOPES,
                 ),
                 auth_backend,
                 USER_AUTH_SECRET,
@@ -146,7 +128,7 @@ def get_application() -> FastAPI:
     include_router_with_global_prefix_prepended(application, query_router)
     include_router_with_global_prefix_prepended(application, chat_router)
     include_router_with_global_prefix_prepended(application, standard_answer_router)
-    include_router_with_global_prefix_prepended(application, ee_oauth_router)
+    include_router_with_global_prefix_prepended(application, oauth_router)
 
     # Enterprise-only global settings
     include_router_with_global_prefix_prepended(
@@ -170,8 +152,4 @@ def get_application() -> FastAPI:
     # environment variable. Used to automate deployment for multiple environments.
     seed_db()
 
-    # for debugging discovered routes
-    # for route in application.router.routes:
-    #     print(f"Path: {route.path}, Methods: {route.methods}")
-
     return application

backend/ee/onyx/onyxbot/slack/handlers/handle_standard_answers.py

@@ -22,7 +22,7 @@ from onyx.onyxbot.slack.blocks import get_restate_blocks
 from onyx.onyxbot.slack.constants import GENERATE_ANSWER_BUTTON_ACTION_ID
 from onyx.onyxbot.slack.handlers.utils import send_team_member_message
 from onyx.onyxbot.slack.models import SlackMessageInfo
-from onyx.onyxbot.slack.utils import respond_in_thread_or_channel
+from onyx.onyxbot.slack.utils import respond_in_thread
 from onyx.onyxbot.slack.utils import update_emote_react
 from onyx.utils.logger import OnyxLoggingAdapter
 from onyx.utils.logger import setup_logger
@@ -216,7 +216,7 @@ def _handle_standard_answers(
         all_blocks = restate_question_blocks + answer_blocks
 
         try:
-            respond_in_thread_or_channel(
+            respond_in_thread(
                 client=client,
                 channel=message_info.channel_to_respond,
                 receiver_ids=receiver_ids,
@@ -231,7 +231,6 @@ def _handle_standard_answers(
                     client=client,
                     channel=message_info.channel_to_respond,
                     thread_ts=slack_thread_id,
-                    receiver_ids=receiver_ids,
                 )
 
             return True

backend/ee/onyx/server/enterprise_settings/api.py

@@ -15,8 +15,8 @@ from sqlalchemy.orm import Session
 
 from ee.onyx.server.enterprise_settings.models import AnalyticsScriptUpload
 from ee.onyx.server.enterprise_settings.models import EnterpriseSettings
-from ee.onyx.server.enterprise_settings.store import get_logo_filename
-from ee.onyx.server.enterprise_settings.store import get_logotype_filename
+from ee.onyx.server.enterprise_settings.store import _LOGO_FILENAME
+from ee.onyx.server.enterprise_settings.store import _LOGOTYPE_FILENAME
 from ee.onyx.server.enterprise_settings.store import load_analytics_script
 from ee.onyx.server.enterprise_settings.store import load_settings
 from ee.onyx.server.enterprise_settings.store import store_analytics_script
@@ -28,7 +28,7 @@ from onyx.auth.users import get_user_manager
 from onyx.auth.users import UserManager
 from onyx.db.engine import get_session
 from onyx.db.models import User
-from onyx.file_store.file_store import PostgresBackedFileStore
+from onyx.file_store.file_store import get_default_file_store
 from onyx.utils.logger import setup_logger
 
 admin_router = APIRouter(prefix="/admin/enterprise-settings")
@@ -131,49 +131,31 @@ def put_logo(
     upload_logo(file=file, db_session=db_session, is_logotype=is_logotype)
 
 
-def fetch_logo_helper(db_session: Session) -> Response:
+def fetch_logo_or_logotype(is_logotype: bool, db_session: Session) -> Response:
     try:
-        file_store = PostgresBackedFileStore(db_session)
-        onyx_file = file_store.get_file_with_mime_type(get_logo_filename())
-        if not onyx_file:
-            raise ValueError("get_onyx_file returned None!")
+        file_store = get_default_file_store(db_session)
+        filename = _LOGOTYPE_FILENAME if is_logotype else _LOGO_FILENAME
+        file_io = file_store.read_file(filename, mode="b")
+        # NOTE: specifying "image/jpeg" here, but it still works for pngs
+        # TODO: do this properly
+        return Response(content=file_io.read(), media_type="image/jpeg")
     except Exception:
         raise HTTPException(
             status_code=404,
-            detail="No logo file found",
+            detail=f"No {'logotype' if is_logotype else 'logo'} file found",
         )
-    else:
-        return Response(content=onyx_file.data, media_type=onyx_file.mime_type)
-
-
-def fetch_logotype_helper(db_session: Session) -> Response:
-    try:
-        file_store = PostgresBackedFileStore(db_session)
-        onyx_file = file_store.get_file_with_mime_type(get_logotype_filename())
-        if not onyx_file:
-            raise ValueError("get_onyx_file returned None!")
-    except Exception:
-        raise HTTPException(
-            status_code=404,
-            detail="No logotype file found",
-        )
-    else:
-        return Response(content=onyx_file.data, media_type=onyx_file.mime_type)
 
 
 @basic_router.get("/logotype")
 def fetch_logotype(db_session: Session = Depends(get_session)) -> Response:
-    return fetch_logotype_helper(db_session)
+    return fetch_logo_or_logotype(is_logotype=True, db_session=db_session)
 
 
 @basic_router.get("/logo")
 def fetch_logo(
     is_logotype: bool = False, db_session: Session = Depends(get_session)
 ) -> Response:
-    if is_logotype:
-        return fetch_logotype_helper(db_session)
-
-    return fetch_logo_helper(db_session)
+    return fetch_logo_or_logotype(is_logotype=is_logotype, db_session=db_session)
 
 
 @admin_router.put("/custom-analytics-script")

backend/ee/onyx/server/enterprise_settings/store.py

@@ -13,7 +13,6 @@ from ee.onyx.server.enterprise_settings.models import EnterpriseSettings
 from onyx.configs.constants import FileOrigin
 from onyx.configs.constants import KV_CUSTOM_ANALYTICS_SCRIPT_KEY
 from onyx.configs.constants import KV_ENTERPRISE_SETTINGS_KEY
-from onyx.configs.constants import ONYX_DEFAULT_APPLICATION_NAME
 from onyx.file_store.file_store import get_default_file_store
 from onyx.key_value_store.factory import get_kv_store
 from onyx.key_value_store.interface import KvKeyNotFoundError
@@ -22,18 +21,8 @@ from onyx.utils.logger import setup_logger
 
 logger = setup_logger()
 
-_LOGO_FILENAME = "__logo__"
-_LOGOTYPE_FILENAME = "__logotype__"
-
 
 def load_settings() -> EnterpriseSettings:
-    """Loads settings data directly from DB. This should be used primarily
-    for checking what is actually in the DB, aka for editing and saving back settings.
-
-    Runtime settings actually used by the application should be checked with
-    load_runtime_settings as defaults may be applied at runtime.
-    """
-
     dynamic_config_store = get_kv_store()
     try:
         settings = EnterpriseSettings(
@@ -47,24 +36,9 @@ def load_settings() -> EnterpriseSettings:
 
 
 def store_settings(settings: EnterpriseSettings) -> None:
-    """Stores settings directly to the kv store / db."""
-
     get_kv_store().store(KV_ENTERPRISE_SETTINGS_KEY, settings.model_dump())
 
 
-def load_runtime_settings() -> EnterpriseSettings:
-    """Loads settings from DB and applies any defaults or transformations for use
-    at runtime.
-
-    Should not be stored back to the DB.
-    """
-    enterprise_settings = load_settings()
-    if not enterprise_settings.application_name:
-        enterprise_settings.application_name = ONYX_DEFAULT_APPLICATION_NAME
-
-    return enterprise_settings
-
-
 _CUSTOM_ANALYTICS_SECRET_KEY = os.environ.get("CUSTOM_ANALYTICS_SECRET_KEY")
 
 
@@ -86,6 +60,10 @@ def store_analytics_script(analytics_script_upload: AnalyticsScriptUpload) -> No
     get_kv_store().store(KV_CUSTOM_ANALYTICS_SCRIPT_KEY, analytics_script_upload.script)
 
 
+_LOGO_FILENAME = "__logo__"
+_LOGOTYPE_FILENAME = "__logotype__"
+
+
 def is_valid_file_type(filename: str) -> bool:
     valid_extensions = (".png", ".jpg", ".jpeg")
     return filename.endswith(valid_extensions)
@@ -138,11 +116,3 @@ def upload_logo(
         file_type=file_type,
     )
     return True
-
-
-def get_logo_filename() -> str:
-    return _LOGO_FILENAME
-
-
-def get_logotype_filename() -> str:
-    return _LOGOTYPE_FILENAME

backend/ee/onyx/server/middleware/tenant_tracking.py

@@ -44,7 +44,7 @@ async def _get_tenant_id_from_request(
     Attempt to extract tenant_id from:
     1) The API key header
     2) The Redis-based token (stored in Cookie: fastapiusersauth)
-    3) The anonymous user cookie
+    3)  Reset token cookie
     Fallback: POSTGRES_DEFAULT_SCHEMA
     """
     # Check for API key
@@ -52,55 +52,41 @@ async def _get_tenant_id_from_request(
     if tenant_id is not None:
         return tenant_id
 
+    # Check for anonymous user cookie
+    anonymous_user_cookie = request.cookies.get(ANONYMOUS_USER_COOKIE_NAME)
+    if anonymous_user_cookie:
+        try:
+            anonymous_user_data = decode_anonymous_user_jwt_token(anonymous_user_cookie)
+            return anonymous_user_data.get("tenant_id", POSTGRES_DEFAULT_SCHEMA)
+        except Exception as e:
+            logger.error(f"Error decoding anonymous user cookie: {str(e)}")
+            # Continue and attempt to authenticate
+
     try:
         # Look up token data in Redis
 
         token_data = await retrieve_auth_token_data_from_redis(request)
 
-        if token_data:
-            tenant_id_from_payload = token_data.get(
-                "tenant_id", POSTGRES_DEFAULT_SCHEMA
+        if not token_data:
+            logger.debug(
+                "Token data not found or expired in Redis, defaulting to POSTGRES_DEFAULT_SCHEMA"
             )
+            # Return POSTGRES_DEFAULT_SCHEMA, so non-authenticated requests are sent to the default schema
+            # The CURRENT_TENANT_ID_CONTEXTVAR is initialized with POSTGRES_DEFAULT_SCHEMA,
+            # so we maintain consistency by returning it here when no valid tenant is found.
+            return POSTGRES_DEFAULT_SCHEMA
 
-            tenant_id = (
-                str(tenant_id_from_payload)
-                if tenant_id_from_payload is not None
-                else None
-            )
+        tenant_id_from_payload = token_data.get("tenant_id", POSTGRES_DEFAULT_SCHEMA)
 
-            if tenant_id and not is_valid_schema_name(tenant_id):
-                raise HTTPException(status_code=400, detail="Invalid tenant ID format")
-
-        # Check for anonymous user cookie
-        anonymous_user_cookie = request.cookies.get(ANONYMOUS_USER_COOKIE_NAME)
-        if anonymous_user_cookie:
-            try:
-                anonymous_user_data = decode_anonymous_user_jwt_token(
-                    anonymous_user_cookie
-                )
-                tenant_id = anonymous_user_data.get(
-                    "tenant_id", POSTGRES_DEFAULT_SCHEMA
-                )
-
-                if not tenant_id or not is_valid_schema_name(tenant_id):
-                    raise HTTPException(
-                        status_code=400, detail="Invalid tenant ID format"
-                    )
-
-                return tenant_id
-
-            except Exception as e:
-                logger.error(f"Error decoding anonymous user cookie: {str(e)}")
-                # Continue and attempt to authenticate
-
-        logger.debug(
-            "Token data not found or expired in Redis, defaulting to POSTGRES_DEFAULT_SCHEMA"
+        # Since token_data.get() can return None, ensure we have a string
+        tenant_id = (
+            str(tenant_id_from_payload)
+            if tenant_id_from_payload is not None
+            else POSTGRES_DEFAULT_SCHEMA
         )
 
-        # Return POSTGRES_DEFAULT_SCHEMA, so non-authenticated requests are sent to the default schema
-        # The CURRENT_TENANT_ID_CONTEXTVAR is initialized with POSTGRES_DEFAULT_SCHEMA,
-        # so we maintain consistency by returning it here when no valid tenant is found.
-        return POSTGRES_DEFAULT_SCHEMA
+        if not is_valid_schema_name(tenant_id):
+            raise HTTPException(status_code=400, detail="Invalid tenant ID format")
 
     except Exception as e:
         logger.error(f"Unexpected error in _get_tenant_id_from_request: {str(e)}")

backend/ee/onyx/server/oauth.py

@@ -0,0 +1,629 @@
+import base64
+import json
+import uuid
+from typing import Any
+from typing import cast
+
+import requests
+from fastapi import APIRouter
+from fastapi import Depends
+from fastapi import HTTPException
+from fastapi.responses import JSONResponse
+from pydantic import BaseModel
+from sqlalchemy.orm import Session
+
+from ee.onyx.configs.app_configs import OAUTH_CONFLUENCE_CLIENT_ID
+from ee.onyx.configs.app_configs import OAUTH_CONFLUENCE_CLIENT_SECRET
+from ee.onyx.configs.app_configs import OAUTH_GOOGLE_DRIVE_CLIENT_ID
+from ee.onyx.configs.app_configs import OAUTH_GOOGLE_DRIVE_CLIENT_SECRET
+from ee.onyx.configs.app_configs import OAUTH_SLACK_CLIENT_ID
+from ee.onyx.configs.app_configs import OAUTH_SLACK_CLIENT_SECRET
+from onyx.auth.users import current_user
+from onyx.configs.app_configs import WEB_DOMAIN
+from onyx.configs.constants import DocumentSource
+from onyx.connectors.google_utils.google_auth import get_google_oauth_creds
+from onyx.connectors.google_utils.google_auth import sanitize_oauth_credentials
+from onyx.connectors.google_utils.shared_constants import (
+    DB_CREDENTIALS_AUTHENTICATION_METHOD,
+)
+from onyx.connectors.google_utils.shared_constants import (
+    DB_CREDENTIALS_DICT_TOKEN_KEY,
+)
+from onyx.connectors.google_utils.shared_constants import (
+    DB_CREDENTIALS_PRIMARY_ADMIN_KEY,
+)
+from onyx.connectors.google_utils.shared_constants import (
+    GoogleOAuthAuthenticationMethod,
+)
+from onyx.db.credentials import create_credential
+from onyx.db.engine import get_session
+from onyx.db.models import User
+from onyx.redis.redis_pool import get_redis_client
+from onyx.server.documents.models import CredentialBase
+from onyx.utils.logger import setup_logger
+from shared_configs.contextvars import get_current_tenant_id
+
+
+logger = setup_logger()
+
+router = APIRouter(prefix="/oauth")
+
+
+class SlackOAuth:
+    # https://knock.app/blog/how-to-authenticate-users-in-slack-using-oauth
+    # Example: https://api.slack.com/authentication/oauth-v2#exchanging
+
+    class OAuthSession(BaseModel):
+        """Stored in redis to be looked up on callback"""
+
+        email: str
+        redirect_on_success: str | None  # Where to send the user if OAuth flow succeeds
+
+    CLIENT_ID = OAUTH_SLACK_CLIENT_ID
+    CLIENT_SECRET = OAUTH_SLACK_CLIENT_SECRET
+
+    TOKEN_URL = "https://slack.com/api/oauth.v2.access"
+
+    # SCOPE is per https://docs.onyx.app/connectors/slack
+    BOT_SCOPE = (
+        "channels:history,"
+        "channels:read,"
+        "groups:history,"
+        "groups:read,"
+        "channels:join,"
+        "im:history,"
+        "users:read,"
+        "users:read.email,"
+        "usergroups:read"
+    )
+
+    REDIRECT_URI = f"{WEB_DOMAIN}/admin/connectors/slack/oauth/callback"
+    DEV_REDIRECT_URI = f"https://redirectmeto.com/{REDIRECT_URI}"
+
+    @classmethod
+    def generate_oauth_url(cls, state: str) -> str:
+        return cls._generate_oauth_url_helper(cls.REDIRECT_URI, state)
+
+    @classmethod
+    def generate_dev_oauth_url(cls, state: str) -> str:
+        """dev mode workaround for localhost testing
+        - https://www.nango.dev/blog/oauth-redirects-on-localhost-with-https
+        """
+
+        return cls._generate_oauth_url_helper(cls.DEV_REDIRECT_URI, state)
+
+    @classmethod
+    def _generate_oauth_url_helper(cls, redirect_uri: str, state: str) -> str:
+        url = (
+            f"https://slack.com/oauth/v2/authorize"
+            f"?client_id={cls.CLIENT_ID}"
+            f"&redirect_uri={redirect_uri}"
+            f"&scope={cls.BOT_SCOPE}"
+            f"&state={state}"
+        )
+        return url
+
+    @classmethod
+    def session_dump_json(cls, email: str, redirect_on_success: str | None) -> str:
+        """Temporary state to store in redis. to be looked up on auth response.
+        Returns a json string.
+        """
+        session = SlackOAuth.OAuthSession(
+            email=email, redirect_on_success=redirect_on_success
+        )
+        return session.model_dump_json()
+
+    @classmethod
+    def parse_session(cls, session_json: str) -> OAuthSession:
+        session = SlackOAuth.OAuthSession.model_validate_json(session_json)
+        return session
+
+
+class ConfluenceCloudOAuth:
+    """work in progress"""
+
+    # https://developer.atlassian.com/cloud/confluence/oauth-2-3lo-apps/
+
+    class OAuthSession(BaseModel):
+        """Stored in redis to be looked up on callback"""
+
+        email: str
+        redirect_on_success: str | None  # Where to send the user if OAuth flow succeeds
+
+    CLIENT_ID = OAUTH_CONFLUENCE_CLIENT_ID
+    CLIENT_SECRET = OAUTH_CONFLUENCE_CLIENT_SECRET
+    TOKEN_URL = "https://auth.atlassian.com/oauth/token"
+
+    # All read scopes per https://developer.atlassian.com/cloud/confluence/scopes-for-oauth-2-3LO-and-forge-apps/
+    CONFLUENCE_OAUTH_SCOPE = (
+        "read:confluence-props%20"
+        "read:confluence-content.all%20"
+        "read:confluence-content.summary%20"
+        "read:confluence-content.permission%20"
+        "read:confluence-user%20"
+        "read:confluence-groups%20"
+        "readonly:content.attachment:confluence"
+    )
+
+    REDIRECT_URI = f"{WEB_DOMAIN}/admin/connectors/confluence/oauth/callback"
+    DEV_REDIRECT_URI = f"https://redirectmeto.com/{REDIRECT_URI}"
+
+    # eventually for Confluence Data Center
+    # oauth_url = (
+    #     f"http://localhost:8090/rest/oauth/v2/authorize?client_id={CONFLUENCE_OAUTH_CLIENT_ID}"
+    #     f"&scope={CONFLUENCE_OAUTH_SCOPE_2}"
+    #     f"&redirect_uri={redirectme_uri}"
+    # )
+
+    @classmethod
+    def generate_oauth_url(cls, state: str) -> str:
+        return cls._generate_oauth_url_helper(cls.REDIRECT_URI, state)
+
+    @classmethod
+    def generate_dev_oauth_url(cls, state: str) -> str:
+        """dev mode workaround for localhost testing
+        - https://www.nango.dev/blog/oauth-redirects-on-localhost-with-https
+        """
+        return cls._generate_oauth_url_helper(cls.DEV_REDIRECT_URI, state)
+
+    @classmethod
+    def _generate_oauth_url_helper(cls, redirect_uri: str, state: str) -> str:
+        url = (
+            "https://auth.atlassian.com/authorize"
+            f"?audience=api.atlassian.com"
+            f"&client_id={cls.CLIENT_ID}"
+            f"&redirect_uri={redirect_uri}"
+            f"&scope={cls.CONFLUENCE_OAUTH_SCOPE}"
+            f"&state={state}"
+            "&response_type=code"
+            "&prompt=consent"
+        )
+        return url
+
+    @classmethod
+    def session_dump_json(cls, email: str, redirect_on_success: str | None) -> str:
+        """Temporary state to store in redis. to be looked up on auth response.
+        Returns a json string.
+        """
+        session = ConfluenceCloudOAuth.OAuthSession(
+            email=email, redirect_on_success=redirect_on_success
+        )
+        return session.model_dump_json()
+
+    @classmethod
+    def parse_session(cls, session_json: str) -> SlackOAuth.OAuthSession:
+        session = SlackOAuth.OAuthSession.model_validate_json(session_json)
+        return session
+
+
+class GoogleDriveOAuth:
+    # https://developers.google.com/identity/protocols/oauth2
+    # https://developers.google.com/identity/protocols/oauth2/web-server
+
+    class OAuthSession(BaseModel):
+        """Stored in redis to be looked up on callback"""
+
+        email: str
+        redirect_on_success: str | None  # Where to send the user if OAuth flow succeeds
+
+    CLIENT_ID = OAUTH_GOOGLE_DRIVE_CLIENT_ID
+    CLIENT_SECRET = OAUTH_GOOGLE_DRIVE_CLIENT_SECRET
+
+    TOKEN_URL = "https://oauth2.googleapis.com/token"
+
+    # SCOPE is per https://docs.onyx.app/connectors/google-drive
+    # TODO: Merge with or use google_utils.GOOGLE_SCOPES
+    SCOPE = (
+        "https://www.googleapis.com/auth/drive.readonly%20"
+        "https://www.googleapis.com/auth/drive.metadata.readonly%20"
+        "https://www.googleapis.com/auth/admin.directory.user.readonly%20"
+        "https://www.googleapis.com/auth/admin.directory.group.readonly"
+    )
+
+    REDIRECT_URI = f"{WEB_DOMAIN}/admin/connectors/google-drive/oauth/callback"
+    DEV_REDIRECT_URI = f"https://redirectmeto.com/{REDIRECT_URI}"
+
+    @classmethod
+    def generate_oauth_url(cls, state: str) -> str:
+        return cls._generate_oauth_url_helper(cls.REDIRECT_URI, state)
+
+    @classmethod
+    def generate_dev_oauth_url(cls, state: str) -> str:
+        """dev mode workaround for localhost testing
+        - https://www.nango.dev/blog/oauth-redirects-on-localhost-with-https
+        """
+
+        return cls._generate_oauth_url_helper(cls.DEV_REDIRECT_URI, state)
+
+    @classmethod
+    def _generate_oauth_url_helper(cls, redirect_uri: str, state: str) -> str:
+        # without prompt=consent, a refresh token is only issued the first time the user approves
+        url = (
+            f"https://accounts.google.com/o/oauth2/v2/auth"
+            f"?client_id={cls.CLIENT_ID}"
+            f"&redirect_uri={redirect_uri}"
+            "&response_type=code"
+            f"&scope={cls.SCOPE}"
+            "&access_type=offline"
+            f"&state={state}"
+            "&prompt=consent"
+        )
+        return url
+
+    @classmethod
+    def session_dump_json(cls, email: str, redirect_on_success: str | None) -> str:
+        """Temporary state to store in redis. to be looked up on auth response.
+        Returns a json string.
+        """
+        session = GoogleDriveOAuth.OAuthSession(
+            email=email, redirect_on_success=redirect_on_success
+        )
+        return session.model_dump_json()
+
+    @classmethod
+    def parse_session(cls, session_json: str) -> OAuthSession:
+        session = GoogleDriveOAuth.OAuthSession.model_validate_json(session_json)
+        return session
+
+
+@router.post("/prepare-authorization-request")
+def prepare_authorization_request(
+    connector: DocumentSource,
+    redirect_on_success: str | None,
+    user: User = Depends(current_user),
+) -> JSONResponse:
+    """Used by the frontend to generate the url for the user's browser during auth request.
+
+    Example: https://www.oauth.com/oauth2-servers/authorization/the-authorization-request/
+    """
+    tenant_id = get_current_tenant_id()
+
+    # create random oauth state param for security and to retrieve user data later
+    oauth_uuid = uuid.uuid4()
+    oauth_uuid_str = str(oauth_uuid)
+
+    # urlsafe b64 encode the uuid for the oauth url
+    oauth_state = (
+        base64.urlsafe_b64encode(oauth_uuid.bytes).rstrip(b"=").decode("utf-8")
+    )
+    session: str
+
+    if connector == DocumentSource.SLACK:
+        oauth_url = SlackOAuth.generate_oauth_url(oauth_state)
+        session = SlackOAuth.session_dump_json(
+            email=user.email, redirect_on_success=redirect_on_success
+        )
+    elif connector == DocumentSource.GOOGLE_DRIVE:
+        oauth_url = GoogleDriveOAuth.generate_oauth_url(oauth_state)
+        session = GoogleDriveOAuth.session_dump_json(
+            email=user.email, redirect_on_success=redirect_on_success
+        )
+    # elif connector == DocumentSource.CONFLUENCE:
+    #     oauth_url = ConfluenceCloudOAuth.generate_oauth_url(oauth_state)
+    #     session = ConfluenceCloudOAuth.session_dump_json(
+    #         email=user.email, redirect_on_success=redirect_on_success
+    #     )
+    # elif connector == DocumentSource.JIRA:
+    #     oauth_url = JiraCloudOAuth.generate_dev_oauth_url(oauth_state)
+    else:
+        oauth_url = None
+
+    if not oauth_url:
+        raise HTTPException(
+            status_code=404,
+            detail=f"The document source type {connector} does not have OAuth implemented",
+        )
+
+    r = get_redis_client(tenant_id=tenant_id)
+
+    # store important session state to retrieve when the user is redirected back
+    # 10 min is the max we want an oauth flow to be valid
+    r.set(f"da_oauth:{oauth_uuid_str}", session, ex=600)
+
+    return JSONResponse(content={"url": oauth_url})
+
+
+@router.post("/connector/slack/callback")
+def handle_slack_oauth_callback(
+    code: str,
+    state: str,
+    user: User = Depends(current_user),
+    db_session: Session = Depends(get_session),
+) -> JSONResponse:
+    if not SlackOAuth.CLIENT_ID or not SlackOAuth.CLIENT_SECRET:
+        raise HTTPException(
+            status_code=500,
+            detail="Slack client ID or client secret is not configured.",
+        )
+
+    r = get_redis_client()
+
+    # recover the state
+    padded_state = state + "=" * (
+        -len(state) % 4
+    )  # Add padding back (Base64 decoding requires padding)
+    uuid_bytes = base64.urlsafe_b64decode(
+        padded_state
+    )  # Decode the Base64 string back to bytes
+
+    # Convert bytes back to a UUID
+    oauth_uuid = uuid.UUID(bytes=uuid_bytes)
+    oauth_uuid_str = str(oauth_uuid)
+
+    r_key = f"da_oauth:{oauth_uuid_str}"
+
+    session_json_bytes = cast(bytes, r.get(r_key))
+    if not session_json_bytes:
+        raise HTTPException(
+            status_code=400,
+            detail=f"Slack OAuth failed - OAuth state key not found: key={r_key}",
+        )
+
+    session_json = session_json_bytes.decode("utf-8")
+    try:
+        session = SlackOAuth.parse_session(session_json)
+
+        # Exchange the authorization code for an access token
+        response = requests.post(
+            SlackOAuth.TOKEN_URL,
+            headers={"Content-Type": "application/x-www-form-urlencoded"},
+            data={
+                "client_id": SlackOAuth.CLIENT_ID,
+                "client_secret": SlackOAuth.CLIENT_SECRET,
+                "code": code,
+                "redirect_uri": SlackOAuth.REDIRECT_URI,
+            },
+        )
+
+        response_data = response.json()
+
+        if not response_data.get("ok"):
+            raise HTTPException(
+                status_code=400,
+                detail=f"Slack OAuth failed: {response_data.get('error')}",
+            )
+
+        # Extract token and team information
+        access_token: str = response_data.get("access_token")
+        team_id: str = response_data.get("team", {}).get("id")
+        authed_user_id: str = response_data.get("authed_user", {}).get("id")
+
+        credential_info = CredentialBase(
+            credential_json={"slack_bot_token": access_token},
+            admin_public=True,
+            source=DocumentSource.SLACK,
+            name="Slack OAuth",
+        )
+
+        create_credential(credential_info, user, db_session)
+    except Exception as e:
+        return JSONResponse(
+            status_code=500,
+            content={
+                "success": False,
+                "message": f"An error occurred during Slack OAuth: {str(e)}",
+            },
+        )
+    finally:
+        r.delete(r_key)
+
+    # return the result
+    return JSONResponse(
+        content={
+            "success": True,
+            "message": "Slack OAuth completed successfully.",
+            "team_id": team_id,
+            "authed_user_id": authed_user_id,
+            "redirect_on_success": session.redirect_on_success,
+        }
+    )
+
+
+# Work in progress
+# @router.post("/connector/confluence/callback")
+# def handle_confluence_oauth_callback(
+#     code: str,
+#     state: str,
+#     user: User = Depends(current_user),
+#     db_session: Session = Depends(get_session),
+#     tenant_id: str | None = Depends(get_current_tenant_id),
+# ) -> JSONResponse:
+#     if not ConfluenceCloudOAuth.CLIENT_ID or not ConfluenceCloudOAuth.CLIENT_SECRET:
+#         raise HTTPException(
+#             status_code=500,
+#             detail="Confluence client ID or client secret is not configured."
+#         )
+
+#     r = get_redis_client(tenant_id=tenant_id)
+
+#     # recover the state
+#     padded_state = state + '=' * (-len(state) % 4)  # Add padding back (Base64 decoding requires padding)
+#     uuid_bytes = base64.urlsafe_b64decode(padded_state)  # Decode the Base64 string back to bytes
+
+#     # Convert bytes back to a UUID
+#     oauth_uuid = uuid.UUID(bytes=uuid_bytes)
+#     oauth_uuid_str = str(oauth_uuid)
+
+#     r_key = f"da_oauth:{oauth_uuid_str}"
+
+#     result = r.get(r_key)
+#     if not result:
+#         raise HTTPException(
+#             status_code=400,
+#             detail=f"Confluence OAuth failed - OAuth state key not found: key={r_key}"
+#         )
+
+#     try:
+#         session = ConfluenceCloudOAuth.parse_session(result)
+
+#         # Exchange the authorization code for an access token
+#         response = requests.post(
+#             ConfluenceCloudOAuth.TOKEN_URL,
+#             headers={"Content-Type": "application/x-www-form-urlencoded"},
+#             data={
+#                 "client_id": ConfluenceCloudOAuth.CLIENT_ID,
+#                 "client_secret": ConfluenceCloudOAuth.CLIENT_SECRET,
+#                 "code": code,
+#                 "redirect_uri": ConfluenceCloudOAuth.DEV_REDIRECT_URI,
+#             },
+#         )
+
+#         response_data = response.json()
+
+#         if not response_data.get("ok"):
+#             raise HTTPException(
+#                 status_code=400,
+#                 detail=f"ConfluenceCloudOAuth OAuth failed: {response_data.get('error')}"
+#             )
+
+#         # Extract token and team information
+#         access_token: str = response_data.get("access_token")
+#         team_id: str = response_data.get("team", {}).get("id")
+#         authed_user_id: str = response_data.get("authed_user", {}).get("id")
+
+#         credential_info = CredentialBase(
+#             credential_json={"slack_bot_token": access_token},
+#             admin_public=True,
+#             source=DocumentSource.CONFLUENCE,
+#             name="Confluence OAuth",
+#         )
+
+#         logger.info(f"Slack access token: {access_token}")
+
+#         credential = create_credential(credential_info, user, db_session)
+
+#         logger.info(f"new_credential_id={credential.id}")
+#     except Exception as e:
+#         return JSONResponse(
+#             status_code=500,
+#             content={
+#                 "success": False,
+#                 "message": f"An error occurred during Slack OAuth: {str(e)}",
+#             },
+#         )
+#     finally:
+#         r.delete(r_key)
+
+#     # return the result
+#     return JSONResponse(
+#         content={
+#             "success": True,
+#             "message": "Slack OAuth completed successfully.",
+#             "team_id": team_id,
+#             "authed_user_id": authed_user_id,
+#             "redirect_on_success": session.redirect_on_success,
+#         }
+#     )
+
+
+@router.post("/connector/google-drive/callback")
+def handle_google_drive_oauth_callback(
+    code: str,
+    state: str,
+    user: User = Depends(current_user),
+    db_session: Session = Depends(get_session),
+) -> JSONResponse:
+    if not GoogleDriveOAuth.CLIENT_ID or not GoogleDriveOAuth.CLIENT_SECRET:
+        raise HTTPException(
+            status_code=500,
+            detail="Google Drive client ID or client secret is not configured.",
+        )
+
+    r = get_redis_client()
+
+    # recover the state
+    padded_state = state + "=" * (
+        -len(state) % 4
+    )  # Add padding back (Base64 decoding requires padding)
+    uuid_bytes = base64.urlsafe_b64decode(
+        padded_state
+    )  # Decode the Base64 string back to bytes
+
+    # Convert bytes back to a UUID
+    oauth_uuid = uuid.UUID(bytes=uuid_bytes)
+    oauth_uuid_str = str(oauth_uuid)
+
+    r_key = f"da_oauth:{oauth_uuid_str}"
+
+    session_json_bytes = cast(bytes, r.get(r_key))
+    if not session_json_bytes:
+        raise HTTPException(
+            status_code=400,
+            detail=f"Google Drive OAuth failed - OAuth state key not found: key={r_key}",
+        )
+
+    session_json = session_json_bytes.decode("utf-8")
+    session: GoogleDriveOAuth.OAuthSession
+    try:
+        session = GoogleDriveOAuth.parse_session(session_json)
+
+        # Exchange the authorization code for an access token
+        response = requests.post(
+            GoogleDriveOAuth.TOKEN_URL,
+            headers={"Content-Type": "application/x-www-form-urlencoded"},
+            data={
+                "client_id": GoogleDriveOAuth.CLIENT_ID,
+                "client_secret": GoogleDriveOAuth.CLIENT_SECRET,
+                "code": code,
+                "redirect_uri": GoogleDriveOAuth.REDIRECT_URI,
+                "grant_type": "authorization_code",
+            },
+        )
+
+        response.raise_for_status()
+
+        authorization_response: dict[str, Any] = response.json()
+
+        # the connector wants us to store the json in its authorized_user_info format
+        # returned from OAuthCredentials.get_authorized_user_info().
+        # So refresh immediately via get_google_oauth_creds with the params filled in
+        # from fields in authorization_response to get the json we need
+        authorized_user_info = {}
+        authorized_user_info["client_id"] = OAUTH_GOOGLE_DRIVE_CLIENT_ID
+        authorized_user_info["client_secret"] = OAUTH_GOOGLE_DRIVE_CLIENT_SECRET
+        authorized_user_info["refresh_token"] = authorization_response["refresh_token"]
+
+        token_json_str = json.dumps(authorized_user_info)
+        oauth_creds = get_google_oauth_creds(
+            token_json_str=token_json_str, source=DocumentSource.GOOGLE_DRIVE
+        )
+        if not oauth_creds:
+            raise RuntimeError("get_google_oauth_creds returned None.")
+
+        # save off the credentials
+        oauth_creds_sanitized_json_str = sanitize_oauth_credentials(oauth_creds)
+
+        credential_dict: dict[str, str] = {}
+        credential_dict[DB_CREDENTIALS_DICT_TOKEN_KEY] = oauth_creds_sanitized_json_str
+        credential_dict[DB_CREDENTIALS_PRIMARY_ADMIN_KEY] = session.email
+        credential_dict[
+            DB_CREDENTIALS_AUTHENTICATION_METHOD
+        ] = GoogleOAuthAuthenticationMethod.OAUTH_INTERACTIVE.value
+
+        credential_info = CredentialBase(
+            credential_json=credential_dict,
+            admin_public=True,
+            source=DocumentSource.GOOGLE_DRIVE,
+            name="OAuth (interactive)",
+        )
+
+        create_credential(credential_info, user, db_session)
+    except Exception as e:
+        return JSONResponse(
+            status_code=500,
+            content={
+                "success": False,
+                "message": f"An error occurred during Google Drive OAuth: {str(e)}",
+            },
+        )
+    finally:
+        r.delete(r_key)
+
+    # return the result
+    return JSONResponse(
+        content={
+            "success": True,
+            "message": "Google Drive OAuth completed successfully.",
+            "redirect_on_success": session.redirect_on_success,
+        }
+    )

backend/ee/onyx/server/oauth/api.py

@@ -1,91 +0,0 @@
-import base64
-import uuid
-
-from fastapi import Depends
-from fastapi import HTTPException
-from fastapi.responses import JSONResponse
-
-from ee.onyx.server.oauth.api_router import router
-from ee.onyx.server.oauth.confluence_cloud import ConfluenceCloudOAuth
-from ee.onyx.server.oauth.google_drive import GoogleDriveOAuth
-from ee.onyx.server.oauth.slack import SlackOAuth
-from onyx.auth.users import current_admin_user
-from onyx.configs.app_configs import DEV_MODE
-from onyx.configs.constants import DocumentSource
-from onyx.db.engine import get_current_tenant_id
-from onyx.db.models import User
-from onyx.redis.redis_pool import get_redis_client
-from onyx.utils.logger import setup_logger
-
-logger = setup_logger()
-
-
-@router.post("/prepare-authorization-request")
-def prepare_authorization_request(
-    connector: DocumentSource,
-    redirect_on_success: str | None,
-    user: User = Depends(current_admin_user),
-    tenant_id: str | None = Depends(get_current_tenant_id),
-) -> JSONResponse:
-    """Used by the frontend to generate the url for the user's browser during auth request.
-
-    Example: https://www.oauth.com/oauth2-servers/authorization/the-authorization-request/
-    """
-
-    # create random oauth state param for security and to retrieve user data later
-    oauth_uuid = uuid.uuid4()
-    oauth_uuid_str = str(oauth_uuid)
-
-    # urlsafe b64 encode the uuid for the oauth url
-    oauth_state = (
-        base64.urlsafe_b64encode(oauth_uuid.bytes).rstrip(b"=").decode("utf-8")
-    )
-
-    session: str | None = None
-    if connector == DocumentSource.SLACK:
-        if not DEV_MODE:
-            oauth_url = SlackOAuth.generate_oauth_url(oauth_state)
-        else:
-            oauth_url = SlackOAuth.generate_dev_oauth_url(oauth_state)
-
-        session = SlackOAuth.session_dump_json(
-            email=user.email, redirect_on_success=redirect_on_success
-        )
-    elif connector == DocumentSource.CONFLUENCE:
-        if not DEV_MODE:
-            oauth_url = ConfluenceCloudOAuth.generate_oauth_url(oauth_state)
-        else:
-            oauth_url = ConfluenceCloudOAuth.generate_dev_oauth_url(oauth_state)
-        session = ConfluenceCloudOAuth.session_dump_json(
-            email=user.email, redirect_on_success=redirect_on_success
-        )
-    elif connector == DocumentSource.GOOGLE_DRIVE:
-        if not DEV_MODE:
-            oauth_url = GoogleDriveOAuth.generate_oauth_url(oauth_state)
-        else:
-            oauth_url = GoogleDriveOAuth.generate_dev_oauth_url(oauth_state)
-        session = GoogleDriveOAuth.session_dump_json(
-            email=user.email, redirect_on_success=redirect_on_success
-        )
-    else:
-        oauth_url = None
-
-    if not oauth_url:
-        raise HTTPException(
-            status_code=404,
-            detail=f"The document source type {connector} does not have OAuth implemented",
-        )
-
-    if not session:
-        raise HTTPException(
-            status_code=500,
-            detail=f"The document source type {connector} failed to generate an OAuth session.",
-        )
-
-    r = get_redis_client(tenant_id=tenant_id)
-
-    # store important session state to retrieve when the user is redirected back
-    # 10 min is the max we want an oauth flow to be valid
-    r.set(f"da_oauth:{oauth_uuid_str}", session, ex=600)
-
-    return JSONResponse(content={"url": oauth_url})

backend/ee/onyx/server/oauth/api_router.py

@@ -1,3 +0,0 @@
-from fastapi import APIRouter
-
-router: APIRouter = APIRouter(prefix="/oauth")

backend/ee/onyx/server/oauth/confluence_cloud.py

@@ -1,362 +0,0 @@
-import base64
-import uuid
-from datetime import datetime
-from datetime import timedelta
-from datetime import timezone
-from typing import Any
-from typing import cast
-
-import requests
-from fastapi import Depends
-from fastapi import HTTPException
-from fastapi.responses import JSONResponse
-from pydantic import BaseModel
-from pydantic import ValidationError
-from sqlalchemy.orm import Session
-
-from ee.onyx.configs.app_configs import OAUTH_CONFLUENCE_CLOUD_CLIENT_ID
-from ee.onyx.configs.app_configs import OAUTH_CONFLUENCE_CLOUD_CLIENT_SECRET
-from ee.onyx.server.oauth.api_router import router
-from onyx.auth.users import current_admin_user
-from onyx.configs.app_configs import DEV_MODE
-from onyx.configs.app_configs import WEB_DOMAIN
-from onyx.configs.constants import DocumentSource
-from onyx.connectors.confluence.utils import CONFLUENCE_OAUTH_TOKEN_URL
-from onyx.db.credentials import create_credential
-from onyx.db.credentials import fetch_credential_by_id_for_user
-from onyx.db.credentials import update_credential_json
-from onyx.db.engine import get_current_tenant_id
-from onyx.db.engine import get_session
-from onyx.db.models import User
-from onyx.redis.redis_pool import get_redis_client
-from onyx.server.documents.models import CredentialBase
-from onyx.utils.logger import setup_logger
-
-logger = setup_logger()
-
-
-class ConfluenceCloudOAuth:
-    # https://developer.atlassian.com/cloud/confluence/oauth-2-3lo-apps/
-
-    class OAuthSession(BaseModel):
-        """Stored in redis to be looked up on callback"""
-
-        email: str
-        redirect_on_success: str | None  # Where to send the user if OAuth flow succeeds
-
-    class TokenResponse(BaseModel):
-        access_token: str
-        expires_in: int
-        token_type: str
-        refresh_token: str
-        scope: str
-
-    class AccessibleResources(BaseModel):
-        id: str
-        name: str
-        url: str
-        scopes: list[str]
-        avatarUrl: str
-
-    CLIENT_ID = OAUTH_CONFLUENCE_CLOUD_CLIENT_ID
-    CLIENT_SECRET = OAUTH_CONFLUENCE_CLOUD_CLIENT_SECRET
-    TOKEN_URL = CONFLUENCE_OAUTH_TOKEN_URL
-
-    ACCESSIBLE_RESOURCE_URL = (
-        "https://api.atlassian.com/oauth/token/accessible-resources"
-    )
-
-    # All read scopes per https://developer.atlassian.com/cloud/confluence/scopes-for-oauth-2-3LO-and-forge-apps/
-    CONFLUENCE_OAUTH_SCOPE = (
-        # classic scope
-        "read:confluence-space.summary%20"
-        "read:confluence-props%20"
-        "read:confluence-content.all%20"
-        "read:confluence-content.summary%20"
-        "read:confluence-content.permission%20"
-        "read:confluence-user%20"
-        "read:confluence-groups%20"
-        "readonly:content.attachment:confluence%20"
-        "search:confluence%20"
-        # granular scope
-        "read:attachment:confluence%20"  # possibly unneeded unless calling v2 attachments api
-        "read:content-details:confluence%20"  # for permission sync
-        "offline_access"
-    )
-
-    REDIRECT_URI = f"{WEB_DOMAIN}/admin/connectors/confluence/oauth/callback"
-    DEV_REDIRECT_URI = f"https://redirectmeto.com/{REDIRECT_URI}"
-
-    # eventually for Confluence Data Center
-    # oauth_url = (
-    #     f"http://localhost:8090/rest/oauth/v2/authorize?client_id={CONFLUENCE_OAUTH_CLIENT_ID}"
-    #     f"&scope={CONFLUENCE_OAUTH_SCOPE_2}"
-    #     f"&redirect_uri={redirectme_uri}"
-    # )
-
-    @classmethod
-    def generate_oauth_url(cls, state: str) -> str:
-        return cls._generate_oauth_url_helper(cls.REDIRECT_URI, state)
-
-    @classmethod
-    def generate_dev_oauth_url(cls, state: str) -> str:
-        """dev mode workaround for localhost testing
-        - https://www.nango.dev/blog/oauth-redirects-on-localhost-with-https
-        """
-        return cls._generate_oauth_url_helper(cls.DEV_REDIRECT_URI, state)
-
-    @classmethod
-    def _generate_oauth_url_helper(cls, redirect_uri: str, state: str) -> str:
-        # https://developer.atlassian.com/cloud/jira/platform/oauth-2-3lo-apps/#1--direct-the-user-to-the-authorization-url-to-get-an-authorization-code
-
-        url = (
-            "https://auth.atlassian.com/authorize"
-            f"?audience=api.atlassian.com"
-            f"&client_id={cls.CLIENT_ID}"
-            f"&scope={cls.CONFLUENCE_OAUTH_SCOPE}"
-            f"&redirect_uri={redirect_uri}"
-            f"&state={state}"
-            "&response_type=code"
-            "&prompt=consent"
-        )
-        return url
-
-    @classmethod
-    def session_dump_json(cls, email: str, redirect_on_success: str | None) -> str:
-        """Temporary state to store in redis. to be looked up on auth response.
-        Returns a json string.
-        """
-        session = ConfluenceCloudOAuth.OAuthSession(
-            email=email, redirect_on_success=redirect_on_success
-        )
-        return session.model_dump_json()
-
-    @classmethod
-    def parse_session(cls, session_json: str) -> OAuthSession:
-        session = ConfluenceCloudOAuth.OAuthSession.model_validate_json(session_json)
-        return session
-
-    @classmethod
-    def generate_finalize_url(cls, credential_id: int) -> str:
-        return f"{WEB_DOMAIN}/admin/connectors/confluence/oauth/finalize?credential={credential_id}"
-
-
-@router.post("/connector/confluence/callback")
-def confluence_oauth_callback(
-    code: str,
-    state: str,
-    user: User = Depends(current_admin_user),
-    db_session: Session = Depends(get_session),
-    tenant_id: str | None = Depends(get_current_tenant_id),
-) -> JSONResponse:
-    """Handles the backend logic for the frontend page that the user is redirected to
-    after visiting the oauth authorization url."""
-
-    if not ConfluenceCloudOAuth.CLIENT_ID or not ConfluenceCloudOAuth.CLIENT_SECRET:
-        raise HTTPException(
-            status_code=500,
-            detail="Confluence Cloud client ID or client secret is not configured.",
-        )
-
-    r = get_redis_client(tenant_id=tenant_id)
-
-    # recover the state
-    padded_state = state + "=" * (
-        -len(state) % 4
-    )  # Add padding back (Base64 decoding requires padding)
-    uuid_bytes = base64.urlsafe_b64decode(
-        padded_state
-    )  # Decode the Base64 string back to bytes
-
-    # Convert bytes back to a UUID
-    oauth_uuid = uuid.UUID(bytes=uuid_bytes)
-    oauth_uuid_str = str(oauth_uuid)
-
-    r_key = f"da_oauth:{oauth_uuid_str}"
-
-    session_json_bytes = cast(bytes, r.get(r_key))
-    if not session_json_bytes:
-        raise HTTPException(
-            status_code=400,
-            detail=f"Confluence Cloud OAuth failed - OAuth state key not found: key={r_key}",
-        )
-
-    session_json = session_json_bytes.decode("utf-8")
-    try:
-        session = ConfluenceCloudOAuth.parse_session(session_json)
-
-        if not DEV_MODE:
-            redirect_uri = ConfluenceCloudOAuth.REDIRECT_URI
-        else:
-            redirect_uri = ConfluenceCloudOAuth.DEV_REDIRECT_URI
-
-        # Exchange the authorization code for an access token
-        response = requests.post(
-            ConfluenceCloudOAuth.TOKEN_URL,
-            headers={"Content-Type": "application/x-www-form-urlencoded"},
-            data={
-                "client_id": ConfluenceCloudOAuth.CLIENT_ID,
-                "client_secret": ConfluenceCloudOAuth.CLIENT_SECRET,
-                "code": code,
-                "redirect_uri": redirect_uri,
-                "grant_type": "authorization_code",
-            },
-        )
-
-        token_response: ConfluenceCloudOAuth.TokenResponse | None = None
-
-        try:
-            token_response = ConfluenceCloudOAuth.TokenResponse.model_validate_json(
-                response.text
-            )
-        except Exception:
-            raise RuntimeError(
-                "Confluence Cloud OAuth failed during code/token exchange."
-            )
-
-        now = datetime.now(timezone.utc)
-        expires_at = now + timedelta(seconds=token_response.expires_in)
-
-        credential_info = CredentialBase(
-            credential_json={
-                "confluence_access_token": token_response.access_token,
-                "confluence_refresh_token": token_response.refresh_token,
-                "created_at": now.isoformat(),
-                "expires_at": expires_at.isoformat(),
-                "expires_in": token_response.expires_in,
-                "scope": token_response.scope,
-            },
-            admin_public=True,
-            source=DocumentSource.CONFLUENCE,
-            name="Confluence Cloud OAuth",
-        )
-
-        credential = create_credential(credential_info, user, db_session)
-    except Exception as e:
-        return JSONResponse(
-            status_code=500,
-            content={
-                "success": False,
-                "message": f"An error occurred during Confluence Cloud OAuth: {str(e)}",
-            },
-        )
-    finally:
-        r.delete(r_key)
-
-    # return the result
-    return JSONResponse(
-        content={
-            "success": True,
-            "message": "Confluence Cloud OAuth completed successfully.",
-            "finalize_url": ConfluenceCloudOAuth.generate_finalize_url(credential.id),
-            "redirect_on_success": session.redirect_on_success,
-        }
-    )
-
-
-@router.get("/connector/confluence/accessible-resources")
-def confluence_oauth_accessible_resources(
-    credential_id: int,
-    user: User = Depends(current_admin_user),
-    db_session: Session = Depends(get_session),
-    tenant_id: str | None = Depends(get_current_tenant_id),
-) -> JSONResponse:
-    """Atlassian's API is weird and does not supply us with enough info to be in a
-    usable state after authorizing.  All API's require a cloud id. We have to list
-    the accessible resources/sites and let the user choose which site to use."""
-
-    credential = fetch_credential_by_id_for_user(credential_id, user, db_session)
-    if not credential:
-        raise HTTPException(400, f"Credential {credential_id} not found.")
-
-    credential_dict = credential.credential_json
-    access_token = credential_dict["confluence_access_token"]
-
-    try:
-        # Exchange the authorization code for an access token
-        response = requests.get(
-            ConfluenceCloudOAuth.ACCESSIBLE_RESOURCE_URL,
-            headers={
-                "Authorization": f"Bearer {access_token}",
-                "Accept": "application/json",
-            },
-        )
-
-        response.raise_for_status()
-        accessible_resources_data = response.json()
-
-        # Validate the list of AccessibleResources
-        try:
-            accessible_resources = [
-                ConfluenceCloudOAuth.AccessibleResources(**resource)
-                for resource in accessible_resources_data
-            ]
-        except ValidationError as e:
-            raise RuntimeError(f"Failed to parse accessible resources: {e}")
-    except Exception as e:
-        return JSONResponse(
-            status_code=500,
-            content={
-                "success": False,
-                "message": f"An error occurred retrieving Confluence Cloud accessible resources: {str(e)}",
-            },
-        )
-
-    # return the result
-    return JSONResponse(
-        content={
-            "success": True,
-            "message": "Confluence Cloud get accessible resources completed successfully.",
-            "accessible_resources": [
-                resource.model_dump() for resource in accessible_resources
-            ],
-        }
-    )
-
-
-@router.post("/connector/confluence/finalize")
-def confluence_oauth_finalize(
-    credential_id: int,
-    cloud_id: str,
-    cloud_name: str,
-    cloud_url: str,
-    user: User = Depends(current_admin_user),
-    db_session: Session = Depends(get_session),
-    tenant_id: str | None = Depends(get_current_tenant_id),
-) -> JSONResponse:
-    """Saves the info for the selected cloud site to the credential.
-    This is the final step in the confluence oauth flow where after the traditional
-    OAuth process, the user has to select a site to associate with the credentials.
-    After this, the credential is usable."""
-
-    credential = fetch_credential_by_id_for_user(credential_id, user, db_session)
-    if not credential:
-        raise HTTPException(
-            status_code=400,
-            detail=f"Confluence Cloud OAuth failed - credential {credential_id} not found.",
-        )
-
-    new_credential_json: dict[str, Any] = dict(credential.credential_json)
-    new_credential_json["cloud_id"] = cloud_id
-    new_credential_json["cloud_name"] = cloud_name
-    new_credential_json["wiki_base"] = cloud_url
-
-    try:
-        update_credential_json(credential_id, new_credential_json, user, db_session)
-    except Exception as e:
-        return JSONResponse(
-            status_code=500,
-            content={
-                "success": False,
-                "message": f"An error occurred during Confluence Cloud OAuth: {str(e)}",
-            },
-        )
-
-    # return the result
-    return JSONResponse(
-        content={
-            "success": True,
-            "message": "Confluence Cloud OAuth finalized successfully.",
-            "redirect_url": f"{WEB_DOMAIN}/admin/connectors/confluence",
-        }
-    )

backend/ee/onyx/server/oauth/google_drive.py

@@ -1,229 +0,0 @@
-import base64
-import json
-import uuid
-from typing import Any
-from typing import cast
-
-import requests
-from fastapi import Depends
-from fastapi import HTTPException
-from fastapi.responses import JSONResponse
-from pydantic import BaseModel
-from sqlalchemy.orm import Session
-
-from ee.onyx.configs.app_configs import OAUTH_GOOGLE_DRIVE_CLIENT_ID
-from ee.onyx.configs.app_configs import OAUTH_GOOGLE_DRIVE_CLIENT_SECRET
-from ee.onyx.server.oauth.api_router import router
-from onyx.auth.users import current_admin_user
-from onyx.configs.app_configs import DEV_MODE
-from onyx.configs.app_configs import WEB_DOMAIN
-from onyx.configs.constants import DocumentSource
-from onyx.connectors.google_utils.google_auth import get_google_oauth_creds
-from onyx.connectors.google_utils.google_auth import sanitize_oauth_credentials
-from onyx.connectors.google_utils.shared_constants import (
-    DB_CREDENTIALS_AUTHENTICATION_METHOD,
-)
-from onyx.connectors.google_utils.shared_constants import (
-    DB_CREDENTIALS_DICT_TOKEN_KEY,
-)
-from onyx.connectors.google_utils.shared_constants import (
-    DB_CREDENTIALS_PRIMARY_ADMIN_KEY,
-)
-from onyx.connectors.google_utils.shared_constants import (
-    GoogleOAuthAuthenticationMethod,
-)
-from onyx.db.credentials import create_credential
-from onyx.db.engine import get_current_tenant_id
-from onyx.db.engine import get_session
-from onyx.db.models import User
-from onyx.redis.redis_pool import get_redis_client
-from onyx.server.documents.models import CredentialBase
-
-
-class GoogleDriveOAuth:
-    # https://developers.google.com/identity/protocols/oauth2
-    # https://developers.google.com/identity/protocols/oauth2/web-server
-
-    class OAuthSession(BaseModel):
-        """Stored in redis to be looked up on callback"""
-
-        email: str
-        redirect_on_success: str | None  # Where to send the user if OAuth flow succeeds
-
-    CLIENT_ID = OAUTH_GOOGLE_DRIVE_CLIENT_ID
-    CLIENT_SECRET = OAUTH_GOOGLE_DRIVE_CLIENT_SECRET
-
-    TOKEN_URL = "https://oauth2.googleapis.com/token"
-
-    # SCOPE is per https://docs.danswer.dev/connectors/google-drive
-    # TODO: Merge with or use google_utils.GOOGLE_SCOPES
-    SCOPE = (
-        "https://www.googleapis.com/auth/drive.readonly%20"
-        "https://www.googleapis.com/auth/drive.metadata.readonly%20"
-        "https://www.googleapis.com/auth/admin.directory.user.readonly%20"
-        "https://www.googleapis.com/auth/admin.directory.group.readonly"
-    )
-
-    REDIRECT_URI = f"{WEB_DOMAIN}/admin/connectors/google-drive/oauth/callback"
-    DEV_REDIRECT_URI = f"https://redirectmeto.com/{REDIRECT_URI}"
-
-    @classmethod
-    def generate_oauth_url(cls, state: str) -> str:
-        return cls._generate_oauth_url_helper(cls.REDIRECT_URI, state)
-
-    @classmethod
-    def generate_dev_oauth_url(cls, state: str) -> str:
-        """dev mode workaround for localhost testing
-        - https://www.nango.dev/blog/oauth-redirects-on-localhost-with-https
-        """
-
-        return cls._generate_oauth_url_helper(cls.DEV_REDIRECT_URI, state)
-
-    @classmethod
-    def _generate_oauth_url_helper(cls, redirect_uri: str, state: str) -> str:
-        # without prompt=consent, a refresh token is only issued the first time the user approves
-        url = (
-            f"https://accounts.google.com/o/oauth2/v2/auth"
-            f"?client_id={cls.CLIENT_ID}"
-            f"&redirect_uri={redirect_uri}"
-            "&response_type=code"
-            f"&scope={cls.SCOPE}"
-            "&access_type=offline"
-            f"&state={state}"
-            "&prompt=consent"
-        )
-        return url
-
-    @classmethod
-    def session_dump_json(cls, email: str, redirect_on_success: str | None) -> str:
-        """Temporary state to store in redis. to be looked up on auth response.
-        Returns a json string.
-        """
-        session = GoogleDriveOAuth.OAuthSession(
-            email=email, redirect_on_success=redirect_on_success
-        )
-        return session.model_dump_json()
-
-    @classmethod
-    def parse_session(cls, session_json: str) -> OAuthSession:
-        session = GoogleDriveOAuth.OAuthSession.model_validate_json(session_json)
-        return session
-
-
-@router.post("/connector/google-drive/callback")
-def handle_google_drive_oauth_callback(
-    code: str,
-    state: str,
-    user: User = Depends(current_admin_user),
-    db_session: Session = Depends(get_session),
-    tenant_id: str | None = Depends(get_current_tenant_id),
-) -> JSONResponse:
-    if not GoogleDriveOAuth.CLIENT_ID or not GoogleDriveOAuth.CLIENT_SECRET:
-        raise HTTPException(
-            status_code=500,
-            detail="Google Drive client ID or client secret is not configured.",
-        )
-
-    r = get_redis_client(tenant_id=tenant_id)
-
-    # recover the state
-    padded_state = state + "=" * (
-        -len(state) % 4
-    )  # Add padding back (Base64 decoding requires padding)
-    uuid_bytes = base64.urlsafe_b64decode(
-        padded_state
-    )  # Decode the Base64 string back to bytes
-
-    # Convert bytes back to a UUID
-    oauth_uuid = uuid.UUID(bytes=uuid_bytes)
-    oauth_uuid_str = str(oauth_uuid)
-
-    r_key = f"da_oauth:{oauth_uuid_str}"
-
-    session_json_bytes = cast(bytes, r.get(r_key))
-    if not session_json_bytes:
-        raise HTTPException(
-            status_code=400,
-            detail=f"Google Drive OAuth failed - OAuth state key not found: key={r_key}",
-        )
-
-    session_json = session_json_bytes.decode("utf-8")
-    try:
-        session = GoogleDriveOAuth.parse_session(session_json)
-
-        if not DEV_MODE:
-            redirect_uri = GoogleDriveOAuth.REDIRECT_URI
-        else:
-            redirect_uri = GoogleDriveOAuth.DEV_REDIRECT_URI
-
-        # Exchange the authorization code for an access token
-        response = requests.post(
-            GoogleDriveOAuth.TOKEN_URL,
-            headers={"Content-Type": "application/x-www-form-urlencoded"},
-            data={
-                "client_id": GoogleDriveOAuth.CLIENT_ID,
-                "client_secret": GoogleDriveOAuth.CLIENT_SECRET,
-                "code": code,
-                "redirect_uri": redirect_uri,
-                "grant_type": "authorization_code",
-            },
-        )
-
-        response.raise_for_status()
-
-        authorization_response: dict[str, Any] = response.json()
-
-        # the connector wants us to store the json in its authorized_user_info format
-        # returned from OAuthCredentials.get_authorized_user_info().
-        # So refresh immediately via get_google_oauth_creds with the params filled in
-        # from fields in authorization_response to get the json we need
-        authorized_user_info = {}
-        authorized_user_info["client_id"] = OAUTH_GOOGLE_DRIVE_CLIENT_ID
-        authorized_user_info["client_secret"] = OAUTH_GOOGLE_DRIVE_CLIENT_SECRET
-        authorized_user_info["refresh_token"] = authorization_response["refresh_token"]
-
-        token_json_str = json.dumps(authorized_user_info)
-        oauth_creds = get_google_oauth_creds(
-            token_json_str=token_json_str, source=DocumentSource.GOOGLE_DRIVE
-        )
-        if not oauth_creds:
-            raise RuntimeError("get_google_oauth_creds returned None.")
-
-        # save off the credentials
-        oauth_creds_sanitized_json_str = sanitize_oauth_credentials(oauth_creds)
-
-        credential_dict: dict[str, str] = {}
-        credential_dict[DB_CREDENTIALS_DICT_TOKEN_KEY] = oauth_creds_sanitized_json_str
-        credential_dict[DB_CREDENTIALS_PRIMARY_ADMIN_KEY] = session.email
-        credential_dict[
-            DB_CREDENTIALS_AUTHENTICATION_METHOD
-        ] = GoogleOAuthAuthenticationMethod.OAUTH_INTERACTIVE.value
-
-        credential_info = CredentialBase(
-            credential_json=credential_dict,
-            admin_public=True,
-            source=DocumentSource.GOOGLE_DRIVE,
-            name="OAuth (interactive)",
-        )
-
-        create_credential(credential_info, user, db_session)
-    except Exception as e:
-        return JSONResponse(
-            status_code=500,
-            content={
-                "success": False,
-                "message": f"An error occurred during Google Drive OAuth: {str(e)}",
-            },
-        )
-    finally:
-        r.delete(r_key)
-
-    # return the result
-    return JSONResponse(
-        content={
-            "success": True,
-            "message": "Google Drive OAuth completed successfully.",
-            "finalize_url": None,
-            "redirect_on_success": session.redirect_on_success,
-        }
-    )

backend/ee/onyx/server/oauth/slack.py

@@ -1,197 +0,0 @@
-import base64
-import uuid
-from typing import cast
-
-import requests
-from fastapi import Depends
-from fastapi import HTTPException
-from fastapi.responses import JSONResponse
-from pydantic import BaseModel
-from sqlalchemy.orm import Session
-
-from ee.onyx.configs.app_configs import OAUTH_SLACK_CLIENT_ID
-from ee.onyx.configs.app_configs import OAUTH_SLACK_CLIENT_SECRET
-from ee.onyx.server.oauth.api_router import router
-from onyx.auth.users import current_admin_user
-from onyx.configs.app_configs import DEV_MODE
-from onyx.configs.app_configs import WEB_DOMAIN
-from onyx.configs.constants import DocumentSource
-from onyx.db.credentials import create_credential
-from onyx.db.engine import get_current_tenant_id
-from onyx.db.engine import get_session
-from onyx.db.models import User
-from onyx.redis.redis_pool import get_redis_client
-from onyx.server.documents.models import CredentialBase
-
-
-class SlackOAuth:
-    # https://knock.app/blog/how-to-authenticate-users-in-slack-using-oauth
-    # Example: https://api.slack.com/authentication/oauth-v2#exchanging
-
-    class OAuthSession(BaseModel):
-        """Stored in redis to be looked up on callback"""
-
-        email: str
-        redirect_on_success: str | None  # Where to send the user if OAuth flow succeeds
-
-    CLIENT_ID = OAUTH_SLACK_CLIENT_ID
-    CLIENT_SECRET = OAUTH_SLACK_CLIENT_SECRET
-
-    TOKEN_URL = "https://slack.com/api/oauth.v2.access"
-
-    # SCOPE is per https://docs.danswer.dev/connectors/slack
-    BOT_SCOPE = (
-        "channels:history,"
-        "channels:read,"
-        "groups:history,"
-        "groups:read,"
-        "channels:join,"
-        "im:history,"
-        "users:read,"
-        "users:read.email,"
-        "usergroups:read"
-    )
-
-    REDIRECT_URI = f"{WEB_DOMAIN}/admin/connectors/slack/oauth/callback"
-    DEV_REDIRECT_URI = f"https://redirectmeto.com/{REDIRECT_URI}"
-
-    @classmethod
-    def generate_oauth_url(cls, state: str) -> str:
-        return cls._generate_oauth_url_helper(cls.REDIRECT_URI, state)
-
-    @classmethod
-    def generate_dev_oauth_url(cls, state: str) -> str:
-        """dev mode workaround for localhost testing
-        - https://www.nango.dev/blog/oauth-redirects-on-localhost-with-https
-        """
-
-        return cls._generate_oauth_url_helper(cls.DEV_REDIRECT_URI, state)
-
-    @classmethod
-    def _generate_oauth_url_helper(cls, redirect_uri: str, state: str) -> str:
-        url = (
-            f"https://slack.com/oauth/v2/authorize"
-            f"?client_id={cls.CLIENT_ID}"
-            f"&redirect_uri={redirect_uri}"
-            f"&scope={cls.BOT_SCOPE}"
-            f"&state={state}"
-        )
-        return url
-
-    @classmethod
-    def session_dump_json(cls, email: str, redirect_on_success: str | None) -> str:
-        """Temporary state to store in redis. to be looked up on auth response.
-        Returns a json string.
-        """
-        session = SlackOAuth.OAuthSession(
-            email=email, redirect_on_success=redirect_on_success
-        )
-        return session.model_dump_json()
-
-    @classmethod
-    def parse_session(cls, session_json: str) -> OAuthSession:
-        session = SlackOAuth.OAuthSession.model_validate_json(session_json)
-        return session
-
-
-@router.post("/connector/slack/callback")
-def handle_slack_oauth_callback(
-    code: str,
-    state: str,
-    user: User = Depends(current_admin_user),
-    db_session: Session = Depends(get_session),
-    tenant_id: str | None = Depends(get_current_tenant_id),
-) -> JSONResponse:
-    if not SlackOAuth.CLIENT_ID or not SlackOAuth.CLIENT_SECRET:
-        raise HTTPException(
-            status_code=500,
-            detail="Slack client ID or client secret is not configured.",
-        )
-
-    r = get_redis_client(tenant_id=tenant_id)
-
-    # recover the state
-    padded_state = state + "=" * (
-        -len(state) % 4
-    )  # Add padding back (Base64 decoding requires padding)
-    uuid_bytes = base64.urlsafe_b64decode(
-        padded_state
-    )  # Decode the Base64 string back to bytes
-
-    # Convert bytes back to a UUID
-    oauth_uuid = uuid.UUID(bytes=uuid_bytes)
-    oauth_uuid_str = str(oauth_uuid)
-
-    r_key = f"da_oauth:{oauth_uuid_str}"
-
-    session_json_bytes = cast(bytes, r.get(r_key))
-    if not session_json_bytes:
-        raise HTTPException(
-            status_code=400,
-            detail=f"Slack OAuth failed - OAuth state key not found: key={r_key}",
-        )
-
-    session_json = session_json_bytes.decode("utf-8")
-    try:
-        session = SlackOAuth.parse_session(session_json)
-
-        if not DEV_MODE:
-            redirect_uri = SlackOAuth.REDIRECT_URI
-        else:
-            redirect_uri = SlackOAuth.DEV_REDIRECT_URI
-
-        # Exchange the authorization code for an access token
-        response = requests.post(
-            SlackOAuth.TOKEN_URL,
-            headers={"Content-Type": "application/x-www-form-urlencoded"},
-            data={
-                "client_id": SlackOAuth.CLIENT_ID,
-                "client_secret": SlackOAuth.CLIENT_SECRET,
-                "code": code,
-                "redirect_uri": redirect_uri,
-            },
-        )
-
-        response_data = response.json()
-
-        if not response_data.get("ok"):
-            raise HTTPException(
-                status_code=400,
-                detail=f"Slack OAuth failed: {response_data.get('error')}",
-            )
-
-        # Extract token and team information
-        access_token: str = response_data.get("access_token")
-        team_id: str = response_data.get("team", {}).get("id")
-        authed_user_id: str = response_data.get("authed_user", {}).get("id")
-
-        credential_info = CredentialBase(
-            credential_json={"slack_bot_token": access_token},
-            admin_public=True,
-            source=DocumentSource.SLACK,
-            name="Slack OAuth",
-        )
-
-        create_credential(credential_info, user, db_session)
-    except Exception as e:
-        return JSONResponse(
-            status_code=500,
-            content={
-                "success": False,
-                "message": f"An error occurred during Slack OAuth: {str(e)}",
-            },
-        )
-    finally:
-        r.delete(r_key)
-
-    # return the result
-    return JSONResponse(
-        content={
-            "success": True,
-            "message": "Slack OAuth completed successfully.",
-            "finalize_url": None,
-            "redirect_on_success": session.redirect_on_success,
-            "team_id": team_id,
-            "authed_user_id": authed_user_id,
-        }
-    )

backend/ee/onyx/server/query_and_chat/chat_backend.py

@@ -1,35 +1,26 @@
 import re
-from typing import cast
 
 from fastapi import APIRouter
 from fastapi import Depends
 from fastapi import HTTPException
 from sqlalchemy.orm import Session
 
-from ee.onyx.server.query_and_chat.models import AgentAnswer
-from ee.onyx.server.query_and_chat.models import AgentSubQuery
-from ee.onyx.server.query_and_chat.models import AgentSubQuestion
 from ee.onyx.server.query_and_chat.models import BasicCreateChatMessageRequest
 from ee.onyx.server.query_and_chat.models import (
     BasicCreateChatMessageWithHistoryRequest,
 )
 from ee.onyx.server.query_and_chat.models import ChatBasicResponse
+from ee.onyx.server.query_and_chat.models import SimpleDoc
 from onyx.auth.users import current_user
 from onyx.chat.chat_utils import combine_message_thread
 from onyx.chat.chat_utils import create_chat_chain
-from onyx.chat.models import AgentAnswerPiece
 from onyx.chat.models import AllCitations
-from onyx.chat.models import ExtendedToolResponse
 from onyx.chat.models import FinalUsedContextDocsResponse
 from onyx.chat.models import LlmDoc
 from onyx.chat.models import LLMRelevanceFilterResponse
 from onyx.chat.models import OnyxAnswerPiece
 from onyx.chat.models import QADocsResponse
-from onyx.chat.models import RefinedAnswerImprovement
 from onyx.chat.models import StreamingError
-from onyx.chat.models import SubQueryPiece
-from onyx.chat.models import SubQuestionIdentifier
-from onyx.chat.models import SubQuestionPiece
 from onyx.chat.process_message import ChatPacketStream
 from onyx.chat.process_message import stream_chat_message_objects
 from onyx.configs.chat_configs import CHAT_TARGET_CHUNK_PERCENTAGE
@@ -55,6 +46,25 @@ logger = setup_logger()
 router = APIRouter(prefix="/chat")
 
 
+def _translate_doc_response_to_simple_doc(
+    doc_response: QADocsResponse,
+) -> list[SimpleDoc]:
+    return [
+        SimpleDoc(
+            id=doc.document_id,
+            semantic_identifier=doc.semantic_identifier,
+            link=doc.link,
+            blurb=doc.blurb,
+            match_highlights=[
+                highlight for highlight in doc.match_highlights if highlight
+            ],
+            source_type=doc.source_type,
+            metadata=doc.metadata,
+        )
+        for doc in doc_response.top_documents
+    ]
+
+
 def _get_final_context_doc_indices(
     final_context_docs: list[LlmDoc] | None,
     top_docs: list[SavedSearchDoc] | None,
@@ -79,26 +89,14 @@ def _convert_packet_stream_to_response(
     final_context_docs: list[LlmDoc] = []
 
     answer = ""
-
-    # accumulate stream data with these dicts
-    agent_sub_questions: dict[tuple[int, int], AgentSubQuestion] = {}
-    agent_answers: dict[tuple[int, int], AgentAnswer] = {}
-    agent_sub_queries: dict[tuple[int, int, int], AgentSubQuery] = {}
-
     for packet in packets:
         if isinstance(packet, OnyxAnswerPiece) and packet.answer_piece:
             answer += packet.answer_piece
         elif isinstance(packet, QADocsResponse):
             response.top_documents = packet.top_documents
 
-            # This is a no-op if agent_sub_questions hasn't already been filled
-            if packet.level is not None and packet.level_question_num is not None:
-                id = (packet.level, packet.level_question_num)
-                if id in agent_sub_questions:
-                    agent_sub_questions[id].document_ids = [
-                        saved_search_doc.document_id
-                        for saved_search_doc in packet.top_documents
-                    ]
+            # TODO: deprecate `simple_search_docs`
+            response.simple_search_docs = _translate_doc_response_to_simple_doc(packet)
         elif isinstance(packet, StreamingError):
             response.error_msg = packet.error
         elif isinstance(packet, ChatMessageDetail):
@@ -115,104 +113,11 @@ def _convert_packet_stream_to_response(
                 citation.citation_num: citation.document_id
                 for citation in packet.citations
             }
-        # agentic packets
-        elif isinstance(packet, SubQuestionPiece):
-            if packet.level is not None and packet.level_question_num is not None:
-                id = (packet.level, packet.level_question_num)
-                if agent_sub_questions.get(id) is None:
-                    agent_sub_questions[id] = AgentSubQuestion(
-                        level=packet.level,
-                        level_question_num=packet.level_question_num,
-                        sub_question=packet.sub_question,
-                        document_ids=[],
-                    )
-                else:
-                    agent_sub_questions[id].sub_question += packet.sub_question
-
-        elif isinstance(packet, AgentAnswerPiece):
-            if packet.level is not None and packet.level_question_num is not None:
-                id = (packet.level, packet.level_question_num)
-                if agent_answers.get(id) is None:
-                    agent_answers[id] = AgentAnswer(
-                        level=packet.level,
-                        level_question_num=packet.level_question_num,
-                        answer=packet.answer_piece,
-                        answer_type=packet.answer_type,
-                    )
-                else:
-                    agent_answers[id].answer += packet.answer_piece
-        elif isinstance(packet, SubQueryPiece):
-            if packet.level is not None and packet.level_question_num is not None:
-                sub_query_id = (
-                    packet.level,
-                    packet.level_question_num,
-                    packet.query_id,
-                )
-                if agent_sub_queries.get(sub_query_id) is None:
-                    agent_sub_queries[sub_query_id] = AgentSubQuery(
-                        level=packet.level,
-                        level_question_num=packet.level_question_num,
-                        sub_query=packet.sub_query,
-                        query_id=packet.query_id,
-                    )
-                else:
-                    agent_sub_queries[sub_query_id].sub_query += packet.sub_query
-        elif isinstance(packet, ExtendedToolResponse):
-            # we shouldn't get this ... it gets intercepted and translated to QADocsResponse
-            logger.warning(
-                "_convert_packet_stream_to_response: Unexpected chat packet type ExtendedToolResponse!"
-            )
-        elif isinstance(packet, RefinedAnswerImprovement):
-            response.agent_refined_answer_improvement = (
-                packet.refined_answer_improvement
-            )
-        else:
-            logger.warning(
-                f"_convert_packet_stream_to_response - Unrecognized chat packet: type={type(packet)}"
-            )
 
     response.final_context_doc_indices = _get_final_context_doc_indices(
         final_context_docs, response.top_documents
     )
 
-    # organize / sort agent metadata for output
-    if len(agent_sub_questions) > 0:
-        response.agent_sub_questions = cast(
-            dict[int, list[AgentSubQuestion]],
-            SubQuestionIdentifier.make_dict_by_level(agent_sub_questions),
-        )
-
-    if len(agent_answers) > 0:
-        # return the agent_level_answer from the first level or the last one depending
-        # on agent_refined_answer_improvement
-        response.agent_answers = cast(
-            dict[int, list[AgentAnswer]],
-            SubQuestionIdentifier.make_dict_by_level(agent_answers),
-        )
-        if response.agent_answers:
-            selected_answer_level = (
-                0
-                if not response.agent_refined_answer_improvement
-                else len(response.agent_answers) - 1
-            )
-            level_answers = response.agent_answers[selected_answer_level]
-            for level_answer in level_answers:
-                if level_answer.answer_type != "agent_level_answer":
-                    continue
-
-                answer = level_answer.answer
-                break
-
-    if len(agent_sub_queries) > 0:
-        # subqueries are often emitted with trailing whitespace ... clean it up here
-        # perhaps fix at the source?
-        for v in agent_sub_queries.values():
-            v.sub_query = v.sub_query.strip()
-
-        response.agent_sub_queries = (
-            AgentSubQuery.make_dict_by_level_and_question_index(agent_sub_queries)
-        )
-
     response.answer = answer
     if answer:
         response.answer_citationless = remove_answer_citations(answer)

backend/ee/onyx/server/query_and_chat/models.py

@@ -1,5 +1,3 @@
-from collections import OrderedDict
-from typing import Literal
 from uuid import UUID
 
 from pydantic import BaseModel
@@ -8,9 +6,9 @@ from pydantic import model_validator
 
 from ee.onyx.server.manage.models import StandardAnswer
 from onyx.chat.models import CitationInfo
+from onyx.chat.models import OnyxContexts
 from onyx.chat.models import PersonaOverrideConfig
 from onyx.chat.models import QADocsResponse
-from onyx.chat.models import SubQuestionIdentifier
 from onyx.chat.models import ThreadMessage
 from onyx.configs.constants import DocumentSource
 from onyx.context.search.enums import LLMEvaluationType
@@ -90,64 +88,6 @@ class SimpleDoc(BaseModel):
     metadata: dict | None
 
 
-class AgentSubQuestion(SubQuestionIdentifier):
-    sub_question: str
-    document_ids: list[str]
-
-
-class AgentAnswer(SubQuestionIdentifier):
-    answer: str
-    answer_type: Literal["agent_sub_answer", "agent_level_answer"]
-
-
-class AgentSubQuery(SubQuestionIdentifier):
-    sub_query: str
-    query_id: int
-
-    @staticmethod
-    def make_dict_by_level_and_question_index(
-        original_dict: dict[tuple[int, int, int], "AgentSubQuery"]
-    ) -> dict[int, dict[int, list["AgentSubQuery"]]]:
-        """Takes a dict of tuple(level, question num, query_id) to sub queries.
-
-        returns a dict of level to dict[question num to list of query_id's]
-        Ordering is asc for readability.
-        """
-        # In this function, when we sort int | None, we deliberately push None to the end
-
-        # map entries to the level_question_dict
-        level_question_dict: dict[int, dict[int, list["AgentSubQuery"]]] = {}
-        for k1, obj in original_dict.items():
-            level = k1[0]
-            question = k1[1]
-
-            if level not in level_question_dict:
-                level_question_dict[level] = {}
-
-            if question not in level_question_dict[level]:
-                level_question_dict[level][question] = []
-
-            level_question_dict[level][question].append(obj)
-
-        # sort each query_id list and question_index
-        for key1, obj1 in level_question_dict.items():
-            for key2, value2 in obj1.items():
-                # sort the query_id list of each question_index
-                level_question_dict[key1][key2] = sorted(
-                    value2, key=lambda o: o.query_id
-                )
-            # sort the question_index dict of level
-            level_question_dict[key1] = OrderedDict(
-                sorted(level_question_dict[key1].items(), key=lambda x: (x is None, x))
-            )
-
-        # sort the top dict of levels
-        sorted_dict = OrderedDict(
-            sorted(level_question_dict.items(), key=lambda x: (x is None, x))
-        )
-        return sorted_dict
-
-
 class ChatBasicResponse(BaseModel):
     # This is built piece by piece, any of these can be None as the flow could break
     answer: str | None = None
@@ -163,14 +103,10 @@ class ChatBasicResponse(BaseModel):
     cited_documents: dict[int, str] | None = None
 
     # FOR BACKWARDS COMPATIBILITY
+    # TODO: deprecate both of these
+    simple_search_docs: list[SimpleDoc] | None = None
     llm_chunks_indices: list[int] | None = None
 
-    # agentic fields
-    agent_sub_questions: dict[int, list[AgentSubQuestion]] | None = None
-    agent_answers: dict[int, list[AgentAnswer]] | None = None
-    agent_sub_queries: dict[int, dict[int, list[AgentSubQuery]]] | None = None
-    agent_refined_answer_improvement: bool | None = None
-
 
 class OneShotQARequest(ChunkContext):
     # Supports simplier APIs that don't deal with chat histories or message edits
@@ -217,3 +153,4 @@ class OneShotQAResponse(BaseModel):
     llm_selected_doc_indices: list[int] | None = None
     error_msg: str | None = None
     chat_message_id: int | None = None
+    contexts: OnyxContexts | None = None

backend/ee/onyx/server/query_history/api.py

@@ -2,7 +2,6 @@ import csv
 import io
 from datetime import datetime
 from datetime import timezone
-from http import HTTPStatus
 from uuid import UUID
 
 from fastapi import APIRouter
@@ -22,10 +21,8 @@ from ee.onyx.server.query_history.models import QuestionAnswerPairSnapshot
 from onyx.auth.users import current_admin_user
 from onyx.auth.users import get_display_email
 from onyx.chat.chat_utils import create_chat_chain
-from onyx.configs.app_configs import ONYX_QUERY_HISTORY_TYPE
 from onyx.configs.constants import MessageType
 from onyx.configs.constants import QAFeedbackType
-from onyx.configs.constants import QueryHistoryType
 from onyx.configs.constants import SessionType
 from onyx.db.chat import get_chat_session_by_id
 from onyx.db.chat import get_chat_sessions_by_user
@@ -38,8 +35,6 @@ from onyx.server.query_and_chat.models import ChatSessionsResponse
 
 router = APIRouter()
 
-ONYX_ANONYMIZED_EMAIL = "anonymous@anonymous.invalid"
-
 
 def fetch_and_process_chat_session_history(
     db_session: Session,
@@ -48,15 +43,10 @@ def fetch_and_process_chat_session_history(
     feedback_type: QAFeedbackType | None,
     limit: int | None = 500,
 ) -> list[ChatSessionSnapshot]:
-    # observed to be slow a scale of 8192 sessions and 4 messages per session
-
-    # this is a little slow (5 seconds)
     chat_sessions = fetch_chat_sessions_eagerly_by_time(
         start=start, end=end, db_session=db_session, limit=limit
     )
 
-    # this is VERY slow (80 seconds) due to create_chat_chain being called
-    # for each session. Needs optimizing.
     chat_session_snapshots = [
         snapshot_from_chat_session(chat_session=chat_session, db_session=db_session)
         for chat_session in chat_sessions
@@ -117,17 +107,6 @@ def get_user_chat_sessions(
     _: User | None = Depends(current_admin_user),
     db_session: Session = Depends(get_session),
 ) -> ChatSessionsResponse:
-    # we specifically don't allow this endpoint if "anonymized" since
-    # this is a direct query on the user id
-    if ONYX_QUERY_HISTORY_TYPE in [
-        QueryHistoryType.DISABLED,
-        QueryHistoryType.ANONYMIZED,
-    ]:
-        raise HTTPException(
-            status_code=HTTPStatus.FORBIDDEN,
-            detail="Per user query history has been disabled by the administrator.",
-        )
-
     try:
         chat_sessions = get_chat_sessions_by_user(
             user_id=user_id, deleted=False, db_session=db_session, limit=0
@@ -143,7 +122,6 @@ def get_user_chat_sessions(
                 name=chat.description,
                 persona_id=chat.persona_id,
                 time_created=chat.time_created.isoformat(),
-                time_updated=chat.time_updated.isoformat(),
                 shared_status=chat.shared_status,
                 folder_id=chat.folder_id,
                 current_alternate_model=chat.current_alternate_model,
@@ -163,12 +141,6 @@ def get_chat_session_history(
     _: User | None = Depends(current_admin_user),
     db_session: Session = Depends(get_session),
 ) -> PaginatedReturn[ChatSessionMinimal]:
-    if ONYX_QUERY_HISTORY_TYPE == QueryHistoryType.DISABLED:
-        raise HTTPException(
-            status_code=HTTPStatus.FORBIDDEN,
-            detail="Query history has been disabled by the administrator.",
-        )
-
     page_of_chat_sessions = get_page_of_chat_sessions(
         page_num=page_num,
         page_size=page_size,
@@ -185,16 +157,11 @@ def get_chat_session_history(
         feedback_filter=feedback_type,
     )
 
-    minimal_chat_sessions: list[ChatSessionMinimal] = []
-
-    for chat_session in page_of_chat_sessions:
-        minimal_chat_session = ChatSessionMinimal.from_chat_session(chat_session)
-        if ONYX_QUERY_HISTORY_TYPE == QueryHistoryType.ANONYMIZED:
-            minimal_chat_session.user_email = ONYX_ANONYMIZED_EMAIL
-        minimal_chat_sessions.append(minimal_chat_session)
-
     return PaginatedReturn(
-        items=minimal_chat_sessions,
+        items=[
+            ChatSessionMinimal.from_chat_session(chat_session)
+            for chat_session in page_of_chat_sessions
+        ],
         total_items=total_filtered_chat_sessions_count,
     )
 
@@ -205,12 +172,6 @@ def get_chat_session_admin(
     _: User | None = Depends(current_admin_user),
     db_session: Session = Depends(get_session),
 ) -> ChatSessionSnapshot:
-    if ONYX_QUERY_HISTORY_TYPE == QueryHistoryType.DISABLED:
-        raise HTTPException(
-            status_code=HTTPStatus.FORBIDDEN,
-            detail="Query history has been disabled by the administrator.",
-        )
-
     try:
         chat_session = get_chat_session_by_id(
             chat_session_id=chat_session_id,
@@ -232,9 +193,6 @@ def get_chat_session_admin(
             f"Could not create snapshot for chat session with id '{chat_session_id}'",
         )
 
-    if ONYX_QUERY_HISTORY_TYPE == QueryHistoryType.ANONYMIZED:
-        snapshot.user_email = ONYX_ANONYMIZED_EMAIL
-
     return snapshot
 
 
@@ -245,14 +203,6 @@ def get_query_history_as_csv(
     end: datetime | None = None,
     db_session: Session = Depends(get_session),
 ) -> StreamingResponse:
-    if ONYX_QUERY_HISTORY_TYPE == QueryHistoryType.DISABLED:
-        raise HTTPException(
-            status_code=HTTPStatus.FORBIDDEN,
-            detail="Query history has been disabled by the administrator.",
-        )
-
-    # this call is very expensive and is timing out via endpoint
-    # TODO: optimize call and/or generate via background task
     complete_chat_session_history = fetch_and_process_chat_session_history(
         db_session=db_session,
         start=start or datetime.fromtimestamp(0, tz=timezone.utc),
@@ -263,9 +213,6 @@ def get_query_history_as_csv(
 
     question_answer_pairs: list[QuestionAnswerPairSnapshot] = []
     for chat_session_snapshot in complete_chat_session_history:
-        if ONYX_QUERY_HISTORY_TYPE == QueryHistoryType.ANONYMIZED:
-            chat_session_snapshot.user_email = ONYX_ANONYMIZED_EMAIL
-
         question_answer_pairs.extend(
             QuestionAnswerPairSnapshot.from_chat_session_snapshot(chat_session_snapshot)
         )

backend/ee/onyx/server/saml.py

@@ -38,7 +38,6 @@ router = APIRouter(prefix="/auth/saml")
 
 
 async def upsert_saml_user(email: str) -> User:
-    logger.debug(f"Attempting to upsert SAML user with email: {email}")
     get_async_session_context = contextlib.asynccontextmanager(
         get_async_session
     )  # type:ignore
@@ -49,13 +48,9 @@ async def upsert_saml_user(email: str) -> User:
         async with get_user_db_context(session) as user_db:
             async with get_user_manager_context(user_db) as user_manager:
                 try:
-                    user = await user_manager.get_by_email(email)
-                    # If user has a non-authenticated role, treat as non-existent
-                    if not user.role.is_web_login():
-                        raise exceptions.UserNotExists()
-                    return user
+                    return await user_manager.get_by_email(email)
                 except exceptions.UserNotExists:
-                    logger.info("Creating user from SAML login")
+                    logger.notice("Creating user from SAML login")
 
                 user_count = await get_user_count()
                 role = UserRole.ADMIN if user_count == 0 else UserRole.BASIC
@@ -64,10 +59,11 @@ async def upsert_saml_user(email: str) -> User:
                 password = fastapi_users_pw_helper.generate()
                 hashed_pass = fastapi_users_pw_helper.hash(password)
 
-                user = await user_manager.create(
+                user: User = await user_manager.create(
                     UserCreate(
                         email=email,
                         password=hashed_pass,
+                        is_verified=True,
                         role=role,
                     )
                 )

backend/ee/onyx/server/tenants/admin_api.py

@@ -1,45 +0,0 @@
-from fastapi import APIRouter
-from fastapi import Depends
-from fastapi import HTTPException
-from fastapi import Response
-
-from ee.onyx.auth.users import current_cloud_superuser
-from ee.onyx.server.tenants.models import ImpersonateRequest
-from ee.onyx.server.tenants.user_mapping import get_tenant_id_for_email
-from onyx.auth.users import auth_backend
-from onyx.auth.users import get_redis_strategy
-from onyx.auth.users import User
-from onyx.db.engine import get_session_with_tenant
-from onyx.db.users import get_user_by_email
-from onyx.utils.logger import setup_logger
-
-logger = setup_logger()
-
-router = APIRouter(prefix="/tenants")
-
-
-@router.post("/impersonate")
-async def impersonate_user(
-    impersonate_request: ImpersonateRequest,
-    _: User = Depends(current_cloud_superuser),
-) -> Response:
-    """Allows a cloud superuser to impersonate another user by generating an impersonation JWT token"""
-    tenant_id = get_tenant_id_for_email(impersonate_request.email)
-
-    with get_session_with_tenant(tenant_id=tenant_id) as tenant_session:
-        user_to_impersonate = get_user_by_email(
-            impersonate_request.email, tenant_session
-        )
-        if user_to_impersonate is None:
-            raise HTTPException(status_code=404, detail="User not found")
-        token = await get_redis_strategy().write_token(user_to_impersonate)
-
-    response = await auth_backend.transport.get_login_response(token)
-    response.set_cookie(
-        key="fastapiusersauth",
-        value=token,
-        httponly=True,
-        secure=True,
-        samesite="lax",
-    )
-    return response

backend/ee/onyx/server/tenants/anonymous_users_api.py

@@ -1,98 +0,0 @@
-from fastapi import APIRouter
-from fastapi import Depends
-from fastapi import HTTPException
-from fastapi import Response
-from sqlalchemy.exc import IntegrityError
-
-from ee.onyx.auth.users import generate_anonymous_user_jwt_token
-from ee.onyx.configs.app_configs import ANONYMOUS_USER_COOKIE_NAME
-from ee.onyx.server.tenants.anonymous_user_path import get_anonymous_user_path
-from ee.onyx.server.tenants.anonymous_user_path import (
-    get_tenant_id_for_anonymous_user_path,
-)
-from ee.onyx.server.tenants.anonymous_user_path import modify_anonymous_user_path
-from ee.onyx.server.tenants.anonymous_user_path import validate_anonymous_user_path
-from ee.onyx.server.tenants.models import AnonymousUserPath
-from onyx.auth.users import anonymous_user_enabled
-from onyx.auth.users import current_admin_user
-from onyx.auth.users import optional_user
-from onyx.auth.users import User
-from onyx.configs.constants import FASTAPI_USERS_AUTH_COOKIE_NAME
-from onyx.db.engine import get_session_with_shared_schema
-from onyx.utils.logger import setup_logger
-from shared_configs.contextvars import get_current_tenant_id
-
-logger = setup_logger()
-
-router = APIRouter(prefix="/tenants")
-
-
-@router.get("/anonymous-user-path")
-async def get_anonymous_user_path_api(
-    _: User | None = Depends(current_admin_user),
-) -> AnonymousUserPath:
-    tenant_id = get_current_tenant_id()
-
-    if tenant_id is None:
-        raise HTTPException(status_code=404, detail="Tenant not found")
-
-    with get_session_with_shared_schema() as db_session:
-        current_path = get_anonymous_user_path(tenant_id, db_session)
-
-    return AnonymousUserPath(anonymous_user_path=current_path)
-
-
-@router.post("/anonymous-user-path")
-async def set_anonymous_user_path_api(
-    anonymous_user_path: str,
-    _: User | None = Depends(current_admin_user),
-) -> None:
-    tenant_id = get_current_tenant_id()
-    try:
-        validate_anonymous_user_path(anonymous_user_path)
-    except ValueError as e:
-        raise HTTPException(status_code=400, detail=str(e))
-
-    with get_session_with_shared_schema() as db_session:
-        try:
-            modify_anonymous_user_path(tenant_id, anonymous_user_path, db_session)
-        except IntegrityError:
-            raise HTTPException(
-                status_code=409,
-                detail="The anonymous user path is already in use. Please choose a different path.",
-            )
-        except Exception as e:
-            logger.exception(f"Failed to modify anonymous user path: {str(e)}")
-            raise HTTPException(
-                status_code=500,
-                detail="An unexpected error occurred while modifying the anonymous user path",
-            )
-
-
-@router.post("/anonymous-user")
-async def login_as_anonymous_user(
-    anonymous_user_path: str,
-    _: User | None = Depends(optional_user),
-) -> Response:
-    with get_session_with_shared_schema() as db_session:
-        tenant_id = get_tenant_id_for_anonymous_user_path(
-            anonymous_user_path, db_session
-        )
-        if not tenant_id:
-            raise HTTPException(status_code=404, detail="Tenant not found")
-
-    if not anonymous_user_enabled(tenant_id=tenant_id):
-        raise HTTPException(status_code=403, detail="Anonymous user is not enabled")
-
-    token = generate_anonymous_user_jwt_token(tenant_id)
-
-    response = Response()
-    response.delete_cookie(FASTAPI_USERS_AUTH_COOKIE_NAME)
-    response.set_cookie(
-        key=ANONYMOUS_USER_COOKIE_NAME,
-        value=token,
-        httponly=True,
-        secure=True,
-        samesite="strict",
-    )
-    return response

backend/ee/onyx/server/tenants/api.py

@@ -1,24 +1,269 @@
+import stripe
 from fastapi import APIRouter
+from fastapi import Depends
+from fastapi import HTTPException
+from fastapi import Response
+from sqlalchemy.exc import IntegrityError
+from sqlalchemy.orm import Session
 
-from ee.onyx.server.tenants.admin_api import router as admin_router
-from ee.onyx.server.tenants.anonymous_users_api import router as anonymous_users_router
-from ee.onyx.server.tenants.billing_api import router as billing_router
-from ee.onyx.server.tenants.team_membership_api import router as team_membership_router
-from ee.onyx.server.tenants.tenant_management_api import (
-    router as tenant_management_router,
-)
-from ee.onyx.server.tenants.user_invitations_api import (
-    router as user_invitations_router,
+from ee.onyx.auth.users import current_cloud_superuser
+from ee.onyx.auth.users import generate_anonymous_user_jwt_token
+from ee.onyx.configs.app_configs import ANONYMOUS_USER_COOKIE_NAME
+from ee.onyx.configs.app_configs import STRIPE_SECRET_KEY
+from ee.onyx.server.tenants.access import control_plane_dep
+from ee.onyx.server.tenants.anonymous_user_path import get_anonymous_user_path
+from ee.onyx.server.tenants.anonymous_user_path import (
+    get_tenant_id_for_anonymous_user_path,
 )
+from ee.onyx.server.tenants.anonymous_user_path import modify_anonymous_user_path
+from ee.onyx.server.tenants.anonymous_user_path import validate_anonymous_user_path
+from ee.onyx.server.tenants.billing import fetch_billing_information
+from ee.onyx.server.tenants.billing import fetch_stripe_checkout_session
+from ee.onyx.server.tenants.billing import fetch_tenant_stripe_information
+from ee.onyx.server.tenants.models import AnonymousUserPath
+from ee.onyx.server.tenants.models import BillingInformation
+from ee.onyx.server.tenants.models import ImpersonateRequest
+from ee.onyx.server.tenants.models import ProductGatingRequest
+from ee.onyx.server.tenants.models import ProductGatingResponse
+from ee.onyx.server.tenants.models import SubscriptionSessionResponse
+from ee.onyx.server.tenants.models import SubscriptionStatusResponse
+from ee.onyx.server.tenants.product_gating import store_product_gating
+from ee.onyx.server.tenants.provisioning import delete_user_from_control_plane
+from ee.onyx.server.tenants.user_mapping import get_tenant_id_for_email
+from ee.onyx.server.tenants.user_mapping import remove_all_users_from_tenant
+from ee.onyx.server.tenants.user_mapping import remove_users_from_tenant
+from onyx.auth.users import anonymous_user_enabled
+from onyx.auth.users import auth_backend
+from onyx.auth.users import current_admin_user
+from onyx.auth.users import get_redis_strategy
+from onyx.auth.users import optional_user
+from onyx.auth.users import User
+from onyx.configs.app_configs import WEB_DOMAIN
+from onyx.configs.constants import FASTAPI_USERS_AUTH_COOKIE_NAME
+from onyx.db.auth import get_user_count
+from onyx.db.engine import get_session
+from onyx.db.engine import get_session_with_shared_schema
+from onyx.db.engine import get_session_with_tenant
+from onyx.db.users import delete_user_from_db
+from onyx.db.users import get_user_by_email
+from onyx.server.manage.models import UserByEmail
+from onyx.utils.logger import setup_logger
+from shared_configs.contextvars import CURRENT_TENANT_ID_CONTEXTVAR
+from shared_configs.contextvars import get_current_tenant_id
 
-# Create a main router to include all sub-routers
-# Note: We don't add a prefix here as each router already has the /tenants prefix
-router = APIRouter()
+stripe.api_key = STRIPE_SECRET_KEY
+logger = setup_logger()
+router = APIRouter(prefix="/tenants")
 
-# Include all the individual routers
-router.include_router(admin_router)
-router.include_router(anonymous_users_router)
-router.include_router(billing_router)
-router.include_router(team_membership_router)
-router.include_router(tenant_management_router)
-router.include_router(user_invitations_router)
+
+@router.get("/anonymous-user-path")
+async def get_anonymous_user_path_api(
+    _: User | None = Depends(current_admin_user),
+) -> AnonymousUserPath:
+    tenant_id = get_current_tenant_id()
+
+    if tenant_id is None:
+        raise HTTPException(status_code=404, detail="Tenant not found")
+
+    with get_session_with_shared_schema() as db_session:
+        current_path = get_anonymous_user_path(tenant_id, db_session)
+
+    return AnonymousUserPath(anonymous_user_path=current_path)
+
+
+@router.post("/anonymous-user-path")
+async def set_anonymous_user_path_api(
+    anonymous_user_path: str,
+    _: User | None = Depends(current_admin_user),
+) -> None:
+    tenant_id = get_current_tenant_id()
+    try:
+        validate_anonymous_user_path(anonymous_user_path)
+    except ValueError as e:
+        raise HTTPException(status_code=400, detail=str(e))
+
+    with get_session_with_shared_schema() as db_session:
+        try:
+            modify_anonymous_user_path(tenant_id, anonymous_user_path, db_session)
+        except IntegrityError:
+            raise HTTPException(
+                status_code=409,
+                detail="The anonymous user path is already in use. Please choose a different path.",
+            )
+        except Exception as e:
+            logger.exception(f"Failed to modify anonymous user path: {str(e)}")
+            raise HTTPException(
+                status_code=500,
+                detail="An unexpected error occurred while modifying the anonymous user path",
+            )
+
+
+@router.post("/anonymous-user")
+async def login_as_anonymous_user(
+    anonymous_user_path: str,
+    _: User | None = Depends(optional_user),
+) -> Response:
+    with get_session_with_shared_schema() as db_session:
+        tenant_id = get_tenant_id_for_anonymous_user_path(
+            anonymous_user_path, db_session
+        )
+        if not tenant_id:
+            raise HTTPException(status_code=404, detail="Tenant not found")
+
+    if not anonymous_user_enabled(tenant_id=tenant_id):
+        raise HTTPException(status_code=403, detail="Anonymous user is not enabled")
+
+    token = generate_anonymous_user_jwt_token(tenant_id)
+
+    response = Response()
+    response.delete_cookie(FASTAPI_USERS_AUTH_COOKIE_NAME)
+    response.set_cookie(
+        key=ANONYMOUS_USER_COOKIE_NAME,
+        value=token,
+        httponly=True,
+        secure=True,
+        samesite="strict",
+    )
+    return response
+
+
+@router.post("/product-gating")
+def gate_product(
+    product_gating_request: ProductGatingRequest, _: None = Depends(control_plane_dep)
+) -> ProductGatingResponse:
+    """
+    Gating the product means that the product is not available to the tenant.
+    They will be directed to the billing page.
+    We gate the product when their subscription has ended.
+    """
+    try:
+        store_product_gating(
+            product_gating_request.tenant_id, product_gating_request.application_status
+        )
+        return ProductGatingResponse(updated=True, error=None)
+
+    except Exception as e:
+        logger.exception("Failed to gate product")
+        return ProductGatingResponse(updated=False, error=str(e))
+
+
+@router.get("/billing-information")
+async def billing_information(
+    _: User = Depends(current_admin_user),
+) -> BillingInformation | SubscriptionStatusResponse:
+    logger.info("Fetching billing information")
+    tenant_id = get_current_tenant_id()
+    return fetch_billing_information(tenant_id)
+
+
+@router.post("/create-customer-portal-session")
+async def create_customer_portal_session(
+    _: User = Depends(current_admin_user),
+) -> dict:
+    tenant_id = get_current_tenant_id()
+
+    try:
+        stripe_info = fetch_tenant_stripe_information(tenant_id)
+        stripe_customer_id = stripe_info.get("stripe_customer_id")
+        if not stripe_customer_id:
+            raise HTTPException(status_code=400, detail="Stripe customer ID not found")
+        logger.info(stripe_customer_id)
+
+        portal_session = stripe.billing_portal.Session.create(
+            customer=stripe_customer_id,
+            return_url=f"{WEB_DOMAIN}/admin/billing",
+        )
+        logger.info(portal_session)
+        return {"url": portal_session.url}
+    except Exception as e:
+        logger.exception("Failed to create customer portal session")
+        raise HTTPException(status_code=500, detail=str(e))
+
+
+@router.post("/create-subscription-session")
+async def create_subscription_session(
+    _: User = Depends(current_admin_user),
+) -> SubscriptionSessionResponse:
+    try:
+        tenant_id = CURRENT_TENANT_ID_CONTEXTVAR.get()
+        if not tenant_id:
+            raise HTTPException(status_code=400, detail="Tenant ID not found")
+        session_id = fetch_stripe_checkout_session(tenant_id)
+        return SubscriptionSessionResponse(sessionId=session_id)
+
+    except Exception as e:
+        logger.exception("Failed to create resubscription session")
+        raise HTTPException(status_code=500, detail=str(e))
+
+
+@router.post("/impersonate")
+async def impersonate_user(
+    impersonate_request: ImpersonateRequest,
+    _: User = Depends(current_cloud_superuser),
+) -> Response:
+    """Allows a cloud superuser to impersonate another user by generating an impersonation JWT token"""
+    tenant_id = get_tenant_id_for_email(impersonate_request.email)
+
+    with get_session_with_tenant(tenant_id=tenant_id) as tenant_session:
+        user_to_impersonate = get_user_by_email(
+            impersonate_request.email, tenant_session
+        )
+        if user_to_impersonate is None:
+            raise HTTPException(status_code=404, detail="User not found")
+        token = await get_redis_strategy().write_token(user_to_impersonate)
+
+    response = await auth_backend.transport.get_login_response(token)
+    response.set_cookie(
+        key="fastapiusersauth",
+        value=token,
+        httponly=True,
+        secure=True,
+        samesite="lax",
+    )
+    return response
+
+
+@router.post("/leave-organization")
+async def leave_organization(
+    user_email: UserByEmail,
+    current_user: User | None = Depends(current_admin_user),
+    db_session: Session = Depends(get_session),
+) -> None:
+    tenant_id = get_current_tenant_id()
+
+    if current_user is None or current_user.email != user_email.user_email:
+        raise HTTPException(
+            status_code=403, detail="You can only leave the organization as yourself"
+        )
+
+    user_to_delete = get_user_by_email(user_email.user_email, db_session)
+    if user_to_delete is None:
+        raise HTTPException(status_code=404, detail="User not found")
+
+    num_admin_users = await get_user_count(only_admin_users=True)
+
+    should_delete_tenant = num_admin_users == 1
+
+    if should_delete_tenant:
+        logger.info(
+            "Last admin user is leaving the organization. Deleting tenant from control plane."
+        )
+        try:
+            await delete_user_from_control_plane(tenant_id, user_to_delete.email)
+            logger.debug("User deleted from control plane")
+        except Exception as e:
+            logger.exception(
+                f"Failed to delete user from control plane for tenant {tenant_id}: {e}"
+            )
+            raise HTTPException(
+                status_code=500,
+                detail=f"Failed to remove user from control plane: {str(e)}",
+            )
+
+    db_session.expunge(user_to_delete)
+    delete_user_from_db(user_to_delete, db_session)
+
+    if should_delete_tenant:
+        remove_all_users_from_tenant(tenant_id)
+    else:
+        remove_users_from_tenant([user_to_delete.email], tenant_id)

backend/ee/onyx/server/tenants/billing.py

@@ -7,7 +7,6 @@ from ee.onyx.configs.app_configs import STRIPE_PRICE_ID
 from ee.onyx.configs.app_configs import STRIPE_SECRET_KEY
 from ee.onyx.server.tenants.access import generate_data_plane_token
 from ee.onyx.server.tenants.models import BillingInformation
-from ee.onyx.server.tenants.models import SubscriptionStatusResponse
 from onyx.configs.app_configs import CONTROL_PLANE_API_BASE_URL
 from onyx.utils.logger import setup_logger
 
@@ -42,9 +41,7 @@ def fetch_tenant_stripe_information(tenant_id: str) -> dict:
     return response.json()
 
 
-def fetch_billing_information(
-    tenant_id: str,
-) -> BillingInformation | SubscriptionStatusResponse:
+def fetch_billing_information(tenant_id: str) -> BillingInformation:
     logger.info("Fetching billing information")
     token = generate_data_plane_token()
     headers = {
@@ -55,19 +52,8 @@ def fetch_billing_information(
     params = {"tenant_id": tenant_id}
     response = requests.get(url, headers=headers, params=params)
     response.raise_for_status()
-
-    response_data = response.json()
-
-    # Check if the response indicates no subscription
-    if (
-        isinstance(response_data, dict)
-        and "subscribed" in response_data
-        and not response_data["subscribed"]
-    ):
-        return SubscriptionStatusResponse(**response_data)
-
-    # Otherwise, parse as BillingInformation
-    return BillingInformation(**response_data)
+    billing_info = BillingInformation(**response.json())
+    return billing_info
 
 
 def register_tenant_users(tenant_id: str, number_of_users: int) -> stripe.Subscription:

backend/ee/onyx/server/tenants/billing_api.py

@@ -1,96 +0,0 @@
-import stripe
-from fastapi import APIRouter
-from fastapi import Depends
-from fastapi import HTTPException
-
-from ee.onyx.auth.users import current_admin_user
-from ee.onyx.configs.app_configs import STRIPE_SECRET_KEY
-from ee.onyx.server.tenants.access import control_plane_dep
-from ee.onyx.server.tenants.billing import fetch_billing_information
-from ee.onyx.server.tenants.billing import fetch_stripe_checkout_session
-from ee.onyx.server.tenants.billing import fetch_tenant_stripe_information
-from ee.onyx.server.tenants.models import BillingInformation
-from ee.onyx.server.tenants.models import ProductGatingRequest
-from ee.onyx.server.tenants.models import ProductGatingResponse
-from ee.onyx.server.tenants.models import SubscriptionSessionResponse
-from ee.onyx.server.tenants.models import SubscriptionStatusResponse
-from ee.onyx.server.tenants.product_gating import store_product_gating
-from onyx.auth.users import User
-from onyx.configs.app_configs import WEB_DOMAIN
-from onyx.utils.logger import setup_logger
-from shared_configs.contextvars import CURRENT_TENANT_ID_CONTEXTVAR
-from shared_configs.contextvars import get_current_tenant_id
-
-stripe.api_key = STRIPE_SECRET_KEY
-logger = setup_logger()
-
-router = APIRouter(prefix="/tenants")
-
-
-@router.post("/product-gating")
-def gate_product(
-    product_gating_request: ProductGatingRequest, _: None = Depends(control_plane_dep)
-) -> ProductGatingResponse:
-    """
-    Gating the product means that the product is not available to the tenant.
-    They will be directed to the billing page.
-    We gate the product when their subscription has ended.
-    """
-    try:
-        store_product_gating(
-            product_gating_request.tenant_id, product_gating_request.application_status
-        )
-        return ProductGatingResponse(updated=True, error=None)
-
-    except Exception as e:
-        logger.exception("Failed to gate product")
-        return ProductGatingResponse(updated=False, error=str(e))
-
-
-@router.get("/billing-information")
-async def billing_information(
-    _: User = Depends(current_admin_user),
-) -> BillingInformation | SubscriptionStatusResponse:
-    logger.info("Fetching billing information")
-    tenant_id = get_current_tenant_id()
-    return fetch_billing_information(tenant_id)
-
-
-@router.post("/create-customer-portal-session")
-async def create_customer_portal_session(
-    _: User = Depends(current_admin_user),
-) -> dict:
-    tenant_id = get_current_tenant_id()
-
-    try:
-        stripe_info = fetch_tenant_stripe_information(tenant_id)
-        stripe_customer_id = stripe_info.get("stripe_customer_id")
-        if not stripe_customer_id:
-            raise HTTPException(status_code=400, detail="Stripe customer ID not found")
-        logger.info(stripe_customer_id)
-
-        portal_session = stripe.billing_portal.Session.create(
-            customer=stripe_customer_id,
-            return_url=f"{WEB_DOMAIN}/admin/billing",
-        )
-        logger.info(portal_session)
-        return {"url": portal_session.url}
-    except Exception as e:
-        logger.exception("Failed to create customer portal session")
-        raise HTTPException(status_code=500, detail=str(e))
-
-
-@router.post("/create-subscription-session")
-async def create_subscription_session(
-    _: User = Depends(current_admin_user),
-) -> SubscriptionSessionResponse:
-    try:
-        tenant_id = CURRENT_TENANT_ID_CONTEXTVAR.get()
-        if not tenant_id:
-            raise HTTPException(status_code=400, detail="Tenant ID not found")
-        session_id = fetch_stripe_checkout_session(tenant_id)
-        return SubscriptionSessionResponse(sessionId=session_id)
-
-    except Exception as e:
-        logger.exception("Failed to create resubscription session")
-        raise HTTPException(status_code=500, detail=str(e))

backend/ee/onyx/server/tenants/models.py

@@ -67,30 +67,3 @@ class ProductGatingResponse(BaseModel):
 
 class SubscriptionSessionResponse(BaseModel):
     sessionId: str
-
-
-class TenantByDomainResponse(BaseModel):
-    tenant_id: str
-    number_of_users: int
-    creator_email: str
-
-
-class TenantByDomainRequest(BaseModel):
-    email: str
-
-
-class RequestInviteRequest(BaseModel):
-    tenant_id: str
-
-
-class RequestInviteResponse(BaseModel):
-    success: bool
-    message: str
-
-
-class PendingUserSnapshot(BaseModel):
-    email: str
-
-
-class ApproveUserRequest(BaseModel):
-    email: str

backend/ee/onyx/server/tenants/product_gating.py

@@ -48,5 +48,4 @@ def store_product_gating(tenant_id: str, application_status: ApplicationStatus)
 
 def get_gated_tenants() -> set[str]:
     redis_client = get_redis_replica_client(tenant_id=ONYX_CLOUD_TENANT_ID)
-    gated_tenants_bytes = cast(set[bytes], redis_client.smembers(GATED_TENANTS_KEY))
-    return {tenant_id.decode("utf-8") for tenant_id in gated_tenants_bytes}
+    return cast(set[str], redis_client.smembers(GATED_TENANTS_KEY))

backend/ee/onyx/server/tenants/provisioning.py

@@ -4,7 +4,6 @@ import uuid
 
 import aiohttp  # Async HTTP client
 import httpx
-import requests
 from fastapi import HTTPException
 from fastapi import Request
 from sqlalchemy import select
@@ -15,7 +14,6 @@ from ee.onyx.configs.app_configs import COHERE_DEFAULT_API_KEY
 from ee.onyx.configs.app_configs import HUBSPOT_TRACKING_URL
 from ee.onyx.configs.app_configs import OPENAI_DEFAULT_API_KEY
 from ee.onyx.server.tenants.access import generate_data_plane_token
-from ee.onyx.server.tenants.models import TenantByDomainResponse
 from ee.onyx.server.tenants.models import TenantCreationPayload
 from ee.onyx.server.tenants.models import TenantDeletionPayload
 from ee.onyx.server.tenants.schema_management import create_schema_if_not_exists
@@ -28,12 +26,11 @@ from onyx.auth.users import exceptions
 from onyx.configs.app_configs import CONTROL_PLANE_API_BASE_URL
 from onyx.configs.app_configs import DEV_MODE
 from onyx.configs.constants import MilestoneRecordType
-from onyx.db.engine import get_session_with_shared_schema
 from onyx.db.engine import get_session_with_tenant
+from onyx.db.engine import get_sqlalchemy_engine
 from onyx.db.llm import update_default_provider
 from onyx.db.llm import upsert_cloud_embedding_provider
 from onyx.db.llm import upsert_llm_provider
-from onyx.db.models import AvailableTenant
 from onyx.db.models import IndexModelStatus
 from onyx.db.models import SearchSettings
 from onyx.db.models import UserTenantMapping
@@ -58,77 +55,43 @@ logger = logging.getLogger(__name__)
 async def get_or_provision_tenant(
     email: str, referral_source: str | None = None, request: Request | None = None
 ) -> str:
-    """
-    Get existing tenant ID for an email or create a new tenant if none exists.
-    This function should only be called after we have verified we want this user's tenant to exist.
-    It returns the tenant ID associated with the email, creating a new tenant if necessary.
-    """
-    # Early return for non-multi-tenant mode
+    """Get existing tenant ID for an email or create a new tenant if none exists."""
     if not MULTI_TENANT:
         return POSTGRES_DEFAULT_SCHEMA
 
     if referral_source and request:
         await submit_to_hubspot(email, referral_source, request)
 
-    # First, check if the user already has a tenant
-    tenant_id: str | None = None
     try:
         tenant_id = get_tenant_id_for_email(email)
-        return tenant_id
     except exceptions.UserNotExists:
-        # User doesn't exist, so we need to create a new tenant or assign an existing one
-        pass
-
-    try:
-        # Try to get a pre-provisioned tenant
-        tenant_id = await get_available_tenant()
-
-        if tenant_id:
-            # If we have a pre-provisioned tenant, assign it to the user
-            await assign_tenant_to_user(tenant_id, email, referral_source)
-            logger.info(f"Assigned pre-provisioned tenant {tenant_id} to user {email}")
-        else:
-            # If no pre-provisioned tenant is available, create a new one on-demand
+        # If tenant does not exist and in Multi tenant mode, provision a new tenant
+        try:
             tenant_id = await create_tenant(email, referral_source)
+        except Exception as e:
+            logger.error(f"Tenant provisioning failed: {e}")
+            raise HTTPException(status_code=500, detail="Failed to provision tenant.")
 
-        # Notify control plane if we have created / assigned a new tenant
-        if not DEV_MODE:
-            await notify_control_plane(tenant_id, email, referral_source)
-
-        return tenant_id
-
-    except Exception as e:
-        # If we've encountered an error, log and raise an exception
-        error_msg = "Failed to provision tenant"
-        logger.error(error_msg, exc_info=e)
+    if not tenant_id:
         raise HTTPException(
-            status_code=500,
-            detail="Failed to provision tenant. Please try again later.",
+            status_code=401, detail="User does not belong to an organization"
         )
 
+    return tenant_id
+
 
 async def create_tenant(email: str, referral_source: str | None = None) -> str:
-    """
-    Create a new tenant on-demand when no pre-provisioned tenants are available.
-    This is the fallback method when we can't use a pre-provisioned tenant.
-
-    """
     tenant_id = TENANT_ID_PREFIX + str(uuid.uuid4())
-    logger.info(f"Creating new tenant {tenant_id} for user {email}")
-
     try:
         # Provision tenant on data plane
         await provision_tenant(tenant_id, email)
-
+        # Notify control plane
+        if not DEV_MODE:
+            await notify_control_plane(tenant_id, email, referral_source)
     except Exception as e:
-        logger.exception(f"Tenant provisioning failed: {str(e)}")
-        # Attempt to rollback the tenant provisioning
-        try:
-            await rollback_tenant_provisioning(tenant_id)
-        except Exception:
-            logger.exception(f"Failed to rollback tenant provisioning for {tenant_id}")
+        logger.error(f"Tenant provisioning failed: {e}")
+        await rollback_tenant_provisioning(tenant_id)
         raise HTTPException(status_code=500, detail="Failed to provision tenant.")
-
     return tenant_id
 
 
@@ -141,26 +104,55 @@ async def provision_tenant(tenant_id: str, email: str) -> None:
             status_code=409, detail="User already belongs to an organization"
         )
 
-    logger.debug(f"Provisioning tenant {tenant_id} for user {email}")
+    logger.info(f"Provisioning tenant: {tenant_id}")
+    token = None
 
     try:
-        # Create the schema for the tenant
         if not create_schema_if_not_exists(tenant_id):
-            logger.debug(f"Created schema for tenant {tenant_id}")
+            logger.info(f"Created schema for tenant {tenant_id}")
         else:
-            logger.debug(f"Schema already exists for tenant {tenant_id}")
+            logger.info(f"Schema already exists for tenant {tenant_id}")
 
-        # Set up the tenant with all necessary configurations
-        await setup_tenant(tenant_id)
+        token = CURRENT_TENANT_ID_CONTEXTVAR.set(tenant_id)
 
-        # Assign the tenant to the user
-        await assign_tenant_to_user(tenant_id, email)
+        # Await the Alembic migrations
+        await asyncio.to_thread(run_alembic_migrations, tenant_id)
+
+        with get_session_with_tenant(tenant_id=tenant_id) as db_session:
+            configure_default_api_keys(db_session)
+
+            current_search_settings = (
+                db_session.query(SearchSettings)
+                .filter_by(status=IndexModelStatus.FUTURE)
+                .first()
+            )
+            cohere_enabled = (
+                current_search_settings is not None
+                and current_search_settings.provider_type == EmbeddingProvider.COHERE
+            )
+            setup_onyx(db_session, tenant_id, cohere_enabled=cohere_enabled)
+
+        add_users_to_tenant([email], tenant_id)
+
+        with get_session_with_tenant(tenant_id=tenant_id) as db_session:
+            create_milestone_and_report(
+                user=None,
+                distinct_id=tenant_id,
+                event_type=MilestoneRecordType.TENANT_CREATED,
+                properties={
+                    "email": email,
+                },
+                db_session=db_session,
+            )
 
     except Exception as e:
         logger.exception(f"Failed to create tenant {tenant_id}")
         raise HTTPException(
             status_code=500, detail=f"Failed to create tenant: {str(e)}"
         )
+    finally:
+        if token is not None:
+            CURRENT_TENANT_ID_CONTEXTVAR.reset(token)
 
 
 async def notify_control_plane(
@@ -191,77 +183,42 @@ async def notify_control_plane(
 
 
 async def rollback_tenant_provisioning(tenant_id: str) -> None:
-    """
-    Logic to rollback tenant provisioning on data plane.
-    Handles each step independently to ensure maximum cleanup even if some steps fail.
-    """
+    # Logic to rollback tenant provisioning on data plane
     logger.info(f"Rolling back tenant provisioning for tenant_id: {tenant_id}")
-
-    # Track if any part of the rollback fails
-    rollback_errors = []
-
-    # 1. Try to drop the tenant's schema
     try:
+        # Drop the tenant's schema to rollback provisioning
         drop_schema(tenant_id)
-        logger.info(f"Successfully dropped schema for tenant {tenant_id}")
+
+        # Remove tenant mapping
+        with Session(get_sqlalchemy_engine()) as db_session:
+            db_session.query(UserTenantMapping).filter(
+                UserTenantMapping.tenant_id == tenant_id
+            ).delete()
+            db_session.commit()
     except Exception as e:
-        error_msg = f"Failed to drop schema for tenant {tenant_id}: {str(e)}"
-        logger.error(error_msg)
-        rollback_errors.append(error_msg)
-
-    # 2. Try to remove tenant mapping
-    try:
-        with get_session_with_shared_schema() as db_session:
-            db_session.begin()
-            try:
-                db_session.query(UserTenantMapping).filter(
-                    UserTenantMapping.tenant_id == tenant_id
-                ).delete()
-                db_session.commit()
-                logger.info(
-                    f"Successfully removed user mappings for tenant {tenant_id}"
-                )
-            except Exception as e:
-                db_session.rollback()
-                raise e
-    except Exception as e:
-        error_msg = f"Failed to remove user mappings for tenant {tenant_id}: {str(e)}"
-        logger.error(error_msg)
-        rollback_errors.append(error_msg)
-
-    # 3. If this tenant was in the available tenants table, remove it
-    try:
-        with get_session_with_shared_schema() as db_session:
-            db_session.begin()
-            try:
-                available_tenant = (
-                    db_session.query(AvailableTenant)
-                    .filter(AvailableTenant.tenant_id == tenant_id)
-                    .first()
-                )
-
-                if available_tenant:
-                    db_session.delete(available_tenant)
-                    db_session.commit()
-                    logger.info(
-                        f"Removed tenant {tenant_id} from available tenants table"
-                    )
-            except Exception as e:
-                db_session.rollback()
-                raise e
-    except Exception as e:
-        error_msg = f"Failed to remove tenant {tenant_id} from available tenants table: {str(e)}"
-        logger.error(error_msg)
-        rollback_errors.append(error_msg)
-
-    # Log summary of rollback operation
-    if rollback_errors:
-        logger.error(f"Tenant rollback completed with {len(rollback_errors)} errors")
-    else:
-        logger.info(f"Tenant rollback completed successfully for tenant {tenant_id}")
+        logger.error(f"Failed to rollback tenant provisioning: {e}")
 
 
 def configure_default_api_keys(db_session: Session) -> None:
+    if OPENAI_DEFAULT_API_KEY:
+        open_provider = LLMProviderUpsertRequest(
+            name="OpenAI",
+            provider=OPENAI_PROVIDER_NAME,
+            api_key=OPENAI_DEFAULT_API_KEY,
+            default_model_name="gpt-4",
+            fast_default_model_name="gpt-4o-mini",
+            model_names=OPEN_AI_MODEL_NAMES,
+        )
+        try:
+            full_provider = upsert_llm_provider(open_provider, db_session)
+            update_default_provider(full_provider.id, db_session)
+        except Exception as e:
+            logger.error(f"Failed to configure OpenAI provider: {e}")
+    else:
+        logger.error(
+            "OPENAI_DEFAULT_API_KEY not set, skipping OpenAI provider configuration"
+        )
+
     if ANTHROPIC_DEFAULT_API_KEY:
         anthropic_provider = LLMProviderUpsertRequest(
             name="Anthropic",
@@ -270,8 +227,6 @@ def configure_default_api_keys(db_session: Session) -> None:
             default_model_name="claude-3-7-sonnet-20250219",
             fast_default_model_name="claude-3-5-sonnet-20241022",
             model_names=ANTHROPIC_MODEL_NAMES,
-            display_model_names=["claude-3-5-sonnet-20241022"],
-            api_key_changed=True,
         )
         try:
             full_provider = upsert_llm_provider(anthropic_provider, db_session)
@@ -283,27 +238,6 @@ def configure_default_api_keys(db_session: Session) -> None:
             "ANTHROPIC_DEFAULT_API_KEY not set, skipping Anthropic provider configuration"
         )
 
-    if OPENAI_DEFAULT_API_KEY:
-        openai_provider = LLMProviderUpsertRequest(
-            name="OpenAI",
-            provider=OPENAI_PROVIDER_NAME,
-            api_key=OPENAI_DEFAULT_API_KEY,
-            default_model_name="gpt-4o",
-            fast_default_model_name="gpt-4o-mini",
-            model_names=OPEN_AI_MODEL_NAMES,
-            display_model_names=["o1", "o3-mini", "gpt-4o", "gpt-4o-mini"],
-            api_key_changed=True,
-        )
-        try:
-            full_provider = upsert_llm_provider(openai_provider, db_session)
-            update_default_provider(full_provider.id, db_session)
-        except Exception as e:
-            logger.error(f"Failed to configure OpenAI provider: {e}")
-    else:
-        logger.error(
-            "OPENAI_DEFAULT_API_KEY not set, skipping OpenAI provider configuration"
-        )
-
     if COHERE_DEFAULT_API_KEY:
         cloud_embedding_provider = CloudEmbeddingProviderCreationRequest(
             provider_type=EmbeddingProvider.COHERE,
@@ -413,154 +347,3 @@ async def delete_user_from_control_plane(tenant_id: str, email: str) -> None:
                 raise Exception(
                     f"Failed to delete tenant on control plane: {error_text}"
                 )
-
-
-def get_tenant_by_domain_from_control_plane(
-    domain: str,
-    tenant_id: str,
-) -> TenantByDomainResponse | None:
-    """
-    Fetches tenant information from the control plane based on the email domain.
-
-    Args:
-        domain: The email domain to search for (e.g., "example.com")
-
-    Returns:
-        A dictionary containing tenant information if found, None otherwise
-    """
-    token = generate_data_plane_token()
-    headers = {
-        "Authorization": f"Bearer {token}",
-        "Content-Type": "application/json",
-    }
-
-    try:
-        response = requests.get(
-            f"{CONTROL_PLANE_API_BASE_URL}/tenant-by-domain",
-            headers=headers,
-            json={"domain": domain, "tenant_id": tenant_id},
-        )
-
-        if response.status_code != 200:
-            logger.error(f"Control plane tenant lookup failed: {response.text}")
-            return None
-
-        response_data = response.json()
-        if not response_data:
-            return None
-
-        return TenantByDomainResponse(
-            tenant_id=response_data.get("tenant_id"),
-            number_of_users=response_data.get("number_of_users"),
-            creator_email=response_data.get("creator_email"),
-        )
-    except Exception as e:
-        logger.error(f"Error fetching tenant by domain: {str(e)}")
-        return None
-
-
-async def get_available_tenant() -> str | None:
-    """
-    Get an available pre-provisioned tenant from the NewAvailableTenant table.
-    Returns the tenant_id if one is available, None otherwise.
-    Uses row-level locking to prevent race conditions when multiple processes
-    try to get an available tenant simultaneously.
-    """
-    if not MULTI_TENANT:
-        return None
-
-    with get_session_with_shared_schema() as db_session:
-        try:
-            db_session.begin()
-
-            # Get the oldest available tenant with FOR UPDATE lock to prevent race conditions
-            available_tenant = (
-                db_session.query(AvailableTenant)
-                .order_by(AvailableTenant.date_created)
-                .with_for_update(skip_locked=True)  # Skip locked rows to avoid blocking
-                .first()
-            )
-
-            if available_tenant:
-                tenant_id = available_tenant.tenant_id
-                # Remove the tenant from the available tenants table
-                db_session.delete(available_tenant)
-                db_session.commit()
-                logger.info(f"Using pre-provisioned tenant {tenant_id}")
-                return tenant_id
-            else:
-                db_session.rollback()
-                return None
-        except Exception:
-            logger.exception("Error getting available tenant")
-            db_session.rollback()
-            return None
-
-
-async def setup_tenant(tenant_id: str) -> None:
-    """
-    Set up a tenant with all necessary configurations.
-    This is a centralized function that handles all tenant setup logic.
-    """
-    token = None
-    try:
-        token = CURRENT_TENANT_ID_CONTEXTVAR.set(tenant_id)
-
-        # Run Alembic migrations in a way that isolates it from the current event loop
-        # Create a new event loop for this synchronous operation
-        loop = asyncio.get_event_loop()
-        # Use run_in_executor which properly isolates the thread execution
-        await loop.run_in_executor(None, lambda: run_alembic_migrations(tenant_id))
-
-        # Configure the tenant with default settings
-        with get_session_with_tenant(tenant_id=tenant_id) as db_session:
-            # Configure default API keys
-            configure_default_api_keys(db_session)
-
-            # Set up Onyx with appropriate settings
-            current_search_settings = (
-                db_session.query(SearchSettings)
-                .filter_by(status=IndexModelStatus.FUTURE)
-                .first()
-            )
-            cohere_enabled = (
-                current_search_settings is not None
-                and current_search_settings.provider_type == EmbeddingProvider.COHERE
-            )
-            setup_onyx(db_session, tenant_id, cohere_enabled=cohere_enabled)
-
-    except Exception as e:
-        logger.exception(f"Failed to set up tenant {tenant_id}")
-        raise e
-    finally:
-        if token is not None:
-            CURRENT_TENANT_ID_CONTEXTVAR.reset(token)
-
-
-async def assign_tenant_to_user(
-    tenant_id: str, email: str, referral_source: str | None = None
-) -> None:
-    """
-    Assign a tenant to a user and perform necessary operations.
-    Uses transaction handling to ensure atomicity and includes retry logic
-    for control plane notifications.
-    """
-    # First, add the user to the tenant in a transaction
-
-    try:
-        add_users_to_tenant([email], tenant_id)
-
-        # Create milestone record in the same transaction context as the tenant assignment
-        with get_session_with_tenant(tenant_id=tenant_id) as db_session:
-            create_milestone_and_report(
-                user=None,
-                distinct_id=tenant_id,
-                event_type=MilestoneRecordType.TENANT_CREATED,
-                properties={
-                    "email": email,
-                },
-                db_session=db_session,
-            )
-    except Exception:
-        logger.exception(f"Failed to assign tenant {tenant_id} to user {email}")
-        raise Exception("Failed to assign tenant to user")

backend/ee/onyx/server/tenants/schema_management.py

@@ -74,21 +74,3 @@ def drop_schema(tenant_id: str) -> None:
             text("DROP SCHEMA IF EXISTS %(schema_name)s CASCADE"),
             {"schema_name": tenant_id},
         )
-
-
-def get_current_alembic_version(tenant_id: str) -> str:
-    """Get the current Alembic version for a tenant."""
-    from alembic.runtime.migration import MigrationContext
-    from sqlalchemy import text
-
-    engine = get_sqlalchemy_engine()
-
-    # Set the search path to the tenant's schema
-    with engine.connect() as connection:
-        connection.execute(text(f'SET search_path TO "{tenant_id}"'))
-
-        # Get the current version from the alembic_version table
-        context = MigrationContext.configure(connection)
-        current_rev = context.get_current_revision()
-
-    return current_rev or "head"

backend/ee/onyx/server/tenants/team_membership_api.py

@@ -1,67 +0,0 @@
-from fastapi import APIRouter
-from fastapi import Depends
-from fastapi import HTTPException
-from sqlalchemy.orm import Session
-
-from ee.onyx.server.tenants.provisioning import delete_user_from_control_plane
-from ee.onyx.server.tenants.user_mapping import remove_all_users_from_tenant
-from ee.onyx.server.tenants.user_mapping import remove_users_from_tenant
-from onyx.auth.users import current_admin_user
-from onyx.auth.users import User
-from onyx.db.auth import get_user_count
-from onyx.db.engine import get_session
-from onyx.db.users import delete_user_from_db
-from onyx.db.users import get_user_by_email
-from onyx.server.manage.models import UserByEmail
-from onyx.utils.logger import setup_logger
-from shared_configs.contextvars import get_current_tenant_id
-
-logger = setup_logger()
-
-router = APIRouter(prefix="/tenants")
-
-
-@router.post("/leave-team")
-async def leave_organization(
-    user_email: UserByEmail,
-    current_user: User | None = Depends(current_admin_user),
-    db_session: Session = Depends(get_session),
-) -> None:
-    tenant_id = get_current_tenant_id()
-
-    if current_user is None or current_user.email != user_email.user_email:
-        raise HTTPException(
-            status_code=403, detail="You can only leave the organization as yourself"
-        )
-
-    user_to_delete = get_user_by_email(user_email.user_email, db_session)
-    if user_to_delete is None:
-        raise HTTPException(status_code=404, detail="User not found")
-
-    num_admin_users = await get_user_count(only_admin_users=True)
-
-    should_delete_tenant = num_admin_users == 1
-
-    if should_delete_tenant:
-        logger.info(
-            "Last admin user is leaving the organization. Deleting tenant from control plane."
-        )
-        try:
-            await delete_user_from_control_plane(tenant_id, user_to_delete.email)
-            logger.debug("User deleted from control plane")
-        except Exception as e:
-            logger.exception(
-                f"Failed to delete user from control plane for tenant {tenant_id}: {e}"
-            )
-            raise HTTPException(
-                status_code=500,
-                detail=f"Failed to remove user from control plane: {str(e)}",
-            )
-
-    db_session.expunge(user_to_delete)
-    delete_user_from_db(user_to_delete, db_session)
-
-    if should_delete_tenant:
-        remove_all_users_from_tenant(tenant_id)
-    else:
-        remove_users_from_tenant([user_to_delete.email], tenant_id)

backend/ee/onyx/server/tenants/tenant_management_api.py

@@ -1,39 +0,0 @@
-from fastapi import APIRouter
-from fastapi import Depends
-
-from ee.onyx.server.tenants.models import TenantByDomainResponse
-from ee.onyx.server.tenants.provisioning import get_tenant_by_domain_from_control_plane
-from onyx.auth.users import current_user
-from onyx.auth.users import User
-from onyx.utils.logger import setup_logger
-from shared_configs.contextvars import get_current_tenant_id
-
-logger = setup_logger()
-
-router = APIRouter(prefix="/tenants")
-
-FORBIDDEN_COMMON_EMAIL_SUBSTRINGS = [
-    "gmail",
-    "outlook",
-    "yahoo",
-    "hotmail",
-    "icloud",
-    "msn",
-    "hotmail",
-    "hotmail.co.uk",
-]
-
-
-@router.get("/existing-team-by-domain")
-def get_existing_tenant_by_domain(
-    user: User | None = Depends(current_user),
-) -> TenantByDomainResponse | None:
-    if not user:
-        return None
-    domain = user.email.split("@")[1]
-    if any(substring in domain for substring in FORBIDDEN_COMMON_EMAIL_SUBSTRINGS):
-        return None
-
-    tenant_id = get_current_tenant_id()
-
-    return get_tenant_by_domain_from_control_plane(domain, tenant_id)

backend/ee/onyx/server/tenants/user_invitations_api.py

@@ -1,90 +0,0 @@
-from fastapi import APIRouter
-from fastapi import Depends
-from fastapi import HTTPException
-
-from ee.onyx.server.tenants.models import ApproveUserRequest
-from ee.onyx.server.tenants.models import PendingUserSnapshot
-from ee.onyx.server.tenants.models import RequestInviteRequest
-from ee.onyx.server.tenants.user_mapping import accept_user_invite
-from ee.onyx.server.tenants.user_mapping import approve_user_invite
-from ee.onyx.server.tenants.user_mapping import deny_user_invite
-from ee.onyx.server.tenants.user_mapping import invite_self_to_tenant
-from onyx.auth.invited_users import get_pending_users
-from onyx.auth.users import current_admin_user
-from onyx.auth.users import current_user
-from onyx.auth.users import User
-from onyx.utils.logger import setup_logger
-from shared_configs.contextvars import get_current_tenant_id
-
-logger = setup_logger()
-
-router = APIRouter(prefix="/tenants")
-
-
-@router.post("/users/invite/request")
-async def request_invite(
-    invite_request: RequestInviteRequest,
-    user: User | None = Depends(current_admin_user),
-) -> None:
-    if user is None:
-        raise HTTPException(status_code=401, detail="User not authenticated")
-    try:
-        invite_self_to_tenant(user.email, invite_request.tenant_id)
-    except Exception as e:
-        logger.exception(
-            f"Failed to invite self to tenant {invite_request.tenant_id}: {e}"
-        )
-        raise HTTPException(status_code=500, detail=str(e))
-
-
-@router.get("/users/pending")
-def list_pending_users(
-    _: User | None = Depends(current_admin_user),
-) -> list[PendingUserSnapshot]:
-    pending_emails = get_pending_users()
-    return [PendingUserSnapshot(email=email) for email in pending_emails]
-
-
-@router.post("/users/invite/approve")
-async def approve_user(
-    approve_user_request: ApproveUserRequest,
-    _: User | None = Depends(current_admin_user),
-) -> None:
-    tenant_id = get_current_tenant_id()
-    approve_user_invite(approve_user_request.email, tenant_id)
-
-
-@router.post("/users/invite/accept")
-async def accept_invite(
-    invite_request: RequestInviteRequest,
-    user: User | None = Depends(current_user),
-) -> None:
-    """
-    Accept an invitation to join a tenant.
-    """
-    if not user:
-        raise HTTPException(status_code=401, detail="Not authenticated")
-
-    try:
-        accept_user_invite(user.email, invite_request.tenant_id)
-    except Exception as e:
-        logger.exception(f"Failed to accept invite: {str(e)}")
-        raise HTTPException(status_code=500, detail="Failed to accept invitation")
-
-
-@router.post("/users/invite/deny")
-async def deny_invite(
-    invite_request: RequestInviteRequest,
-    user: User | None = Depends(current_user),
-) -> None:
-    """
-    Deny an invitation to join a tenant.
-    """
-    if not user:
-        raise HTTPException(status_code=401, detail="Not authenticated")
-
-    try:
-        deny_user_invite(user.email, invite_request.tenant_id)
-    except Exception as e:
-        logger.exception(f"Failed to deny invite: {str(e)}")
-        raise HTTPException(status_code=500, detail="Failed to deny invitation")

backend/ee/onyx/server/tenants/user_mapping.py

@@ -1,63 +1,34 @@
+import logging
+
 from fastapi_users import exceptions
 from sqlalchemy import select
+from sqlalchemy.orm import Session
 
-from onyx.auth.invited_users import get_invited_users
-from onyx.auth.invited_users import get_pending_users
-from onyx.auth.invited_users import write_invited_users
-from onyx.auth.invited_users import write_pending_users
-from onyx.db.engine import get_session_with_shared_schema
 from onyx.db.engine import get_session_with_tenant
+from onyx.db.engine import get_sqlalchemy_engine
 from onyx.db.models import UserTenantMapping
-from onyx.server.manage.models import TenantSnapshot
-from onyx.setup import setup_logger
 from shared_configs.configs import MULTI_TENANT
 from shared_configs.configs import POSTGRES_DEFAULT_SCHEMA
-from shared_configs.contextvars import CURRENT_TENANT_ID_CONTEXTVAR
 
-logger = setup_logger()
+logger = logging.getLogger(__name__)
 
 
 def get_tenant_id_for_email(email: str) -> str:
     if not MULTI_TENANT:
         return POSTGRES_DEFAULT_SCHEMA
     # Implement logic to get tenant_id from the mapping table
-    try:
-        with get_session_with_shared_schema() as db_session:
-            # First try to get an active tenant
-            result = db_session.execute(
-                select(UserTenantMapping).where(
-                    UserTenantMapping.email == email,
-                    UserTenantMapping.active == True,  # noqa: E712
-                )
-            )
-            mapping = result.scalar_one_or_none()
-            tenant_id = mapping.tenant_id if mapping else None
-
-            # If no active tenant found, try to get the first inactive one
-            if tenant_id is None:
-                result = db_session.execute(
-                    select(UserTenantMapping).where(
-                        UserTenantMapping.email == email,
-                        UserTenantMapping.active == False,  # noqa: E712
-                    )
-                )
-                mapping = result.scalar_one_or_none()
-                if mapping:
-                    # Mark this mapping as active
-                    mapping.active = True
-                    db_session.commit()
-                    tenant_id = mapping.tenant_id
-
-    except Exception as e:
-        logger.exception(f"Error getting tenant id for email {email}: {e}")
-        raise exceptions.UserNotExists()
+    with Session(get_sqlalchemy_engine()) as db_session:
+        result = db_session.execute(
+            select(UserTenantMapping.tenant_id).where(UserTenantMapping.email == email)
+        )
+        tenant_id = result.scalar_one_or_none()
     if tenant_id is None:
         raise exceptions.UserNotExists()
     return tenant_id
 
 
 def user_owns_a_tenant(email: str) -> bool:
-    with get_session_with_tenant(tenant_id=POSTGRES_DEFAULT_SCHEMA) as db_session:
+    with get_session_with_tenant(tenant_id=None) as db_session:
         result = (
             db_session.query(UserTenantMapping)
             .filter(UserTenantMapping.email == email)
@@ -67,60 +38,17 @@ def user_owns_a_tenant(email: str) -> bool:
 
 
 def add_users_to_tenant(emails: list[str], tenant_id: str) -> None:
-    """
-    Add users to a tenant with proper transaction handling.
-    Checks if users already have a tenant mapping to avoid duplicates.
-    If a user already has an active mapping to any tenant, the new mapping will be added as inactive.
-    """
-    with get_session_with_tenant(tenant_id=POSTGRES_DEFAULT_SCHEMA) as db_session:
+    with get_session_with_tenant(tenant_id=None) as db_session:
         try:
-            # Start a transaction
-            db_session.begin()
-
             for email in emails:
-                # Check if the user already has a mapping to this tenant
-                existing_mapping = (
-                    db_session.query(UserTenantMapping)
-                    .filter(
-                        UserTenantMapping.email == email,
-                        UserTenantMapping.tenant_id == tenant_id,
-                    )
-                    .with_for_update()
-                    .first()
-                )
-
-                # If user already has an active mapping, add this one as inactive
-                if not existing_mapping:
-                    # Check if the user already has an active mapping to any tenant
-                    has_active_mapping = (
-                        db_session.query(UserTenantMapping)
-                        .filter(
-                            UserTenantMapping.email == email,
-                            UserTenantMapping.active == True,  # noqa: E712
-                        )
-                        .first()
-                    )
-
-                    db_session.add(
-                        UserTenantMapping(
-                            email=email,
-                            tenant_id=tenant_id,
-                            active=False if has_active_mapping else True,
-                        )
-                    )
-
-            # Commit the transaction
-            db_session.commit()
-            logger.info(f"Successfully added users {emails} to tenant {tenant_id}")
-
+                db_session.add(UserTenantMapping(email=email, tenant_id=tenant_id))
         except Exception:
             logger.exception(f"Failed to add users to tenant {tenant_id}")
-            db_session.rollback()
-            raise
+        db_session.commit()
 
 
 def remove_users_from_tenant(emails: list[str], tenant_id: str) -> None:
-    with get_session_with_tenant(tenant_id=POSTGRES_DEFAULT_SCHEMA) as db_session:
+    with get_session_with_tenant(tenant_id=None) as db_session:
         try:
             mappings_to_delete = (
                 db_session.query(UserTenantMapping)
@@ -143,192 +71,8 @@ def remove_users_from_tenant(emails: list[str], tenant_id: str) -> None:
 
 
 def remove_all_users_from_tenant(tenant_id: str) -> None:
-    with get_session_with_tenant(tenant_id=POSTGRES_DEFAULT_SCHEMA) as db_session:
+    with get_session_with_tenant(tenant_id=None) as db_session:
         db_session.query(UserTenantMapping).filter(
             UserTenantMapping.tenant_id == tenant_id
         ).delete()
         db_session.commit()
-
-
-def invite_self_to_tenant(email: str, tenant_id: str) -> None:
-    token = CURRENT_TENANT_ID_CONTEXTVAR.set(tenant_id)
-    try:
-        pending_users = get_pending_users()
-        if email in pending_users:
-            return
-        write_pending_users(pending_users + [email])
-    finally:
-        CURRENT_TENANT_ID_CONTEXTVAR.reset(token)
-
-
-def approve_user_invite(email: str, tenant_id: str) -> None:
-    """
-    Approve a user invite to a tenant.
-    This will delete all existing records for this email and create a new mapping entry for the user in this tenant.
-    """
-    with get_session_with_shared_schema() as db_session:
-        # Delete all existing records for this email
-        db_session.query(UserTenantMapping).filter(
-            UserTenantMapping.email == email
-        ).delete()
-
-        # Create a new mapping entry for the user in this tenant
-        new_mapping = UserTenantMapping(email=email, tenant_id=tenant_id, active=True)
-        db_session.add(new_mapping)
-        db_session.commit()
-
-    # Also remove the user from pending users list
-    # Remove from pending users
-    pending_users = get_pending_users()
-    if email in pending_users:
-        pending_users.remove(email)
-        write_pending_users(pending_users)
-
-    # Add to invited users
-    invited_users = get_invited_users()
-    if email not in invited_users:
-        invited_users.append(email)
-        write_invited_users(invited_users)
-
-
-def accept_user_invite(email: str, tenant_id: str) -> None:
-    """
-    Accept an invitation to join a tenant.
-    This activates the user's mapping to the tenant.
-    """
-    with get_session_with_shared_schema() as db_session:
-        try:
-            # First check if there's an active mapping for this user and tenant
-            active_mapping = (
-                db_session.query(UserTenantMapping)
-                .filter(
-                    UserTenantMapping.email == email,
-                    UserTenantMapping.active == True,  # noqa: E712
-                )
-                .first()
-            )
-
-            # If an active mapping exists, delete it
-            if active_mapping:
-                db_session.delete(active_mapping)
-                logger.info(
-                    f"Deleted existing active mapping for user {email} in tenant {tenant_id}"
-                )
-
-            # Find the inactive mapping for this user and tenant
-            mapping = (
-                db_session.query(UserTenantMapping)
-                .filter(
-                    UserTenantMapping.email == email,
-                    UserTenantMapping.tenant_id == tenant_id,
-                    UserTenantMapping.active == False,  # noqa: E712
-                )
-                .first()
-            )
-
-            if mapping:
-                # Set all other mappings for this user to inactive
-                db_session.query(UserTenantMapping).filter(
-                    UserTenantMapping.email == email,
-                    UserTenantMapping.active == True,  # noqa: E712
-                ).update({"active": False})
-
-                # Activate this mapping
-                mapping.active = True
-                db_session.commit()
-                logger.info(f"User {email} accepted invitation to tenant {tenant_id}")
-            else:
-                logger.warning(
-                    f"No invitation found for user {email} in tenant {tenant_id}"
-                )
-
-        except Exception as e:
-            db_session.rollback()
-            logger.exception(
-                f"Failed to accept invitation for user {email} to tenant {tenant_id}: {str(e)}"
-            )
-            raise
-
-
-def deny_user_invite(email: str, tenant_id: str) -> None:
-    """
-    Deny an invitation to join a tenant.
-    This removes the user's mapping to the tenant.
-    """
-    with get_session_with_shared_schema() as db_session:
-        # Delete the mapping for this user and tenant
-        result = (
-            db_session.query(UserTenantMapping)
-            .filter(
-                UserTenantMapping.email == email,
-                UserTenantMapping.tenant_id == tenant_id,
-                UserTenantMapping.active == False,  # noqa: E712
-            )
-            .delete()
-        )
-
-        db_session.commit()
-        if result:
-            logger.info(f"User {email} denied invitation to tenant {tenant_id}")
-        else:
-            logger.warning(
-                f"No invitation found for user {email} in tenant {tenant_id}"
-            )
-    token = CURRENT_TENANT_ID_CONTEXTVAR.set(tenant_id)
-    try:
-        pending_users = get_invited_users()
-        if email in pending_users:
-            pending_users.remove(email)
-            write_invited_users(pending_users)
-    finally:
-        CURRENT_TENANT_ID_CONTEXTVAR.reset(token)
-
-
-def get_tenant_count(tenant_id: str) -> int:
-    """
-    Get the number of active users for this tenant
-    """
-    with get_session_with_shared_schema() as db_session:
-        # Count the number of active users for this tenant
-        user_count = (
-            db_session.query(UserTenantMapping)
-            .filter(
-                UserTenantMapping.tenant_id == tenant_id,
-                UserTenantMapping.active == True,  # noqa: E712
-            )
-            .count()
-        )
-
-        return user_count
-
-
-def get_tenant_invitation(email: str) -> TenantSnapshot | None:
-    """
-    Get the first tenant invitation for this user
-    """
-    with get_session_with_shared_schema() as db_session:
-        # Get the first tenant invitation for this user
-        invitation = (
-            db_session.query(UserTenantMapping)
-            .filter(
-                UserTenantMapping.email == email,
-                UserTenantMapping.active == False,  # noqa: E712
-            )
-            .first()
-        )
-
-        if invitation:
-            # Get the user count for this tenant
-            user_count = (
-                db_session.query(UserTenantMapping)
-                .filter(
-                    UserTenantMapping.tenant_id == invitation.tenant_id,
-                    UserTenantMapping.active == True,  # noqa: E712
-                )
-                .count()
-            )
-            return TenantSnapshot(
-                tenant_id=invitation.tenant_id, number_of_users=user_count
-            )
-
-        return None

backend/model_server/constants.py

@@ -3,11 +3,10 @@ from shared_configs.enums import EmbedTextType
 
 
 MODEL_WARM_UP_STRING = "hi " * 512
-INFORMATION_CONTENT_MODEL_WARM_UP_STRING = "hi " * 16
 DEFAULT_OPENAI_MODEL = "text-embedding-3-small"
 DEFAULT_COHERE_MODEL = "embed-english-light-v3.0"
 DEFAULT_VOYAGE_MODEL = "voyage-large-2-instruct"
-DEFAULT_VERTEX_MODEL = "text-embedding-005"
+DEFAULT_VERTEX_MODEL = "text-embedding-004"
 
 
 class EmbeddingModelTextType:

backend/model_server/custom_models.py

@@ -1,14 +1,11 @@
-import numpy as np
 import torch
 import torch.nn.functional as F
 from fastapi import APIRouter
 from huggingface_hub import snapshot_download  # type: ignore
-from setfit import SetFitModel  # type: ignore[import]
 from transformers import AutoTokenizer  # type: ignore
 from transformers import BatchEncoding  # type: ignore
 from transformers import PreTrainedTokenizer  # type: ignore
 
-from model_server.constants import INFORMATION_CONTENT_MODEL_WARM_UP_STRING
 from model_server.constants import MODEL_WARM_UP_STRING
 from model_server.onyx_torch_model import ConnectorClassifier
 from model_server.onyx_torch_model import HybridClassifier
@@ -16,22 +13,11 @@ from model_server.utils import simple_log_function_time
 from onyx.utils.logger import setup_logger
 from shared_configs.configs import CONNECTOR_CLASSIFIER_MODEL_REPO
 from shared_configs.configs import CONNECTOR_CLASSIFIER_MODEL_TAG
-from shared_configs.configs import (
-    INDEXING_INFORMATION_CONTENT_CLASSIFICATION_CUTOFF_LENGTH,
-)
-from shared_configs.configs import INDEXING_INFORMATION_CONTENT_CLASSIFICATION_MAX
-from shared_configs.configs import INDEXING_INFORMATION_CONTENT_CLASSIFICATION_MIN
-from shared_configs.configs import (
-    INDEXING_INFORMATION_CONTENT_CLASSIFICATION_TEMPERATURE,
-)
 from shared_configs.configs import INDEXING_ONLY
-from shared_configs.configs import INFORMATION_CONTENT_MODEL_TAG
-from shared_configs.configs import INFORMATION_CONTENT_MODEL_VERSION
 from shared_configs.configs import INTENT_MODEL_TAG
 from shared_configs.configs import INTENT_MODEL_VERSION
 from shared_configs.model_server_models import ConnectorClassificationRequest
 from shared_configs.model_server_models import ConnectorClassificationResponse
-from shared_configs.model_server_models import ContentClassificationPrediction
 from shared_configs.model_server_models import IntentRequest
 from shared_configs.model_server_models import IntentResponse
 
@@ -45,10 +31,6 @@ _CONNECTOR_CLASSIFIER_MODEL: ConnectorClassifier | None = None
 _INTENT_TOKENIZER: AutoTokenizer | None = None
 _INTENT_MODEL: HybridClassifier | None = None
 
-_INFORMATION_CONTENT_MODEL: SetFitModel | None = None
-
-_INFORMATION_CONTENT_MODEL_PROMPT_PREFIX: str = ""  # spec to model version!
-
 
 def get_connector_classifier_tokenizer() -> AutoTokenizer:
     global _CONNECTOR_CLASSIFIER_TOKENIZER
@@ -103,7 +85,7 @@ def get_intent_model_tokenizer() -> AutoTokenizer:
 
 def get_local_intent_model(
     model_name_or_path: str = INTENT_MODEL_VERSION,
-    tag: str | None = INTENT_MODEL_TAG,
+    tag: str = INTENT_MODEL_TAG,
 ) -> HybridClassifier:
     global _INTENT_MODEL
     if _INTENT_MODEL is None:
@@ -120,9 +102,7 @@ def get_local_intent_model(
             try:
                 # Attempt to download the model snapshot
                 logger.notice(f"Downloading model snapshot for {model_name_or_path}")
-                local_path = snapshot_download(
-                    repo_id=model_name_or_path, revision=tag, local_files_only=False
-                )
+                local_path = snapshot_download(repo_id=model_name_or_path, revision=tag)
                 _INTENT_MODEL = HybridClassifier.from_pretrained(local_path)
             except Exception as e:
                 logger.error(
@@ -132,44 +112,6 @@ def get_local_intent_model(
     return _INTENT_MODEL
 
 
-def get_local_information_content_model(
-    model_name_or_path: str = INFORMATION_CONTENT_MODEL_VERSION,
-    tag: str | None = INFORMATION_CONTENT_MODEL_TAG,
-) -> SetFitModel:
-    global _INFORMATION_CONTENT_MODEL
-    if _INFORMATION_CONTENT_MODEL is None:
-        try:
-            # Calculate where the cache should be, then load from local if available
-            logger.notice(
-                f"Loading content information model from local cache: {model_name_or_path}"
-            )
-            local_path = snapshot_download(
-                repo_id=model_name_or_path, revision=tag, local_files_only=True
-            )
-            _INFORMATION_CONTENT_MODEL = SetFitModel.from_pretrained(local_path)
-            logger.notice(
-                f"Loaded content information model from local cache: {local_path}"
-            )
-        except Exception as e:
-            logger.warning(f"Failed to load content information model directly: {e}")
-            try:
-                # Attempt to download the model snapshot
-                logger.notice(
-                    f"Downloading content information model snapshot for {model_name_or_path}"
-                )
-                local_path = snapshot_download(
-                    repo_id=model_name_or_path, revision=tag, local_files_only=False
-                )
-                _INFORMATION_CONTENT_MODEL = SetFitModel.from_pretrained(local_path)
-            except Exception as e:
-                logger.error(
-                    f"Failed to load content information model even after attempted snapshot download: {e}"
-                )
-                raise
-
-    return _INFORMATION_CONTENT_MODEL
-
-
 def tokenize_connector_classification_query(
     connectors: list[str],
     query: str,
@@ -253,13 +195,6 @@ def warm_up_intent_model() -> None:
     )
 
 
-def warm_up_information_content_model() -> None:
-    logger.notice("Warming up Content Model")  # TODO: add version if needed
-
-    information_content_model = get_local_information_content_model()
-    information_content_model(INFORMATION_CONTENT_MODEL_WARM_UP_STRING)
-
-
 @simple_log_function_time()
 def run_inference(tokens: BatchEncoding) -> tuple[list[float], list[float]]:
     intent_model = get_local_intent_model()
@@ -283,117 +218,6 @@ def run_inference(tokens: BatchEncoding) -> tuple[list[float], list[float]]:
     return intent_probabilities.tolist(), token_positive_probs
 
 
-@simple_log_function_time()
-def run_content_classification_inference(
-    text_inputs: list[str],
-) -> list[ContentClassificationPrediction]:
-    """
-    Assign a score to the segments in question. The model stored in get_local_information_content_model()
-    creates the 'model score' based on its training, and the scores are then converted to a 0.0-1.0 scale.
-    In the code outside of the model/inference model servers that score will be converted into the actual
-    boost factor.
-    """
-
-    def _prob_to_score(prob: float) -> float:
-        """
-        Conversion of base score to 0.0 - 1.0 score. Note that the min/max values depend on the model!
-        """
-        _MIN_BASE_SCORE = 0.25
-        _MAX_BASE_SCORE = 0.75
-        if prob < _MIN_BASE_SCORE:
-            raw_score = 0.0
-        elif prob < _MAX_BASE_SCORE:
-            raw_score = (prob - _MIN_BASE_SCORE) / (_MAX_BASE_SCORE - _MIN_BASE_SCORE)
-        else:
-            raw_score = 1.0
-        return (
-            INDEXING_INFORMATION_CONTENT_CLASSIFICATION_MIN
-            + (
-                INDEXING_INFORMATION_CONTENT_CLASSIFICATION_MAX
-                - INDEXING_INFORMATION_CONTENT_CLASSIFICATION_MIN
-            )
-            * raw_score
-        )
-
-    _BATCH_SIZE = 32
-    content_model = get_local_information_content_model()
-
-    # Process inputs in batches
-    all_output_classes: list[int] = []
-    all_base_output_probabilities: list[float] = []
-
-    for i in range(0, len(text_inputs), _BATCH_SIZE):
-        batch = text_inputs[i : i + _BATCH_SIZE]
-        batch_with_prefix = []
-        batch_indices = []
-
-        # Pre-allocate results for this batch
-        batch_output_classes: list[np.ndarray] = [np.array(1)] * len(batch)
-        batch_probabilities: list[np.ndarray] = [np.array(1.0)] * len(batch)
-
-        # Pre-process batch to handle long input exceptions
-        for j, text in enumerate(batch):
-            if len(text) == 0:
-                # if no input, treat as non-informative from the model's perspective
-                batch_output_classes[j] = np.array(0)
-                batch_probabilities[j] = np.array(0.0)
-                logger.warning("Input for Content Information Model is empty")
-
-            elif (
-                len(text.split())
-                <= INDEXING_INFORMATION_CONTENT_CLASSIFICATION_CUTOFF_LENGTH
-            ):
-                # if input is short, use the model
-                batch_with_prefix.append(
-                    _INFORMATION_CONTENT_MODEL_PROMPT_PREFIX + text
-                )
-                batch_indices.append(j)
-            else:
-                # if longer than cutoff, treat as informative (stay with default), but issue warning
-                logger.warning("Input for Content Information Model too long")
-
-        if batch_with_prefix:  # Only run model if we have valid inputs
-            # Get predictions for the batch
-            model_output_classes = content_model(batch_with_prefix)
-            model_output_probabilities = content_model.predict_proba(batch_with_prefix)
-
-            # Place results in the correct positions
-            for idx, batch_idx in enumerate(batch_indices):
-                batch_output_classes[batch_idx] = model_output_classes[idx].numpy()
-                batch_probabilities[batch_idx] = model_output_probabilities[idx][
-                    1
-                ].numpy()  # x[1] is prob of the positive class
-
-        all_output_classes.extend([int(x) for x in batch_output_classes])
-        all_base_output_probabilities.extend([float(x) for x in batch_probabilities])
-
-    logits = [
-        np.log(p / (1 - p)) if p != 0.0 and p != 1.0 else (100 if p == 1.0 else -100)
-        for p in all_base_output_probabilities
-    ]
-    scaled_logits = [
-        logit / INDEXING_INFORMATION_CONTENT_CLASSIFICATION_TEMPERATURE
-        for logit in logits
-    ]
-    output_probabilities_with_temp = [
-        np.exp(scaled_logit) / (1 + np.exp(scaled_logit))
-        for scaled_logit in scaled_logits
-    ]
-
-    prediction_scores = [
-        _prob_to_score(p_temp) for p_temp in output_probabilities_with_temp
-    ]
-
-    content_classification_predictions = [
-        ContentClassificationPrediction(
-            predicted_label=predicted_label, content_boost_factor=output_score
-        )
-        for predicted_label, output_score in zip(all_output_classes, prediction_scores)
-    ]
-
-    return content_classification_predictions
-
-
 def map_keywords(
     input_ids: torch.Tensor, tokenizer: AutoTokenizer, is_keyword: list[bool]
 ) -> list[str]:
@@ -538,10 +362,3 @@ async def process_analysis_request(
 
     is_keyword, keywords = run_analysis(intent_request)
     return IntentResponse(is_keyword=is_keyword, keywords=keywords)
-
-
-@router.post("/content-classification")
-async def process_content_classification_request(
-    content_classification_requests: list[str],
-) -> list[ContentClassificationPrediction]:
-    return run_content_classification_inference(content_classification_requests)

backend/model_server/encoders.py

@@ -5,7 +5,6 @@ from types import TracebackType
 from typing import cast
 from typing import Optional
 
-import aioboto3  # type: ignore
 import httpx
 import openai
 import vertexai  # type: ignore
@@ -29,13 +28,11 @@ from model_server.constants import DEFAULT_VERTEX_MODEL
 from model_server.constants import DEFAULT_VOYAGE_MODEL
 from model_server.constants import EmbeddingModelTextType
 from model_server.constants import EmbeddingProvider
-from model_server.utils import pass_aws_key
 from model_server.utils import simple_log_function_time
 from onyx.utils.logger import setup_logger
 from shared_configs.configs import API_BASED_EMBEDDING_TIMEOUT
 from shared_configs.configs import INDEXING_ONLY
 from shared_configs.configs import OPENAI_EMBEDDING_TIMEOUT
-from shared_configs.configs import VERTEXAI_EMBEDDING_LOCAL_BATCH_SIZE
 from shared_configs.enums import EmbedTextType
 from shared_configs.enums import RerankerProvider
 from shared_configs.model_server_models import Embedding
@@ -62,60 +59,6 @@ _OPENAI_MAX_INPUT_LEN = 2048
 # Cohere allows up to 96 embeddings in a single embedding calling
 _COHERE_MAX_INPUT_LEN = 96
 
-# Authentication error string constants
-_AUTH_ERROR_401 = "401"
-_AUTH_ERROR_UNAUTHORIZED = "unauthorized"
-_AUTH_ERROR_INVALID_API_KEY = "invalid api key"
-_AUTH_ERROR_PERMISSION = "permission"
-
-
-def is_authentication_error(error: Exception) -> bool:
-    """Check if an exception is related to authentication issues.
-
-    Args:
-        error: The exception to check
-
-    Returns:
-        bool: True if the error appears to be authentication-related
-    """
-    error_str = str(error).lower()
-    return (
-        _AUTH_ERROR_401 in error_str
-        or _AUTH_ERROR_UNAUTHORIZED in error_str
-        or _AUTH_ERROR_INVALID_API_KEY in error_str
-        or _AUTH_ERROR_PERMISSION in error_str
-    )
-
-
-def format_embedding_error(
-    error: Exception,
-    service_name: str,
-    model: str | None,
-    provider: EmbeddingProvider,
-    status_code: int | None = None,
-) -> str:
-    """
-    Format a standardized error string for embedding errors.
-    """
-    detail = f"Status {status_code}" if status_code else f"{type(error)}"
-
-    return (
-        f"{'HTTP error' if status_code else 'Exception'} embedding text with {service_name} - {detail}: "
-        f"Model: {model} "
-        f"Provider: {provider} "
-        f"Exception: {error}"
-    )
-
-
-# Custom exception for authentication errors
-class AuthenticationError(Exception):
-    """Raised when authentication fails with a provider."""
-
-    def __init__(self, provider: str, message: str = "API key is invalid or expired"):
-        self.provider = provider
-        self.message = message
-        super().__init__(f"{provider} authentication failed: {message}")
-
 
 class CloudEmbedding:
     def __init__(
@@ -135,7 +78,7 @@ class CloudEmbedding:
         self._closed = False
 
     async def _embed_openai(
-        self, texts: list[str], model: str | None, reduced_dimension: int | None
+        self, texts: list[str], model: str | None
     ) -> list[Embedding]:
         if not model:
             model = DEFAULT_OPENAI_MODEL
@@ -146,17 +89,27 @@ class CloudEmbedding:
         )
 
         final_embeddings: list[Embedding] = []
+        try:
+            for text_batch in batch_list(texts, _OPENAI_MAX_INPUT_LEN):
+                response = await client.embeddings.create(input=text_batch, model=model)
+                final_embeddings.extend(
+                    [embedding.embedding for embedding in response.data]
+                )
+            return final_embeddings
+        except Exception as e:
+            error_string = (
+                f"Exception embedding text with OpenAI - {type(e)}: "
+                f"Model: {model} "
+                f"Provider: {self.provider} "
+                f"Exception: {e}"
+            )
+            logger.error(error_string)
 
-        for text_batch in batch_list(texts, _OPENAI_MAX_INPUT_LEN):
-            response = await client.embeddings.create(
-                input=text_batch,
-                model=model,
-                dimensions=reduced_dimension or openai.NOT_GIVEN,
-            )
-            final_embeddings.extend(
-                [embedding.embedding for embedding in response.data]
-            )
-        return final_embeddings
+            # only log text when it's not an authentication error.
+            if not isinstance(e, openai.AuthenticationError):
+                logger.debug(f"Exception texts: {texts}")
+
+            raise RuntimeError(error_string)
 
     async def _embed_cohere(
         self, texts: list[str], model: str | None, embedding_type: str
@@ -195,6 +148,7 @@ class CloudEmbedding:
             input_type=embedding_type,
             truncation=True,
         )
+
         return response.embeddings
 
     async def _embed_azure(
@@ -224,24 +178,17 @@ class CloudEmbedding:
         vertexai.init(project=project_id, credentials=credentials)
         client = TextEmbeddingModel.from_pretrained(model)
 
-        inputs = [TextEmbeddingInput(text, embedding_type) for text in texts]
-
-        # Split into batches of 25 texts
-        max_texts_per_batch = VERTEXAI_EMBEDDING_LOCAL_BATCH_SIZE
-        batches = [
-            inputs[i : i + max_texts_per_batch]
-            for i in range(0, len(inputs), max_texts_per_batch)
-        ]
-
-        # Dispatch all embedding calls asynchronously at once
-        tasks = [
-            client.get_embeddings_async(batch, auto_truncate=True) for batch in batches
-        ]
-
-        # Wait for all tasks to complete in parallel
-        results = await asyncio.gather(*tasks)
-
-        return [embedding.values for batch in results for embedding in batch]
+        embeddings = await client.get_embeddings_async(
+            [
+                TextEmbeddingInput(
+                    text,
+                    embedding_type,
+                )
+                for text in texts
+            ],
+            auto_truncate=True,  # This is the default
+        )
+        return [embedding.values for embedding in embeddings]
 
     async def _embed_litellm_proxy(
         self, texts: list[str], model_name: str | None
@@ -276,53 +223,23 @@ class CloudEmbedding:
         text_type: EmbedTextType,
         model_name: str | None = None,
         deployment_name: str | None = None,
-        reduced_dimension: int | None = None,
     ) -> list[Embedding]:
-        try:
-            if self.provider == EmbeddingProvider.OPENAI:
-                return await self._embed_openai(texts, model_name, reduced_dimension)
-            elif self.provider == EmbeddingProvider.AZURE:
-                return await self._embed_azure(texts, f"azure/{deployment_name}")
-            elif self.provider == EmbeddingProvider.LITELLM:
-                return await self._embed_litellm_proxy(texts, model_name)
+        if self.provider == EmbeddingProvider.OPENAI:
+            return await self._embed_openai(texts, model_name)
+        elif self.provider == EmbeddingProvider.AZURE:
+            return await self._embed_azure(texts, f"azure/{deployment_name}")
+        elif self.provider == EmbeddingProvider.LITELLM:
+            return await self._embed_litellm_proxy(texts, model_name)
 
-            embedding_type = EmbeddingModelTextType.get_type(self.provider, text_type)
-            if self.provider == EmbeddingProvider.COHERE:
-                return await self._embed_cohere(texts, model_name, embedding_type)
-            elif self.provider == EmbeddingProvider.VOYAGE:
-                return await self._embed_voyage(texts, model_name, embedding_type)
-            elif self.provider == EmbeddingProvider.GOOGLE:
-                return await self._embed_vertex(texts, model_name, embedding_type)
-            else:
-                raise ValueError(f"Unsupported provider: {self.provider}")
-        except openai.AuthenticationError:
-            raise AuthenticationError(provider="OpenAI")
-        except httpx.HTTPStatusError as e:
-            if e.response.status_code == 401:
-                raise AuthenticationError(provider=str(self.provider))
-
-            error_string = format_embedding_error(
-                e,
-                str(self.provider),
-                model_name or deployment_name,
-                self.provider,
-                status_code=e.response.status_code,
-            )
-            logger.error(error_string)
-            logger.debug(f"Exception texts: {texts}")
-
-            raise RuntimeError(error_string)
-        except Exception as e:
-            if is_authentication_error(e):
-                raise AuthenticationError(provider=str(self.provider))
-
-            error_string = format_embedding_error(
-                e, str(self.provider), model_name or deployment_name, self.provider
-            )
-            logger.error(error_string)
-            logger.debug(f"Exception texts: {texts}")
-
-            raise RuntimeError(error_string)
+        embedding_type = EmbeddingModelTextType.get_type(self.provider, text_type)
+        if self.provider == EmbeddingProvider.COHERE:
+            return await self._embed_cohere(texts, model_name, embedding_type)
+        elif self.provider == EmbeddingProvider.VOYAGE:
+            return await self._embed_voyage(texts, model_name, embedding_type)
+        elif self.provider == EmbeddingProvider.GOOGLE:
+            return await self._embed_vertex(texts, model_name, embedding_type)
+        else:
+            raise ValueError(f"Unsupported provider: {self.provider}")
 
     @staticmethod
     def create(
@@ -409,7 +326,6 @@ async def embed_text(
     prefix: str | None,
     api_url: str | None,
     api_version: str | None,
-    reduced_dimension: int | None,
     gpu_type: str = "UNKNOWN",
 ) -> list[Embedding]:
     if not all(texts):
@@ -453,7 +369,6 @@ async def embed_text(
                 model_name=model_name,
                 deployment_name=deployment_name,
                 text_type=text_type,
-                reduced_dimension=reduced_dimension,
             )
 
         if any(embedding is None for embedding in embeddings):
@@ -525,7 +440,7 @@ async def local_rerank(query: str, docs: list[str], model_name: str) -> list[flo
     )
 
 
-async def cohere_rerank_api(
+async def cohere_rerank(
     query: str, docs: list[str], model_name: str, api_key: str
 ) -> list[float]:
     cohere_client = CohereAsyncClient(api_key=api_key)
@@ -535,45 +450,6 @@ async def cohere_rerank_api(
     return [result.relevance_score for result in sorted_results]
 
 
-async def cohere_rerank_aws(
-    query: str,
-    docs: list[str],
-    model_name: str,
-    region_name: str,
-    aws_access_key_id: str,
-    aws_secret_access_key: str,
-) -> list[float]:
-    session = aioboto3.Session(
-        aws_access_key_id=aws_access_key_id, aws_secret_access_key=aws_secret_access_key
-    )
-    async with session.client(
-        "bedrock-runtime", region_name=region_name
-    ) as bedrock_client:
-        body = json.dumps(
-            {
-                "query": query,
-                "documents": docs,
-                "api_version": 2,
-            }
-        )
-        # Invoke the Bedrock model asynchronously
-        response = await bedrock_client.invoke_model(
-            modelId=model_name,
-            accept="application/json",
-            contentType="application/json",
-            body=body,
-        )
-
-        # Read the response asynchronously
-        response_body = json.loads(await response["body"].read())
-
-        # Extract and sort the results
-        results = response_body.get("results", [])
-        sorted_results = sorted(results, key=lambda item: item["index"])
-
-        return [result["relevance_score"] for result in sorted_results]
-
-
 async def litellm_rerank(
     query: str, docs: list[str], api_url: str, model_name: str, api_key: str | None
 ) -> list[float]:
@@ -632,18 +508,10 @@ async def process_embed_request(
             text_type=embed_request.text_type,
             api_url=embed_request.api_url,
             api_version=embed_request.api_version,
-            reduced_dimension=embed_request.reduced_dimension,
             prefix=prefix,
             gpu_type=gpu_type,
         )
         return EmbedResponse(embeddings=embeddings)
-    except AuthenticationError as e:
-        # Handle authentication errors consistently
-        logger.error(f"Authentication error: {e.provider}")
-        raise HTTPException(
-            status_code=401,
-            detail=f"Authentication failed: {e.message}",
-        )
     except RateLimitError as e:
         raise HTTPException(
             status_code=429,
@@ -696,32 +564,15 @@ async def process_rerank_request(rerank_request: RerankRequest) -> RerankRespons
         elif rerank_request.provider_type == RerankerProvider.COHERE:
             if rerank_request.api_key is None:
                 raise RuntimeError("Cohere Rerank Requires an API Key")
-            sim_scores = await cohere_rerank_api(
+            sim_scores = await cohere_rerank(
                 query=rerank_request.query,
                 docs=rerank_request.documents,
                 model_name=rerank_request.model_name,
                 api_key=rerank_request.api_key,
             )
             return RerankResponse(scores=sim_scores)
-
-        elif rerank_request.provider_type == RerankerProvider.BEDROCK:
-            if rerank_request.api_key is None:
-                raise RuntimeError("Bedrock Rerank Requires an API Key")
-            aws_access_key_id, aws_secret_access_key, aws_region = pass_aws_key(
-                rerank_request.api_key
-            )
-            sim_scores = await cohere_rerank_aws(
-                query=rerank_request.query,
-                docs=rerank_request.documents,
-                model_name=rerank_request.model_name,
-                region_name=aws_region,
-                aws_access_key_id=aws_access_key_id,
-                aws_secret_access_key=aws_secret_access_key,
-            )
-            return RerankResponse(scores=sim_scores)
         else:
             raise ValueError(f"Unsupported provider: {rerank_request.provider_type}")
-
     except Exception as e:
         logger.exception(f"Error during reranking process:\n{str(e)}")
         raise HTTPException(

backend/model_server/main.py

@@ -1,4 +1,3 @@
-import logging
 import os
 import shutil
 from collections.abc import AsyncGenerator
@@ -9,21 +8,17 @@ import sentry_sdk
 import torch
 import uvicorn
 from fastapi import FastAPI
-from prometheus_fastapi_instrumentator import Instrumentator
 from sentry_sdk.integrations.fastapi import FastApiIntegration
 from sentry_sdk.integrations.starlette import StarletteIntegration
 from transformers import logging as transformer_logging  # type:ignore
 
 from model_server.custom_models import router as custom_models_router
-from model_server.custom_models import warm_up_information_content_model
 from model_server.custom_models import warm_up_intent_model
 from model_server.encoders import router as encoders_router
 from model_server.management_endpoints import router as management_router
 from model_server.utils import get_gpu_type
 from onyx import __version__
 from onyx.utils.logger import setup_logger
-from onyx.utils.logger import setup_uvicorn_logger
-from onyx.utils.middleware import add_onyx_request_id_middleware
 from shared_configs.configs import INDEXING_ONLY
 from shared_configs.configs import MIN_THREADS_ML_MODELS
 from shared_configs.configs import MODEL_SERVER_ALLOWED_HOST
@@ -40,12 +35,6 @@ transformer_logging.set_verbosity_error()
 
 logger = setup_logger()
 
-file_handlers = [
-    h for h in logger.logger.handlers if isinstance(h, logging.FileHandler)
-]
-
-setup_uvicorn_logger(shared_file_handlers=file_handlers)
-
 
 def _move_files_recursively(source: Path, dest: Path, overwrite: bool = False) -> None:
     """
@@ -75,31 +64,19 @@ async def lifespan(app: FastAPI) -> AsyncGenerator:
 
     app.state.gpu_type = gpu_type
 
-    try:
-        if TEMP_HF_CACHE_PATH.is_dir():
-            logger.notice("Moving contents of temp_huggingface to huggingface cache.")
-            _move_files_recursively(TEMP_HF_CACHE_PATH, HF_CACHE_PATH)
-            shutil.rmtree(TEMP_HF_CACHE_PATH, ignore_errors=True)
-            logger.notice("Moved contents of temp_huggingface to huggingface cache.")
-    except Exception as e:
-        logger.warning(
-            f"Error moving contents of temp_huggingface to huggingface cache: {e}. "
-            "This is not a critical error and the model server will continue to run."
-        )
+    if TEMP_HF_CACHE_PATH.is_dir():
+        logger.notice("Moving contents of temp_huggingface to huggingface cache.")
+        _move_files_recursively(TEMP_HF_CACHE_PATH, HF_CACHE_PATH)
+        shutil.rmtree(TEMP_HF_CACHE_PATH, ignore_errors=True)
+        logger.notice("Moved contents of temp_huggingface to huggingface cache.")
 
     torch.set_num_threads(max(MIN_THREADS_ML_MODELS, torch.get_num_threads()))
     logger.notice(f"Torch Threads: {torch.get_num_threads()}")
 
     if not INDEXING_ONLY:
-        logger.notice(
-            "The intent model should run on the model server. The information content model should not run here."
-        )
         warm_up_intent_model()
     else:
-        logger.notice(
-            "The content information model should run on the indexing model server. The intent model should not run here."
-        )
-        warm_up_information_content_model()
+        logger.notice("This model server should only run document indexing.")
 
     yield
 
@@ -122,15 +99,6 @@ def get_model_app() -> FastAPI:
     application.include_router(encoders_router)
     application.include_router(custom_models_router)
 
-    request_id_prefix = "INF"
-    if INDEXING_ONLY:
-        request_id_prefix = "IDX"
-
-    add_onyx_request_id_middleware(application, request_id_prefix, logger)
-
-    # Initialize and instrument the app
-    Instrumentator().instrument(application).expose(application)
-
     return application
 
 

backend/model_server/utils.py

@@ -70,32 +70,3 @@ def get_gpu_type() -> str:
         return GPUStatus.MAC_MPS
 
     return GPUStatus.NONE
-
-
-def pass_aws_key(api_key: str) -> tuple[str, str, str]:
-    """Parse AWS API key string into components.
-
-    Args:
-        api_key: String in format 'aws_ACCESSKEY_SECRETKEY_REGION'
-
-    Returns:
-        Tuple of (access_key, secret_key, region)
-
-    Raises:
-        ValueError: If key format is invalid
-    """
-    if not api_key.startswith("aws"):
-        raise ValueError("API key must start with 'aws' prefix")
-
-    parts = api_key.split("_")
-    if len(parts) != 4:
-        raise ValueError(
-            f"API key must be in format 'aws_ACCESSKEY_SECRETKEY_REGION', got {len(parts) - 1} parts"
-            "this is an onyx specific format for formatting the aws secrets for bedrock"
-        )
-
-    try:
-        _, aws_access_key_id, aws_secret_access_key, aws_region = parts
-        return aws_access_key_id, aws_secret_access_key, aws_region
-    except Exception as e:
-        raise ValueError(f"Failed to parse AWS key components: {str(e)}")

backend/onyx/access/access.py

@@ -18,7 +18,7 @@ def _get_access_for_document(
         document_id=document_id,
     )
 
-    doc_access = DocumentAccess.build(
+    return DocumentAccess.build(
         user_emails=info[1] if info and info[1] else [],
         user_groups=[],
         external_user_emails=[],
@@ -26,8 +26,6 @@ def _get_access_for_document(
         is_public=info[2] if info else False,
     )
 
-    return doc_access
-
 
 def get_access_for_document(
     document_id: str,
@@ -40,12 +38,12 @@ def get_access_for_document(
 
 
 def get_null_document_access() -> DocumentAccess:
-    return DocumentAccess.build(
-        user_emails=[],
-        user_groups=[],
+    return DocumentAccess(
+        user_emails=set(),
+        user_groups=set(),
         is_public=False,
-        external_user_emails=[],
-        external_user_group_ids=[],
+        external_user_emails=set(),
+        external_user_group_ids=set(),
     )
 
 
@@ -57,18 +55,19 @@ def _get_access_for_documents(
         db_session=db_session,
         document_ids=document_ids,
     )
-    doc_access = {}
-    for document_id, user_emails, is_public in document_access_info:
-        doc_access[document_id] = DocumentAccess.build(
-            user_emails=[email for email in user_emails if email],
+    doc_access = {
+        document_id: DocumentAccess(
+            user_emails=set([email for email in user_emails if email]),
             # MIT version will wipe all groups and external groups on update
-            user_groups=[],
+            user_groups=set(),
             is_public=is_public,
-            external_user_emails=[],
-            external_user_group_ids=[],
+            external_user_emails=set(),
+            external_user_group_ids=set(),
         )
+        for document_id, user_emails, is_public in document_access_info
+    }
 
-    # Sometimes the document has not been indexed by the indexing job yet, in those cases
+    # Sometimes the document has not be indexed by the indexing job yet, in those cases
     # the document does not exist and so we use least permissive. Specifically the EE version
     # checks the MIT version permissions and creates a superset. This ensures that this flow
     # does not fail even if the Document has not yet been indexed.

backend/onyx/access/models.py

@@ -15,28 +15,12 @@ class ExternalAccess:
     # Whether the document is public in the external system or Onyx
     is_public: bool
 
-    def __str__(self) -> str:
-        """Prevent extremely long logs"""
-
-        def truncate_set(s: set[str], max_len: int = 100) -> str:
-            s_str = str(s)
-            if len(s_str) > max_len:
-                return f"{s_str[:max_len]}... ({len(s)} items)"
-            return s_str
-
-        return (
-            f"ExternalAccess("
-            f"external_user_emails={truncate_set(self.external_user_emails)}, "
-            f"external_user_group_ids={truncate_set(self.external_user_group_ids)}, "
-            f"is_public={self.is_public})"
-        )
-
 
 @dataclass(frozen=True)
 class DocExternalAccess:
     """
     This is just a class to wrap the external access and the document ID
-    together. It's used for syncing document permissions to Vespa.
+    together. It's used for syncing document permissions to Redis.
     """
 
     external_access: ExternalAccess
@@ -72,45 +56,33 @@ class DocExternalAccess:
         )
 
 
-@dataclass(frozen=True, init=False)
+@dataclass(frozen=True)
 class DocumentAccess(ExternalAccess):
     # User emails for Onyx users, None indicates admin
     user_emails: set[str | None]
-
     # Names of user groups associated with this document
     user_groups: set[str]
 
-    external_user_emails: set[str]
-    external_user_group_ids: set[str]
-    is_public: bool
-
-    def __init__(self) -> None:
-        raise TypeError(
-            "Use `DocumentAccess.build(...)` instead of creating an instance directly."
-        )
-
     def to_acl(self) -> set[str]:
-        # the acl's emitted by this function are prefixed by type
-        # to get the native objects, access the member variables directly
-
-        acl_set: set[str] = set()
-        for user_email in self.user_emails:
-            if user_email:
-                acl_set.add(prefix_user_email(user_email))
-
-        for group_name in self.user_groups:
-            acl_set.add(prefix_user_group(group_name))
-
-        for external_user_email in self.external_user_emails:
-            acl_set.add(prefix_user_email(external_user_email))
-
-        for external_group_id in self.external_user_group_ids:
-            acl_set.add(prefix_external_group(external_group_id))
-
-        if self.is_public:
-            acl_set.add(PUBLIC_DOC_PAT)
-
-        return acl_set
+        return set(
+            [
+                prefix_user_email(user_email)
+                for user_email in self.user_emails
+                if user_email
+            ]
+            + [prefix_user_group(group_name) for group_name in self.user_groups]
+            + [
+                prefix_user_email(user_email)
+                for user_email in self.external_user_emails
+            ]
+            + [
+                # The group names are already prefixed by the source type
+                # This adds an additional prefix of "external_group:"
+                prefix_external_group(group_name)
+                for group_name in self.external_user_group_ids
+            ]
+            + ([PUBLIC_DOC_PAT] if self.is_public else [])
+        )
 
     @classmethod
     def build(
@@ -121,32 +93,29 @@ class DocumentAccess(ExternalAccess):
         external_user_group_ids: list[str],
         is_public: bool,
     ) -> "DocumentAccess":
-        """Don't prefix incoming data wth acl type, prefix on read from to_acl!"""
-
-        obj = object.__new__(cls)
-        object.__setattr__(
-            obj, "user_emails", {user_email for user_email in user_emails if user_email}
+        return cls(
+            external_user_emails={
+                prefix_user_email(external_email)
+                for external_email in external_user_emails
+            },
+            external_user_group_ids={
+                prefix_external_group(external_group_id)
+                for external_group_id in external_user_group_ids
+            },
+            user_emails={
+                prefix_user_email(user_email)
+                for user_email in user_emails
+                if user_email
+            },
+            user_groups=set(user_groups),
+            is_public=is_public,
         )
-        object.__setattr__(obj, "user_groups", set(user_groups))
-        object.__setattr__(
-            obj,
-            "external_user_emails",
-            {external_email for external_email in external_user_emails},
-        )
-        object.__setattr__(
-            obj,
-            "external_user_group_ids",
-            {external_group_id for external_group_id in external_user_group_ids},
-        )
-        object.__setattr__(obj, "is_public", is_public)
-
-        return obj
 
 
-default_public_access = DocumentAccess.build(
-    external_user_emails=[],
-    external_user_group_ids=[],
-    user_emails=[],
-    user_groups=[],
+default_public_access = DocumentAccess(
+    external_user_emails=set(),
+    external_user_group_ids=set(),
+    user_emails=set(),
+    user_groups=set(),
     is_public=True,
 )

backend/onyx/agents/agent_search/basic/utils.py

@@ -7,6 +7,7 @@ from langgraph.types import StreamWriter
 
 from onyx.agents.agent_search.shared_graph_utils.utils import write_custom_event
 from onyx.chat.models import LlmDoc
+from onyx.chat.models import OnyxContext
 from onyx.chat.stream_processing.answer_response_handler import AnswerResponseHandler
 from onyx.chat.stream_processing.answer_response_handler import CitationResponseHandler
 from onyx.chat.stream_processing.answer_response_handler import (
@@ -23,7 +24,7 @@ def process_llm_stream(
     should_stream_answer: bool,
     writer: StreamWriter,
     final_search_results: list[LlmDoc] | None = None,
-    displayed_search_results: list[LlmDoc] | None = None,
+    displayed_search_results: list[OnyxContext] | list[LlmDoc] | None = None,
 ) -> AIMessageChunk:
     tool_call_chunk = AIMessageChunk(content="")
 

backend/onyx/agents/agent_search/deep_search/initial/generate_individual_sub_answer/nodes/check_sub_answer.py

@@ -31,7 +31,6 @@ from onyx.agents.agent_search.shared_graph_utils.utils import (
     get_langgraph_node_log_string,
 )
 from onyx.agents.agent_search.shared_graph_utils.utils import parse_question_id
-from onyx.configs.agent_configs import AGENT_MAX_TOKENS_VALIDATION
 from onyx.configs.agent_configs import AGENT_TIMEOUT_CONNECT_LLM_SUBANSWER_CHECK
 from onyx.configs.agent_configs import AGENT_TIMEOUT_LLM_SUBANSWER_CHECK
 from onyx.llm.chat_llm import LLMRateLimitError
@@ -93,7 +92,6 @@ def check_sub_answer(
             fast_llm.invoke,
             prompt=msg,
             timeout_override=AGENT_TIMEOUT_CONNECT_LLM_SUBANSWER_CHECK,
-            max_tokens=AGENT_MAX_TOKENS_VALIDATION,
         )
 
         quality_str: str = cast(str, response.content)

backend/onyx/agents/agent_search/deep_search/initial/generate_individual_sub_answer/nodes/generate_sub_answer.py

@@ -46,7 +46,6 @@ from onyx.chat.models import StreamStopInfo
 from onyx.chat.models import StreamStopReason
 from onyx.chat.models import StreamType
 from onyx.configs.agent_configs import AGENT_MAX_ANSWER_CONTEXT_DOCS
-from onyx.configs.agent_configs import AGENT_MAX_TOKENS_SUBANSWER_GENERATION
 from onyx.configs.agent_configs import AGENT_TIMEOUT_CONNECT_LLM_SUBANSWER_GENERATION
 from onyx.configs.agent_configs import AGENT_TIMEOUT_LLM_SUBANSWER_GENERATION
 from onyx.llm.chat_llm import LLMRateLimitError
@@ -120,7 +119,6 @@ def generate_sub_answer(
             for message in fast_llm.stream(
                 prompt=msg,
                 timeout_override=AGENT_TIMEOUT_CONNECT_LLM_SUBANSWER_GENERATION,
-                max_tokens=AGENT_MAX_TOKENS_SUBANSWER_GENERATION,
             ):
                 # TODO: in principle, the answer here COULD contain images, but we don't support that yet
                 content = message.content

backend/onyx/agents/agent_search/deep_search/initial/generate_initial_answer/nodes/generate_initial_answer.py

@@ -43,7 +43,6 @@ from onyx.agents.agent_search.shared_graph_utils.models import LLMNodeErrorStrin
 from onyx.agents.agent_search.shared_graph_utils.operators import (
     dedup_inference_section_list,
 )
-from onyx.agents.agent_search.shared_graph_utils.utils import _should_restrict_tokens
 from onyx.agents.agent_search.shared_graph_utils.utils import (
     dispatch_main_answer_stop_info,
 )
@@ -63,7 +62,6 @@ from onyx.chat.models import StreamingError
 from onyx.configs.agent_configs import AGENT_ANSWER_GENERATION_BY_FAST_LLM
 from onyx.configs.agent_configs import AGENT_MAX_ANSWER_CONTEXT_DOCS
 from onyx.configs.agent_configs import AGENT_MAX_STREAMED_DOCS_FOR_INITIAL_ANSWER
-from onyx.configs.agent_configs import AGENT_MAX_TOKENS_ANSWER_GENERATION
 from onyx.configs.agent_configs import AGENT_MIN_ORIG_QUESTION_DOCS
 from onyx.configs.agent_configs import (
     AGENT_TIMEOUT_CONNECT_LLM_INITIAL_ANSWER_GENERATION,
@@ -155,8 +153,8 @@ def generate_initial_answer(
     )
     for tool_response in yield_search_responses(
         query=question,
-        get_retrieved_sections=lambda: answer_generation_documents.context_documents,
-        get_final_context_sections=lambda: answer_generation_documents.context_documents,
+        reranked_sections=answer_generation_documents.streaming_documents,
+        final_context_sections=answer_generation_documents.context_documents,
         search_query_info=query_info,
         get_section_relevance=lambda: relevance_list,
         search_tool=graph_config.tooling.search_tool,
@@ -280,9 +278,6 @@ def generate_initial_answer(
             for message in model.stream(
                 msg,
                 timeout_override=AGENT_TIMEOUT_CONNECT_LLM_INITIAL_ANSWER_GENERATION,
-                max_tokens=AGENT_MAX_TOKENS_ANSWER_GENERATION
-                if _should_restrict_tokens(model.config)
-                else None,
             ):
                 # TODO: in principle, the answer here COULD contain images, but we don't support that yet
                 content = message.content

backend/onyx/agents/agent_search/deep_search/initial/generate_sub_answers/nodes/decompose_orig_question.py

@@ -34,7 +34,6 @@ from onyx.chat.models import StreamStopInfo
 from onyx.chat.models import StreamStopReason
 from onyx.chat.models import StreamType
 from onyx.chat.models import SubQuestionPiece
-from onyx.configs.agent_configs import AGENT_MAX_TOKENS_SUBQUESTION_GENERATION
 from onyx.configs.agent_configs import AGENT_NUM_DOCS_FOR_DECOMPOSITION
 from onyx.configs.agent_configs import (
     AGENT_TIMEOUT_CONNECT_LLM_SUBQUESTION_GENERATION,
@@ -142,7 +141,6 @@ def decompose_orig_question(
             model.stream(
                 msg,
                 timeout_override=AGENT_TIMEOUT_CONNECT_LLM_SUBQUESTION_GENERATION,
-                max_tokens=AGENT_MAX_TOKENS_SUBQUESTION_GENERATION,
             ),
             dispatch_subquestion(0, writer),
             sep_callback=dispatch_subquestion_sep(0, writer),

backend/onyx/agents/agent_search/deep_search/main/nodes/compare_answers.py

@@ -33,7 +33,6 @@ from onyx.agents.agent_search.shared_graph_utils.utils import (
 )
 from onyx.agents.agent_search.shared_graph_utils.utils import write_custom_event
 from onyx.chat.models import RefinedAnswerImprovement
-from onyx.configs.agent_configs import AGENT_MAX_TOKENS_VALIDATION
 from onyx.configs.agent_configs import AGENT_TIMEOUT_CONNECT_LLM_COMPARE_ANSWERS
 from onyx.configs.agent_configs import AGENT_TIMEOUT_LLM_COMPARE_ANSWERS
 from onyx.llm.chat_llm import LLMRateLimitError
@@ -113,7 +112,6 @@ def compare_answers(
             model.invoke,
             prompt=msg,
             timeout_override=AGENT_TIMEOUT_CONNECT_LLM_COMPARE_ANSWERS,
-            max_tokens=AGENT_MAX_TOKENS_VALIDATION,
         )
 
     except (LLMTimeoutError, TimeoutError):

backend/onyx/agents/agent_search/deep_search/main/nodes/create_refined_sub_questions.py

@@ -43,7 +43,6 @@ from onyx.agents.agent_search.shared_graph_utils.utils import (
 from onyx.agents.agent_search.shared_graph_utils.utils import make_question_id
 from onyx.agents.agent_search.shared_graph_utils.utils import write_custom_event
 from onyx.chat.models import StreamingError
-from onyx.configs.agent_configs import AGENT_MAX_TOKENS_SUBQUESTION_GENERATION
 from onyx.configs.agent_configs import (
     AGENT_TIMEOUT_CONNECT_LLM_REFINED_SUBQUESTION_GENERATION,
 )
@@ -145,7 +144,6 @@ def create_refined_sub_questions(
             model.stream(
                 msg,
                 timeout_override=AGENT_TIMEOUT_CONNECT_LLM_REFINED_SUBQUESTION_GENERATION,
-                max_tokens=AGENT_MAX_TOKENS_SUBQUESTION_GENERATION,
             ),
             dispatch_subquestion(1, writer),
             sep_callback=dispatch_subquestion_sep(1, writer),

backend/onyx/agents/agent_search/deep_search/main/nodes/decide_refinement_need.py

@@ -50,7 +50,13 @@ def decide_refinement_need(
         )
     ]
 
-    return RequireRefinemenEvalUpdate(
-        require_refined_answer_eval=graph_config.behavior.allow_refinement and decision,
-        log_messages=log_messages,
-    )
+    if graph_config.behavior.allow_refinement:
+        return RequireRefinemenEvalUpdate(
+            require_refined_answer_eval=decision,
+            log_messages=log_messages,
+        )
+    else:
+        return RequireRefinemenEvalUpdate(
+            require_refined_answer_eval=False,
+            log_messages=log_messages,
+        )

backend/onyx/agents/agent_search/deep_search/main/nodes/extract_entities_terms.py

@@ -21,7 +21,6 @@ from onyx.agents.agent_search.shared_graph_utils.utils import format_docs
 from onyx.agents.agent_search.shared_graph_utils.utils import (
     get_langgraph_node_log_string,
 )
-from onyx.configs.agent_configs import AGENT_MAX_TOKENS_ENTITY_TERM_EXTRACTION
 from onyx.configs.agent_configs import (
     AGENT_TIMEOUT_CONNECT_LLM_ENTITY_TERM_EXTRACTION,
 )
@@ -97,7 +96,6 @@ def extract_entities_terms(
             fast_llm.invoke,
             prompt=msg,
             timeout_override=AGENT_TIMEOUT_CONNECT_LLM_ENTITY_TERM_EXTRACTION,
-            max_tokens=AGENT_MAX_TOKENS_ENTITY_TERM_EXTRACTION,
         )
 
         cleaned_response = (

backend/onyx/agents/agent_search/deep_search/main/nodes/generate_validate_refined_answer.py

@@ -46,7 +46,6 @@ from onyx.agents.agent_search.shared_graph_utils.models import RefinedAgentStats
 from onyx.agents.agent_search.shared_graph_utils.operators import (
     dedup_inference_section_list,
 )
-from onyx.agents.agent_search.shared_graph_utils.utils import _should_restrict_tokens
 from onyx.agents.agent_search.shared_graph_utils.utils import (
     dispatch_main_answer_stop_info,
 )
@@ -69,8 +68,6 @@ from onyx.chat.models import StreamingError
 from onyx.configs.agent_configs import AGENT_ANSWER_GENERATION_BY_FAST_LLM
 from onyx.configs.agent_configs import AGENT_MAX_ANSWER_CONTEXT_DOCS
 from onyx.configs.agent_configs import AGENT_MAX_STREAMED_DOCS_FOR_REFINED_ANSWER
-from onyx.configs.agent_configs import AGENT_MAX_TOKENS_ANSWER_GENERATION
-from onyx.configs.agent_configs import AGENT_MAX_TOKENS_VALIDATION
 from onyx.configs.agent_configs import AGENT_MIN_ORIG_QUESTION_DOCS
 from onyx.configs.agent_configs import (
     AGENT_TIMEOUT_CONNECT_LLM_REFINED_ANSWER_GENERATION,
@@ -182,8 +179,8 @@ def generate_validate_refined_answer(
     )
     for tool_response in yield_search_responses(
         query=question,
-        get_retrieved_sections=lambda: answer_generation_documents.context_documents,
-        get_final_context_sections=lambda: answer_generation_documents.context_documents,
+        reranked_sections=answer_generation_documents.streaming_documents,
+        final_context_sections=answer_generation_documents.context_documents,
         search_query_info=query_info,
         get_section_relevance=lambda: relevance_list,
         search_tool=graph_config.tooling.search_tool,
@@ -305,11 +302,7 @@ def generate_validate_refined_answer(
 
     def stream_refined_answer() -> list[str]:
         for message in model.stream(
-            msg,
-            timeout_override=AGENT_TIMEOUT_CONNECT_LLM_REFINED_ANSWER_GENERATION,
-            max_tokens=AGENT_MAX_TOKENS_ANSWER_GENERATION
-            if _should_restrict_tokens(model.config)
-            else None,
+            msg, timeout_override=AGENT_TIMEOUT_CONNECT_LLM_REFINED_ANSWER_GENERATION
         ):
             # TODO: in principle, the answer here COULD contain images, but we don't support that yet
             content = message.content
@@ -416,7 +409,6 @@ def generate_validate_refined_answer(
             validation_model.invoke,
             prompt=msg,
             timeout_override=AGENT_TIMEOUT_CONNECT_LLM_REFINED_ANSWER_VALIDATION,
-            max_tokens=AGENT_MAX_TOKENS_VALIDATION,
         )
         refined_answer_quality = binary_string_test_after_answer_separator(
             text=cast(str, validation_response.content),

backend/onyx/agents/agent_search/deep_search/main/operations.py

@@ -13,6 +13,7 @@ from onyx.chat.models import StreamStopInfo
 from onyx.chat.models import StreamStopReason
 from onyx.chat.models import StreamType
 from onyx.chat.models import SubQuestionPiece
+from onyx.context.search.models import IndexFilters
 from onyx.tools.models import SearchQueryInfo
 from onyx.utils.logger import setup_logger
 
@@ -143,6 +144,8 @@ def get_query_info(results: list[QueryRetrievalResult]) -> SearchQueryInfo:
         if result.query_info is not None:
             query_info = result.query_info
             break
-
-    assert query_info is not None, "must have query info"
-    return query_info
+    return query_info or SearchQueryInfo(
+        predicted_search=None,
+        final_filters=IndexFilters(access_control_list=None),
+        recency_bias_multiplier=1.0,
+    )

backend/onyx/agents/agent_search/deep_search/shared/expanded_retrieval/nodes/expand_queries.py

@@ -33,7 +33,6 @@ from onyx.agents.agent_search.shared_graph_utils.utils import (
     get_langgraph_node_log_string,
 )
 from onyx.agents.agent_search.shared_graph_utils.utils import parse_question_id
-from onyx.configs.agent_configs import AGENT_MAX_TOKENS_SUBQUERY_GENERATION
 from onyx.configs.agent_configs import (
     AGENT_TIMEOUT_CONNECT_LLM_QUERY_REWRITING_GENERATION,
 )
@@ -97,7 +96,6 @@ def expand_queries(
             model.stream(
                 prompt=msg,
                 timeout_override=AGENT_TIMEOUT_CONNECT_LLM_QUERY_REWRITING_GENERATION,
-                max_tokens=AGENT_MAX_TOKENS_SUBQUERY_GENERATION,
             ),
             dispatch_subquery(level, question_num, writer),
         )

backend/onyx/agents/agent_search/deep_search/shared/expanded_retrieval/nodes/format_results.py

@@ -56,8 +56,8 @@ def format_results(
         relevance_list = relevance_from_docs(reranked_documents)
         for tool_response in yield_search_responses(
             query=state.question,
-            get_retrieved_sections=lambda: reranked_documents,
-            get_final_context_sections=lambda: reranked_documents,
+            reranked_sections=state.retrieved_documents,
+            final_context_sections=reranked_documents,
             search_query_info=query_info,
             get_section_relevance=lambda: relevance_list,
             search_tool=graph_config.tooling.search_tool,

backend/onyx/agents/agent_search/deep_search/shared/expanded_retrieval/nodes/retrieve_documents.py

@@ -91,7 +91,7 @@ def retrieve_documents(
     retrieved_docs = retrieved_docs[:AGENT_MAX_QUERY_RETRIEVAL_RESULTS]
 
     if AGENT_RETRIEVAL_STATS:
-        pre_rerank_docs = callback_container[0] if callback_container else []
+        pre_rerank_docs = callback_container[0]
         fit_scores = get_fit_scores(
             pre_rerank_docs,
             retrieved_docs,

backend/onyx/agents/agent_search/deep_search/shared/expanded_retrieval/nodes/verify_documents.py

@@ -25,7 +25,6 @@ from onyx.agents.agent_search.shared_graph_utils.models import LLMNodeErrorStrin
 from onyx.agents.agent_search.shared_graph_utils.utils import (
     get_langgraph_node_log_string,
 )
-from onyx.configs.agent_configs import AGENT_MAX_TOKENS_VALIDATION
 from onyx.configs.agent_configs import AGENT_TIMEOUT_CONNECT_LLM_DOCUMENT_VERIFICATION
 from onyx.configs.agent_configs import AGENT_TIMEOUT_LLM_DOCUMENT_VERIFICATION
 from onyx.llm.chat_llm import LLMRateLimitError
@@ -94,7 +93,6 @@ def verify_documents(
             fast_llm.invoke,
             prompt=msg,
             timeout_override=AGENT_TIMEOUT_CONNECT_LLM_DOCUMENT_VERIFICATION,
-            max_tokens=AGENT_MAX_TOKENS_VALIDATION,
         )
 
         assert isinstance(response.content, str)

backend/onyx/agents/agent_search/orchestration/nodes/call_tool.py

@@ -44,9 +44,7 @@ def call_tool(
     tool = tool_choice.tool
     tool_args = tool_choice.tool_args
     tool_id = tool_choice.id
-    tool_runner = ToolRunner(
-        tool, tool_args, override_kwargs=tool_choice.search_tool_override_kwargs
-    )
+    tool_runner = ToolRunner(tool, tool_args)
     tool_kickoff = tool_runner.kickoff()
 
     emit_packet(tool_kickoff, writer)

backend/onyx/agents/agent_search/orchestration/nodes/choose_tool.py

@@ -15,17 +15,8 @@ from onyx.chat.tool_handling.tool_response_handler import get_tool_by_name
 from onyx.chat.tool_handling.tool_response_handler import (
     get_tool_call_for_non_tool_calling_llm_impl,
 )
-from onyx.context.search.preprocessing.preprocessing import query_analysis
-from onyx.context.search.retrieval.search_runner import get_query_embedding
-from onyx.tools.models import SearchToolOverrideKwargs
 from onyx.tools.tool import Tool
-from onyx.tools.tool_implementations.search.search_tool import SearchTool
 from onyx.utils.logger import setup_logger
-from onyx.utils.threadpool_concurrency import run_in_background
-from onyx.utils.threadpool_concurrency import TimeoutThread
-from onyx.utils.threadpool_concurrency import wait_on_background
-from onyx.utils.timing import log_function_time
-from shared_configs.model_server_models import Embedding
 
 logger = setup_logger()
 
@@ -34,7 +25,6 @@ logger = setup_logger()
 # and a function that handles extracting the necessary fields
 # from the state and config
 # TODO: fan-out to multiple tool call nodes? Make this configurable?
-@log_function_time(print_only=True)
 def choose_tool(
     state: ToolChoiceState,
     config: RunnableConfig,
@@ -47,31 +37,6 @@ def choose_tool(
     should_stream_answer = state.should_stream_answer
 
     agent_config = cast(GraphConfig, config["metadata"]["config"])
-
-    force_use_tool = agent_config.tooling.force_use_tool
-
-    embedding_thread: TimeoutThread[Embedding] | None = None
-    keyword_thread: TimeoutThread[tuple[bool, list[str]]] | None = None
-    override_kwargs: SearchToolOverrideKwargs | None = None
-    if (
-        not agent_config.behavior.use_agentic_search
-        and agent_config.tooling.search_tool is not None
-        and (
-            not force_use_tool.force_use or force_use_tool.tool_name == SearchTool.name
-        )
-    ):
-        override_kwargs = SearchToolOverrideKwargs()
-        # Run in a background thread to avoid blocking the main thread
-        embedding_thread = run_in_background(
-            get_query_embedding,
-            agent_config.inputs.search_request.query,
-            agent_config.persistence.db_session,
-        )
-        keyword_thread = run_in_background(
-            query_analysis,
-            agent_config.inputs.search_request.query,
-        )
-
     using_tool_calling_llm = agent_config.tooling.using_tool_calling_llm
     prompt_builder = state.prompt_snapshot or agent_config.inputs.prompt_builder
 
@@ -82,6 +47,7 @@ def choose_tool(
     tools = [
         tool for tool in (agent_config.tooling.tools or []) if tool.name in state.tools
     ]
+    force_use_tool = agent_config.tooling.force_use_tool
 
     tool, tool_args = None, None
     if force_use_tool.force_use and force_use_tool.args is not None:
@@ -105,22 +71,11 @@ def choose_tool(
     # If we have a tool and tool args, we are ready to request a tool call.
     # This only happens if the tool call was forced or we are using a non-tool calling LLM.
     if tool and tool_args:
-        if embedding_thread and tool.name == SearchTool._NAME:
-            # Wait for the embedding thread to finish
-            embedding = wait_on_background(embedding_thread)
-            assert override_kwargs is not None, "must have override kwargs"
-            override_kwargs.precomputed_query_embedding = embedding
-        if keyword_thread and tool.name == SearchTool._NAME:
-            is_keyword, keywords = wait_on_background(keyword_thread)
-            assert override_kwargs is not None, "must have override kwargs"
-            override_kwargs.precomputed_is_keyword = is_keyword
-            override_kwargs.precomputed_keywords = keywords
         return ToolChoiceUpdate(
             tool_choice=ToolChoice(
                 tool=tool,
                 tool_args=tool_args,
                 id=str(uuid4()),
-                search_tool_override_kwargs=override_kwargs,
             ),
         )
 
@@ -143,16 +98,8 @@ def choose_tool(
         # For tool calling LLMs, we want to insert the task prompt as part of this flow, this is because the LLM
         # may choose to not call any tools and just generate the answer, in which case the task prompt is needed.
         prompt=built_prompt,
-        tools=(
-            [tool.tool_definition() for tool in tools] or None
-            if using_tool_calling_llm
-            else None
-        ),
-        tool_choice=(
-            "required"
-            if tools and force_use_tool.force_use and using_tool_calling_llm
-            else None
-        ),
+        tools=[tool.tool_definition() for tool in tools] or None,
+        tool_choice=("required" if tools and force_use_tool.force_use else None),
         structured_response_format=structured_response_format,
     )
 
@@ -198,22 +145,10 @@ def choose_tool(
     logger.debug(f"Selected tool: {selected_tool.name}")
     logger.debug(f"Selected tool call request: {selected_tool_call_request}")
 
-    if embedding_thread and selected_tool.name == SearchTool._NAME:
-        # Wait for the embedding thread to finish
-        embedding = wait_on_background(embedding_thread)
-        assert override_kwargs is not None, "must have override kwargs"
-        override_kwargs.precomputed_query_embedding = embedding
-    if keyword_thread and selected_tool.name == SearchTool._NAME:
-        is_keyword, keywords = wait_on_background(keyword_thread)
-        assert override_kwargs is not None, "must have override kwargs"
-        override_kwargs.precomputed_is_keyword = is_keyword
-        override_kwargs.precomputed_keywords = keywords
-
     return ToolChoiceUpdate(
         tool_choice=ToolChoice(
             tool=selected_tool,
             tool_args=selected_tool_call_request["args"],
             id=selected_tool_call_request["id"],
-            search_tool_override_kwargs=override_kwargs,
         ),
     )

backend/onyx/agents/agent_search/orchestration/nodes/use_tool_response.py

@@ -9,21 +9,18 @@ from onyx.agents.agent_search.basic.states import BasicState
 from onyx.agents.agent_search.basic.utils import process_llm_stream
 from onyx.agents.agent_search.models import GraphConfig
 from onyx.chat.models import LlmDoc
+from onyx.chat.models import OnyxContexts
 from onyx.tools.tool_implementations.search.search_tool import (
-    SEARCH_RESPONSE_SUMMARY_ID,
+    SEARCH_DOC_CONTENT_ID,
 )
-from onyx.tools.tool_implementations.search.search_tool import SearchResponseSummary
-from onyx.tools.tool_implementations.search.search_utils import section_to_llm_doc
 from onyx.tools.tool_implementations.search_like_tool_utils import (
     FINAL_CONTEXT_DOCUMENTS_ID,
 )
 from onyx.utils.logger import setup_logger
-from onyx.utils.timing import log_function_time
 
 logger = setup_logger()
 
 
-@log_function_time(print_only=True)
 def basic_use_tool_response(
     state: BasicState, config: RunnableConfig, writer: StreamWriter = lambda _: None
 ) -> BasicOutput:
@@ -53,11 +50,11 @@ def basic_use_tool_response(
     for yield_item in tool_call_responses:
         if yield_item.id == FINAL_CONTEXT_DOCUMENTS_ID:
             final_search_results = cast(list[LlmDoc], yield_item.response)
-        elif yield_item.id == SEARCH_RESPONSE_SUMMARY_ID:
-            search_response_summary = cast(SearchResponseSummary, yield_item.response)
-            for section in search_response_summary.top_sections:
-                if section.center_chunk.document_id not in initial_search_results:
-                    initial_search_results.append(section_to_llm_doc(section))
+        elif yield_item.id == SEARCH_DOC_CONTENT_ID:
+            search_contexts = cast(OnyxContexts, yield_item.response).contexts
+            for doc in search_contexts:
+                if doc.document_id not in initial_search_results:
+                    initial_search_results.append(doc)
 
     new_tool_call_chunk = AIMessageChunk(content="")
     if not agent_config.behavior.skip_gen_ai_answer_generation: