fix(cli): add JSON type discriminator and URL validation in onboarding

Wrap --json events with {"type": "...", "event": {...}} so consumers
can distinguish event types without inspecting payload fields.

Validate server URL scheme (http/https) during onboarding setup.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

.cursor/skills/onyx-cli/SKILL.md

@@ -0,0 +1,153 @@
+---
+name: onyx-cli
+description: Query the Onyx knowledge base using the onyx-cli command. Use when the user wants to search company documents, ask questions about internal knowledge, query connected data sources, or look up information stored in Onyx.
+---
+
+# Onyx CLI — Agent Tool
+
+Onyx is an enterprise search and Gen-AI platform that connects to company documents, apps, and people. The `onyx-cli` CLI provides non-interactive commands to query the Onyx knowledge base and list available agents.
+
+## Prerequisites
+
+### 1. Check if installed
+
+```bash
+which onyx-cli
+```
+
+### 2. Install (if needed)
+
+**Primary — pip:**
+
+```bash
+pip install onyx-cli
+```
+
+**From source (Go):**
+
+```bash
+cd cli && go build -o onyx-cli . && sudo mv onyx-cli /usr/local/bin/
+```
+
+### 3. Check if configured
+
+The CLI is configured when `~/.config/onyx-cli/config.json` exists and contains an `api_key`. Check with:
+
+```bash
+test -s ~/.config/onyx-cli/config.json && echo "configured" || echo "not configured"
+```
+
+If unconfigured, you have two options:
+
+**Option A — Interactive setup (requires user input):**
+
+```bash
+onyx-cli configure
+```
+
+This prompts for the Onyx server URL and API key, tests the connection, and saves config.
+
+**Option B — Environment variables (non-interactive, preferred for agents):**
+
+```bash
+export ONYX_SERVER_URL="https://your-onyx-server.com"  # default: http://localhost:3000
+export ONYX_API_KEY="your-api-key"
+```
+
+Environment variables override the config file. If these are set, no config file is needed.
+
+| Variable | Required | Description |
+|----------|----------|-------------|
+| `ONYX_SERVER_URL` | No | Onyx server base URL (default: `http://localhost:3000`) |
+| `ONYX_API_KEY` | Yes | API key for authentication |
+| `ONYX_PERSONA_ID` | No | Default agent/persona ID |
+
+If neither the config file nor environment variables are set, tell the user that `onyx-cli` needs to be configured and ask them to either:
+- Run `onyx-cli configure` interactively, or
+- Set `ONYX_SERVER_URL` and `ONYX_API_KEY` environment variables
+
+## Commands
+
+### List available agents
+
+```bash
+onyx-cli agents
+```
+
+Prints a table of agent IDs, names, and descriptions. Use `--json` for structured output:
+
+```bash
+onyx-cli agents --json
+```
+
+Use agent IDs with `ask --agent-id` to query a specific agent.
+
+### Basic query (plain text output)
+
+```bash
+onyx-cli ask "What is our company's PTO policy?"
+```
+
+Streams the answer as plain text to stdout. Exit code 0 on success, non-zero on error.
+
+### JSON output (structured events)
+
+```bash
+onyx-cli ask --json "What authentication methods do we support?"
+```
+
+Outputs JSON-encoded parsed stream events (one object per line). Key event objects include message deltas, stop, errors, search-start, and citation payloads.
+
+| Event Type | Description |
+|------------|-------------|
+| `MessageDeltaEvent` | Content token — concatenate all `content` fields for the full answer |
+| `StopEvent` | Stream complete |
+| `ErrorEvent` | Error with `error` message field |
+| `SearchStartEvent` | Onyx started searching documents |
+| `CitationEvent` | Source citation with `citation_number` and `document_id` |
+
+### Specify an agent
+
+```bash
+onyx-cli ask --agent-id 5 "Summarize our Q4 roadmap"
+```
+
+Uses a specific Onyx agent/persona instead of the default.
+
+### All flags
+
+| Flag | Type | Description |
+|------|------|-------------|
+| `--agent-id` | int | Agent ID to use (overrides default) |
+| `--json` | bool | Output raw NDJSON events instead of plain text |
+
+## When to Use
+
+Use `onyx-cli ask` when:
+
+- The user asks about company-specific information (policies, docs, processes)
+- You need to search internal knowledge bases or connected data sources
+- The user references Onyx, asks you to "search Onyx", or wants to query their documents
+- You need context from company wikis, Confluence, Google Drive, Slack, or other connected sources
+
+Do NOT use when:
+
+- The question is about general programming knowledge (use your own knowledge)
+- The user is asking about code in the current repository (use grep/read tools)
+- The user hasn't mentioned Onyx and the question doesn't require internal company data
+
+## Examples
+
+```bash
+# Simple question
+onyx-cli ask "What are the steps to deploy to production?"
+
+# Get structured output for parsing
+onyx-cli ask --json "List all active API integrations"
+
+# Use a specialized agent
+onyx-cli ask --agent-id 3 "What were the action items from last week's standup?"
+
+# Pipe the answer into another command
+onyx-cli ask "What is the database schema for users?" | head -20
+```

.github/actions/build-integration-image/action.yml

@@ -54,7 +54,6 @@ runs:
       shell: bash
       env:
         RUNS_ON_ECR_CACHE: ${{ inputs.runs-on-ecr-cache }}
-        INTEGRATION_REPOSITORY: ${{ inputs.runs-on-ecr-cache }}
         TAG: nightly-llm-it-${{ inputs.run-id }}
         CACHE_SUFFIX: ${{ steps.format-branch.outputs.cache-suffix }}
         HEAD_SHA: ${{ inputs.github-sha }}

.github/workflows/deployment.yml

@@ -426,9 +426,8 @@ jobs:
             ONYX_VERSION=${{ github.ref_name }}
             NODE_OPTIONS=--max-old-space-size=8192
           cache-from: |
-            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:web-cache-amd64
-            type=registry,ref=${{ env.REGISTRY_IMAGE }}:edge
             type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
+            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:web-cache-amd64
           cache-to: |
             type=inline
             type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:web-cache-amd64,mode=max
@@ -500,9 +499,8 @@ jobs:
             ONYX_VERSION=${{ github.ref_name }}
             NODE_OPTIONS=--max-old-space-size=8192
           cache-from: |
-            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:web-cache-arm64
-            type=registry,ref=${{ env.REGISTRY_IMAGE }}:edge
             type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
+            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:web-cache-arm64
           cache-to: |
             type=inline
             type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:web-cache-arm64,mode=max
@@ -648,8 +646,8 @@ jobs:
             NEXT_PUBLIC_INCLUDE_ERROR_POPUP_SUPPORT_LINK=true
             NODE_OPTIONS=--max-old-space-size=8192
           cache-from: |
-            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:cloudweb-cache-amd64
             type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
+            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:cloudweb-cache-amd64
           cache-to: |
             type=inline
             type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:cloudweb-cache-amd64,mode=max
@@ -730,8 +728,8 @@ jobs:
             NEXT_PUBLIC_INCLUDE_ERROR_POPUP_SUPPORT_LINK=true
             NODE_OPTIONS=--max-old-space-size=8192
           cache-from: |
-            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:cloudweb-cache-arm64
             type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
+            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:cloudweb-cache-arm64
           cache-to: |
             type=inline
             type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:cloudweb-cache-arm64,mode=max
@@ -864,9 +862,8 @@ jobs:
           build-args: |
             ONYX_VERSION=${{ github.ref_name }}
           cache-from: |
-            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:backend-cache-amd64
-            type=registry,ref=${{ env.REGISTRY_IMAGE }}:edge
             type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
+            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:backend-cache-amd64
           cache-to: |
             type=inline
             type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:backend-cache-amd64,mode=max
@@ -937,9 +934,8 @@ jobs:
           build-args: |
             ONYX_VERSION=${{ github.ref_name }}
           cache-from: |
-            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:backend-cache-arm64
-            type=registry,ref=${{ env.REGISTRY_IMAGE }}:edge
             type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
+            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:backend-cache-arm64
           cache-to: |
             type=inline
             type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:backend-cache-arm64,mode=max
@@ -1076,8 +1072,8 @@ jobs:
             ONYX_VERSION=${{ github.ref_name }}
             ENABLE_CRAFT=true
           cache-from: |
-            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:backend-craft-cache-amd64
             type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
+            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:backend-craft-cache-amd64
           cache-to: |
             type=inline
             type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:backend-craft-cache-amd64,mode=max
@@ -1149,8 +1145,8 @@ jobs:
             ONYX_VERSION=${{ github.ref_name }}
             ENABLE_CRAFT=true
           cache-from: |
-            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:backend-craft-cache-arm64
             type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
+            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:backend-craft-cache-arm64
           cache-to: |
             type=inline
             type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:backend-craft-cache-arm64,mode=max
@@ -1291,9 +1287,8 @@ jobs:
           build-args: |
             ONYX_VERSION=${{ github.ref_name }}
           cache-from: |
-            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:model-server-cache-amd64
-            type=registry,ref=${{ env.REGISTRY_IMAGE }}:edge
             type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
+            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:model-server-cache-amd64
           cache-to: |
             type=inline
             type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:model-server-cache-amd64,mode=max
@@ -1371,9 +1366,8 @@ jobs:
           build-args: |
             ONYX_VERSION=${{ github.ref_name }}
           cache-from: |
-            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:model-server-cache-arm64
-            type=registry,ref=${{ env.REGISTRY_IMAGE }}:edge
             type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
+            type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:model-server-cache-arm64
           cache-to: |
             type=inline
             type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:model-server-cache-arm64,mode=max

.github/workflows/nightly-llm-provider-chat.yml

@@ -15,10 +15,6 @@ permissions:
 jobs:
   provider-chat-test:
     uses: ./.github/workflows/reusable-nightly-llm-provider-chat.yml
-    secrets: inherit
-    permissions:
-      contents: read
-      id-token: write
     with:
       openai_models: ${{ vars.NIGHTLY_LLM_OPENAI_MODELS }}
       anthropic_models: ${{ vars.NIGHTLY_LLM_ANTHROPIC_MODELS }}
@@ -29,6 +25,16 @@ jobs:
       ollama_models: ${{ vars.NIGHTLY_LLM_OLLAMA_MODELS }}
       openrouter_models: ${{ vars.NIGHTLY_LLM_OPENROUTER_MODELS }}
       strict: true
+    secrets:
+      openai_api_key: ${{ secrets.OPENAI_API_KEY }}
+      anthropic_api_key: ${{ secrets.ANTHROPIC_API_KEY }}
+      bedrock_api_key: ${{ secrets.BEDROCK_API_KEY }}
+      vertex_ai_custom_config_json: ${{ secrets.NIGHTLY_LLM_VERTEX_AI_CUSTOM_CONFIG_JSON }}
+      azure_api_key: ${{ secrets.AZURE_API_KEY }}
+      ollama_api_key: ${{ secrets.OLLAMA_API_KEY }}
+      openrouter_api_key: ${{ secrets.OPENROUTER_API_KEY }}
+      DOCKER_USERNAME: ${{ secrets.DOCKER_USERNAME }}
+      DOCKER_TOKEN: ${{ secrets.DOCKER_TOKEN }}
 
   notify-slack-on-failure:
     needs: [provider-chat-test]

.github/workflows/post-merge-beta-cherry-pick.yml

@@ -114,10 +114,8 @@ jobs:
 
       - name: Mark workflow as failed if cherry-pick failed
         if: steps.gate.outputs.should_cherrypick == 'true' && steps.run_cherry_pick.outputs.status == 'failure'
-        env:
-          CHERRY_PICK_REASON: ${{ steps.run_cherry_pick.outputs.reason }}
         run: |
-          echo "::error::Automated cherry-pick failed (${CHERRY_PICK_REASON})."
+          echo "::error::Automated cherry-pick failed (${{ steps.run_cherry_pick.outputs.reason }})."
           exit 1
 
   notify-slack-on-cherry-pick-failure:

.github/workflows/pr-external-dependency-unit-tests.yml

@@ -160,7 +160,7 @@ jobs:
           cd deployment/docker_compose
 
           # Get list of running containers
-          containers=$(docker compose -f docker-compose.yml -f docker-compose.dev.yml ps -q)
+          containers=$(docker compose -f docker-compose.yml -f docker-compose.dev.yml -f docker-compose.opensearch.yml ps -q)
 
           # Collect logs from each container
           for container in $containers; do

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

@@ -335,6 +335,7 @@ jobs:
           # TODO(Nik): https://linear.app/onyx-app/issue/ENG-1/update-test-infra-to-use-test-license
           LICENSE_ENFORCEMENT_ENABLED=false
           CHECK_TTL_MANAGEMENT_TASK_FREQUENCY_IN_HOURS=0.001
+          USE_LIGHTWEIGHT_BACKGROUND_WORKER=false
           EOF
           fi
 
@@ -470,13 +471,13 @@ jobs:
           path: ${{ github.workspace }}/docker-compose.log
       # ------------------------------------------------------------
 
-  onyx-lite-tests:
+  no-vectordb-tests:
     needs: [build-backend-image, build-integration-image]
     runs-on:
       [
         runs-on,
         runner=4cpu-linux-arm64,
-        "run-id=${{ github.run_id }}-onyx-lite-tests",
+        "run-id=${{ github.run_id }}-no-vectordb-tests",
         "extras=ecr-cache",
       ]
     timeout-minutes: 45
@@ -494,12 +495,13 @@ jobs:
           username: ${{ secrets.DOCKER_USERNAME }}
           password: ${{ secrets.DOCKER_TOKEN }}
 
-      - name: Create .env file for Onyx Lite Docker Compose
+      - name: Create .env file for no-vectordb Docker Compose
         env:
           ECR_CACHE: ${{ env.RUNS_ON_ECR_CACHE }}
           RUN_ID: ${{ github.run_id }}
         run: |
           cat <<EOF > deployment/docker_compose/.env
+          COMPOSE_PROFILES=s3-filestore
           ENABLE_PAID_ENTERPRISE_EDITION_FEATURES=true
           LICENSE_ENFORCEMENT_ENABLED=false
           AUTH_TYPE=basic
@@ -507,23 +509,28 @@ jobs:
           POSTGRES_USE_NULL_POOL=true
           REQUIRE_EMAIL_VERIFICATION=false
           DISABLE_TELEMETRY=true
+          DISABLE_VECTOR_DB=true
           ONYX_BACKEND_IMAGE=${ECR_CACHE}:integration-test-backend-test-${RUN_ID}
           INTEGRATION_TESTS_MODE=true
+          USE_LIGHTWEIGHT_BACKGROUND_WORKER=true
           EOF
 
-      # Start only the services needed for Onyx Lite (Postgres + API server)
-      - name: Start Docker containers (onyx-lite)
+      # Start only the services needed for no-vectordb mode (no Vespa, no model servers)
+      - name: Start Docker containers (no-vectordb)
         run: |
           cd deployment/docker_compose
-          docker compose -f docker-compose.yml -f docker-compose.onyx-lite.yml -f docker-compose.dev.yml up \
+          docker compose -f docker-compose.yml -f docker-compose.no-vectordb.yml -f docker-compose.dev.yml up \
             relational_db \
+            cache \
+            minio \
             api_server \
+            background \
             -d
-        id: start_docker_onyx_lite
+        id: start_docker_no_vectordb
 
       - name: Wait for services to be ready
         run: |
-          echo "Starting wait-for-service script (onyx-lite)..."
+          echo "Starting wait-for-service script (no-vectordb)..."
           start_time=$(date +%s)
           timeout=300
           while true; do
@@ -545,14 +552,14 @@ jobs:
             sleep 5
           done
 
-      - name: Run Onyx Lite Integration Tests
+      - name: Run No-VectorDB Integration Tests
         uses: nick-fields/retry@ce71cc2ab81d554ebbe88c79ab5975992d79ba08 # ratchet:nick-fields/retry@v3
         with:
           timeout_minutes: 20
           max_attempts: 3
           retry_wait_seconds: 10
           command: |
-            echo "Running onyx-lite integration tests..."
+            echo "Running no-vectordb integration tests..."
             docker run --rm --network onyx_default \
               --name test-runner \
               -e POSTGRES_HOST=relational_db \
@@ -563,38 +570,39 @@ jobs:
               -e DB_READONLY_PASSWORD=password \
               -e POSTGRES_POOL_PRE_PING=true \
               -e POSTGRES_USE_NULL_POOL=true \
+              -e REDIS_HOST=cache \
               -e API_SERVER_HOST=api_server \
               -e OPENAI_API_KEY=${OPENAI_API_KEY} \
               -e TEST_WEB_HOSTNAME=test-runner \
               ${{ env.RUNS_ON_ECR_CACHE }}:integration-test-${{ github.run_id }} \
               /app/tests/integration/tests/no_vectordb
 
-      - name: Dump API server logs (onyx-lite)
+      - name: Dump API server logs (no-vectordb)
         if: always()
         run: |
           cd deployment/docker_compose
-          docker compose -f docker-compose.yml -f docker-compose.onyx-lite.yml -f docker-compose.dev.yml \
-            logs --no-color api_server > $GITHUB_WORKSPACE/api_server_onyx_lite.log || true
+          docker compose -f docker-compose.yml -f docker-compose.no-vectordb.yml -f docker-compose.dev.yml \
+            logs --no-color api_server > $GITHUB_WORKSPACE/api_server_no_vectordb.log || true
 
-      - name: Dump all-container logs (onyx-lite)
+      - name: Dump all-container logs (no-vectordb)
         if: always()
         run: |
           cd deployment/docker_compose
-          docker compose -f docker-compose.yml -f docker-compose.onyx-lite.yml -f docker-compose.dev.yml \
-            logs --no-color > $GITHUB_WORKSPACE/docker-compose-onyx-lite.log || true
+          docker compose -f docker-compose.yml -f docker-compose.no-vectordb.yml -f docker-compose.dev.yml \
+            logs --no-color > $GITHUB_WORKSPACE/docker-compose-no-vectordb.log || true
 
-      - name: Upload logs (onyx-lite)
+      - name: Upload logs (no-vectordb)
         if: always()
         uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
         with:
-          name: docker-all-logs-onyx-lite
-          path: ${{ github.workspace }}/docker-compose-onyx-lite.log
+          name: docker-all-logs-no-vectordb
+          path: ${{ github.workspace }}/docker-compose-no-vectordb.log
 
-      - name: Stop Docker containers (onyx-lite)
+      - name: Stop Docker containers (no-vectordb)
         if: always()
         run: |
           cd deployment/docker_compose
-          docker compose -f docker-compose.yml -f docker-compose.onyx-lite.yml -f docker-compose.dev.yml down -v
+          docker compose -f docker-compose.yml -f docker-compose.no-vectordb.yml -f docker-compose.dev.yml down -v
 
   multitenant-tests:
     needs:
@@ -736,7 +744,7 @@ jobs:
     # NOTE: Github-hosted runners have about 20s faster queue times and are preferred here.
     runs-on: ubuntu-slim
     timeout-minutes: 45
-    needs: [integration-tests, onyx-lite-tests, multitenant-tests]
+    needs: [integration-tests, no-vectordb-tests, multitenant-tests]
     if: ${{ always() }}
     steps:
       - name: Check job status

.github/workflows/pr-python-checks.yml

@@ -8,7 +8,7 @@ on:
   pull_request:
     branches:
       - main
-      - "release/**"
+      - 'release/**'
   push:
     tags:
       - "v*.*.*"
@@ -21,13 +21,7 @@ jobs:
     # See https://runs-on.com/runners/linux/
     # Note: Mypy seems quite optimized for x64 compared to arm64.
     # Similarly, mypy is single-threaded and incremental, so 2cpu is sufficient.
-    runs-on:
-      [
-        runs-on,
-        runner=2cpu-linux-x64,
-        "run-id=${{ github.run_id }}-mypy-check",
-        "extras=s3-cache",
-      ]
+    runs-on: [runs-on, runner=2cpu-linux-x64, "run-id=${{ github.run_id }}-mypy-check", "extras=s3-cache"]
     timeout-minutes: 45
 
     steps:
@@ -58,14 +52,21 @@ jobs:
         if: ${{ vars.DISABLE_MYPY_CACHE != 'true' }}
         uses: runs-on/cache@50350ad4242587b6c8c2baa2e740b1bc11285ff4 # ratchet:runs-on/cache@v4
         with:
-          path: .mypy_cache
-          key: mypy-${{ runner.os }}-${{ github.base_ref || github.event.merge_group.base_ref || 'main' }}-${{ hashFiles('**/*.py', '**/*.pyi', 'pyproject.toml') }}
+          path: backend/.mypy_cache
+          key: mypy-${{ runner.os }}-${{ github.base_ref || github.event.merge_group.base_ref || 'main' }}-${{ hashFiles('**/*.py', '**/*.pyi', 'backend/pyproject.toml') }}
           restore-keys: |
             mypy-${{ runner.os }}-${{ github.base_ref || github.event.merge_group.base_ref || 'main' }}-
             mypy-${{ runner.os }}-
 
       - name: Run MyPy
+        working-directory: ./backend
         env:
           MYPY_FORCE_COLOR: 1
           TERM: xterm-256color
         run: mypy .
+
+      - name: Run MyPy (tools/)
+        env:
+          MYPY_FORCE_COLOR: 1
+          TERM: xterm-256color
+        run: mypy tools/

.github/workflows/reusable-nightly-llm-provider-chat.yml

@@ -48,10 +48,28 @@ on:
         required: false
         default: true
         type: boolean
+    secrets:
+      openai_api_key:
+        required: false
+      anthropic_api_key:
+        required: false
+      bedrock_api_key:
+        required: false
+      vertex_ai_custom_config_json:
+        required: false
+      azure_api_key:
+        required: false
+      ollama_api_key:
+        required: false
+      openrouter_api_key:
+        required: false
+      DOCKER_USERNAME:
+        required: true
+      DOCKER_TOKEN:
+        required: true
 
 permissions:
   contents: read
-  id-token: write
 
 jobs:
   build-backend-image:
@@ -63,7 +81,6 @@ jobs:
         "extras=ecr-cache",
       ]
     timeout-minutes: 45
-    environment: ci-protected
     steps:
       - uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
 
@@ -72,19 +89,6 @@ jobs:
         with:
           persist-credentials: false
 
-      - name: Configure AWS credentials
-        uses: aws-actions/configure-aws-credentials@61815dcd50bd041e203e49132bacad1fd04d2708
-        with:
-          role-to-assume: ${{ secrets.AWS_OIDC_ROLE_ARN }}
-          aws-region: us-east-2
-
-      - name: Get AWS Secrets
-        uses: aws-actions/aws-secretsmanager-get-secrets@a9a7eb4e2f2871d30dc5b892576fde60a2ecc802
-        with:
-          secret-ids: |
-            DOCKER_USERNAME, test/docker-username
-            DOCKER_TOKEN, test/docker-token
-
       - name: Build backend image
         uses: ./.github/actions/build-backend-image
         with:
@@ -93,8 +97,8 @@ jobs:
           pr-number: ${{ github.event.pull_request.number }}
           github-sha: ${{ github.sha }}
           run-id: ${{ github.run_id }}
-          docker-username: ${{ env.DOCKER_USERNAME }}
-          docker-token: ${{ env.DOCKER_TOKEN }}
+          docker-username: ${{ secrets.DOCKER_USERNAME }}
+          docker-token: ${{ secrets.DOCKER_TOKEN }}
           docker-no-cache: ${{ vars.DOCKER_NO_CACHE == 'true' && 'true' || 'false' }}
 
   build-model-server-image:
@@ -106,7 +110,6 @@ jobs:
         "extras=ecr-cache",
       ]
     timeout-minutes: 45
-    environment: ci-protected
     steps:
       - uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
 
@@ -115,19 +118,6 @@ jobs:
         with:
           persist-credentials: false
 
-      - name: Configure AWS credentials
-        uses: aws-actions/configure-aws-credentials@61815dcd50bd041e203e49132bacad1fd04d2708
-        with:
-          role-to-assume: ${{ secrets.AWS_OIDC_ROLE_ARN }}
-          aws-region: us-east-2
-
-      - name: Get AWS Secrets
-        uses: aws-actions/aws-secretsmanager-get-secrets@a9a7eb4e2f2871d30dc5b892576fde60a2ecc802
-        with:
-          secret-ids: |
-            DOCKER_USERNAME, test/docker-username
-            DOCKER_TOKEN, test/docker-token
-
       - name: Build model server image
         uses: ./.github/actions/build-model-server-image
         with:
@@ -136,8 +126,8 @@ jobs:
           pr-number: ${{ github.event.pull_request.number }}
           github-sha: ${{ github.sha }}
           run-id: ${{ github.run_id }}
-          docker-username: ${{ env.DOCKER_USERNAME }}
-          docker-token: ${{ env.DOCKER_TOKEN }}
+          docker-username: ${{ secrets.DOCKER_USERNAME }}
+          docker-token: ${{ secrets.DOCKER_TOKEN }}
 
   build-integration-image:
     runs-on:
@@ -148,7 +138,6 @@ jobs:
         "extras=ecr-cache",
       ]
     timeout-minutes: 45
-    environment: ci-protected
     steps:
       - uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
 
@@ -157,19 +146,6 @@ jobs:
         with:
           persist-credentials: false
 
-      - name: Configure AWS credentials
-        uses: aws-actions/configure-aws-credentials@61815dcd50bd041e203e49132bacad1fd04d2708
-        with:
-          role-to-assume: ${{ secrets.AWS_OIDC_ROLE_ARN }}
-          aws-region: us-east-2
-
-      - name: Get AWS Secrets
-        uses: aws-actions/aws-secretsmanager-get-secrets@a9a7eb4e2f2871d30dc5b892576fde60a2ecc802
-        with:
-          secret-ids: |
-            DOCKER_USERNAME, test/docker-username
-            DOCKER_TOKEN, test/docker-token
-
       - name: Build integration image
         uses: ./.github/actions/build-integration-image
         with:
@@ -178,8 +154,8 @@ jobs:
           pr-number: ${{ github.event.pull_request.number }}
           github-sha: ${{ github.sha }}
           run-id: ${{ github.run_id }}
-          docker-username: ${{ env.DOCKER_USERNAME }}
-          docker-token: ${{ env.DOCKER_TOKEN }}
+          docker-username: ${{ secrets.DOCKER_USERNAME }}
+          docker-token: ${{ secrets.DOCKER_TOKEN }}
 
   provider-chat-test:
     needs:
@@ -194,56 +170,56 @@ jobs:
         include:
           - provider: openai
             models: ${{ inputs.openai_models }}
-            api_key_env: OPENAI_API_KEY
-            custom_config_env: ""
+            api_key_secret: openai_api_key
+            custom_config_secret: ""
             api_base: ""
             api_version: ""
             deployment_name: ""
             required: true
           - provider: anthropic
             models: ${{ inputs.anthropic_models }}
-            api_key_env: ANTHROPIC_API_KEY
-            custom_config_env: ""
+            api_key_secret: anthropic_api_key
+            custom_config_secret: ""
             api_base: ""
             api_version: ""
             deployment_name: ""
             required: true
           - provider: bedrock
             models: ${{ inputs.bedrock_models }}
-            api_key_env: BEDROCK_API_KEY
-            custom_config_env: ""
+            api_key_secret: bedrock_api_key
+            custom_config_secret: ""
             api_base: ""
             api_version: ""
             deployment_name: ""
             required: false
           - provider: vertex_ai
             models: ${{ inputs.vertex_ai_models }}
-            api_key_env: ""
-            custom_config_env: NIGHTLY_LLM_VERTEX_AI_CUSTOM_CONFIG_JSON
+            api_key_secret: ""
+            custom_config_secret: vertex_ai_custom_config_json
             api_base: ""
             api_version: ""
             deployment_name: ""
             required: false
           - provider: azure
             models: ${{ inputs.azure_models }}
-            api_key_env: AZURE_API_KEY
-            custom_config_env: ""
+            api_key_secret: azure_api_key
+            custom_config_secret: ""
             api_base: ${{ inputs.azure_api_base }}
             api_version: "2025-04-01-preview"
             deployment_name: ""
             required: false
           - provider: ollama_chat
             models: ${{ inputs.ollama_models }}
-            api_key_env: OLLAMA_API_KEY
-            custom_config_env: ""
+            api_key_secret: ollama_api_key
+            custom_config_secret: ""
             api_base: "https://ollama.com"
             api_version: ""
             deployment_name: ""
             required: false
           - provider: openrouter
             models: ${{ inputs.openrouter_models }}
-            api_key_env: OPENROUTER_API_KEY
-            custom_config_env: ""
+            api_key_secret: openrouter_api_key
+            custom_config_secret: ""
             api_base: "https://openrouter.ai/api/v1"
             api_version: ""
             deployment_name: ""
@@ -254,7 +230,6 @@ jobs:
       - "run-id=${{ github.run_id }}-nightly-${{ matrix.provider }}-provider-chat-test"
       - extras=ecr-cache
     timeout-minutes: 45
-    environment: ci-protected
     steps:
       - uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
 
@@ -263,43 +238,21 @@ jobs:
         with:
           persist-credentials: false
 
-      - name: Configure AWS credentials
-        uses: aws-actions/configure-aws-credentials@61815dcd50bd041e203e49132bacad1fd04d2708
-        with:
-          role-to-assume: ${{ secrets.AWS_OIDC_ROLE_ARN }}
-          aws-region: us-east-2
-
-      - name: Get AWS Secrets
-        uses: aws-actions/aws-secretsmanager-get-secrets@a9a7eb4e2f2871d30dc5b892576fde60a2ecc802
-        with:
-          # Keep JSON values unparsed so vertex custom config is passed as raw JSON.
-          parse-json-secrets: false
-          secret-ids: |
-            DOCKER_USERNAME, test/docker-username
-            DOCKER_TOKEN, test/docker-token
-            OPENAI_API_KEY, test/openai-api-key
-            ANTHROPIC_API_KEY, test/anthropic-api-key
-            BEDROCK_API_KEY, test/bedrock-api-key
-            NIGHTLY_LLM_VERTEX_AI_CUSTOM_CONFIG_JSON, test/nightly-llm-vertex-ai-custom-config-json
-            AZURE_API_KEY, test/azure-api-key
-            OLLAMA_API_KEY, test/ollama-api-key
-            OPENROUTER_API_KEY, test/openrouter-api-key
-
       - name: Run nightly provider chat test
         uses: ./.github/actions/run-nightly-provider-chat-test
         with:
           provider: ${{ matrix.provider }}
           models: ${{ matrix.models }}
-          provider-api-key: ${{ matrix.api_key_env && env[matrix.api_key_env] || '' }}
+          provider-api-key: ${{ matrix.api_key_secret && secrets[matrix.api_key_secret] || '' }}
           strict: ${{ inputs.strict && 'true' || 'false' }}
           api-base: ${{ matrix.api_base }}
           api-version: ${{ matrix.api_version }}
           deployment-name: ${{ matrix.deployment_name }}
-          custom-config-json: ${{ matrix.custom_config_env && env[matrix.custom_config_env] || '' }}
+          custom-config-json: ${{ matrix.custom_config_secret && secrets[matrix.custom_config_secret] || '' }}
           runs-on-ecr-cache: ${{ env.RUNS_ON_ECR_CACHE }}
           run-id: ${{ github.run_id }}
-          docker-username: ${{ env.DOCKER_USERNAME }}
-          docker-token: ${{ env.DOCKER_TOKEN }}
+          docker-username: ${{ secrets.DOCKER_USERNAME }}
+          docker-token: ${{ secrets.DOCKER_TOKEN }}
 
       - name: Dump API server logs
         if: always()

.pre-commit-config.yaml

@@ -122,7 +122,7 @@ repos:
     rev: 9f61b0f53f80672872fced07b6874397c3ed197b # frozen: v2.7.2
     hooks:
       - id: golangci-lint
-        entry: bash -c "find tools/ -name go.mod -print0 | xargs -0 -I{} bash -c 'cd \"$(dirname {})\" && golangci-lint run ./...'"
+        entry: bash -c "find . -name go.mod -print0 | xargs -0 -I{} bash -c 'cd \"$(dirname {})\" && golangci-lint run ./...'"
 
   - repo: https://github.com/astral-sh/ruff-pre-commit
     # Ruff version.

.vscode/launch.json

@@ -40,7 +40,19 @@
       }
     },
     {
-      "name": "Celery",
+      "name": "Celery (lightweight mode)",
+      "configurations": [
+        "Celery primary",
+        "Celery background",
+        "Celery beat"
+      ],
+      "presentation": {
+        "group": "1"
+      },
+      "stopAll": true
+    },
+    {
+      "name": "Celery (standard mode)",
       "configurations": [
         "Celery primary",
         "Celery light",
@@ -241,6 +253,35 @@
       },
       "consoleTitle": "Celery light Console"
     },
+    {
+      "name": "Celery background",
+      "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.background",
+        "worker",
+        "--pool=threads",
+        "--concurrency=20",
+        "--prefetch-multiplier=4",
+        "--loglevel=INFO",
+        "--hostname=background@%n",
+        "-Q",
+        "vespa_metadata_sync,connector_deletion,doc_permissions_upsert,checkpoint_cleanup,index_attempt_cleanup,docprocessing,connector_doc_fetching,connector_pruning,connector_doc_permissions_sync,connector_external_group_sync,csv_generation,kg_processing,monitoring,user_file_processing,user_file_project_sync,user_file_delete,opensearch_migration"
+      ],
+      "presentation": {
+        "group": "2"
+      },
+      "consoleTitle": "Celery background Console"
+    },
     {
       "name": "Celery heavy",
       "type": "debugpy",

AGENTS.md

@@ -86,6 +86,37 @@ Onyx uses Celery for asynchronous task processing with multiple specialized work
      - Monitoring tasks (every 5 minutes)
      - Cleanup tasks (hourly)
 
+#### Worker Deployment Modes
+
+Onyx supports two deployment modes for background workers, controlled by the `USE_LIGHTWEIGHT_BACKGROUND_WORKER` environment variable:
+
+**Lightweight Mode** (default, `USE_LIGHTWEIGHT_BACKGROUND_WORKER=true`):
+
+- Runs a single consolidated `background` worker that handles all background tasks:
+  - Light worker tasks (Vespa operations, permissions sync, deletion)
+  - Document processing (indexing pipeline)
+  - Document fetching (connector data retrieval)
+  - Pruning operations (from `heavy` worker)
+  - Knowledge graph processing (from `kg_processing` worker)
+  - Monitoring tasks (from `monitoring` worker)
+  - User file processing (from `user_file_processing` worker)
+- Lower resource footprint (fewer worker processes)
+- Suitable for smaller deployments or development environments
+- Default concurrency: 20 threads (increased to handle combined workload)
+
+**Standard Mode** (`USE_LIGHTWEIGHT_BACKGROUND_WORKER=false`):
+
+- Runs separate specialized workers as documented above (light, docprocessing, docfetching, heavy, kg_processing, monitoring, user_file_processing)
+- Better isolation and scalability
+- Can scale individual workers independently based on workload
+- Suitable for production deployments with higher load
+
+The deployment mode affects:
+
+- **Backend**: Worker processes spawned by supervisord or dev scripts
+- **Helm**: Which Kubernetes deployments are created
+- **Dev Environment**: Which workers `dev_run_background_jobs.py` spawns
+
 #### Key Features
 
 - **Thread-based Workers**: All workers use thread pools (not processes) for stability
@@ -586,45 +617,6 @@ Keep it high level. You can reference certain files or functions though.
 
 Before writing your plan, make sure to do research. Explore the relevant sections in the codebase.
 
-## Error Handling
-
-**Always raise `OnyxError` from `onyx.error_handling.exceptions` instead of `HTTPException`.
-Never hardcode status codes or use `starlette.status` / `fastapi.status` constants directly.**
-
-A global FastAPI exception handler converts `OnyxError` into a JSON response with the standard
-`{"error_code": "...", "message": "..."}` shape. This eliminates boilerplate and keeps error
-handling consistent across the entire backend.
-
-```python
-from onyx.error_handling.error_codes import OnyxErrorCode
-from onyx.error_handling.exceptions import OnyxError
-
-# ✅ Good
-raise OnyxError(OnyxErrorCode.NOT_FOUND, "Session not found")
-
-# ✅ Good — no extra message needed
-raise OnyxError(OnyxErrorCode.UNAUTHENTICATED)
-
-# ✅ Good — upstream service with dynamic status code
-raise OnyxError(OnyxErrorCode.BAD_GATEWAY, detail, status_code_override=upstream_status)
-
-# ❌ Bad — using HTTPException directly
-raise HTTPException(status_code=404, detail="Session not found")
-
-# ❌ Bad — starlette constant
-raise HTTPException(status_code=status.HTTP_403_FORBIDDEN, detail="Access denied")
-```
-
-Available error codes are defined in `backend/onyx/error_handling/error_codes.py`. If a new error
-category is needed, add it there first — do not invent ad-hoc codes.
-
-**Upstream service errors:** When forwarding errors from an upstream service where the HTTP
-status code is dynamic (comes from the upstream response), use `status_code_override`:
-
-```python
-raise OnyxError(OnyxErrorCode.BAD_GATEWAY, detail, status_code_override=e.response.status_code)
-```
-
 ## Best Practices
 
 In addition to the other content in this file, best practices for contributing

backend/alembic/versions/2664261bfaab_add_cache_store_table.py

@@ -1,37 +0,0 @@
-"""add cache_store table
-
-Revision ID: 2664261bfaab
-Revises: 4a1e4b1c89d2
-Create Date: 2026-02-27 00:00:00.000000
-
-"""
-
-from alembic import op
-import sqlalchemy as sa
-
-# revision identifiers, used by Alembic.
-revision = "2664261bfaab"
-down_revision = "4a1e4b1c89d2"
-branch_labels: None = None
-depends_on: None = None
-
-
-def upgrade() -> None:
-    op.create_table(
-        "cache_store",
-        sa.Column("key", sa.String(), nullable=False),
-        sa.Column("value", sa.LargeBinary(), nullable=True),
-        sa.Column("expires_at", sa.DateTime(timezone=True), nullable=True),
-        sa.PrimaryKeyConstraint("key"),
-    )
-    op.create_index(
-        "ix_cache_store_expires",
-        "cache_store",
-        ["expires_at"],
-        postgresql_where=sa.text("expires_at IS NOT NULL"),
-    )
-
-
-def downgrade() -> None:
-    op.drop_index("ix_cache_store_expires", table_name="cache_store")
-    op.drop_table("cache_store")

backend/alembic/versions/4a1e4b1c89d2_add_indexing_to_userfilestatus.py

@@ -1,51 +0,0 @@
-"""Add INDEXING to UserFileStatus
-
-Revision ID: 4a1e4b1c89d2
-Revises: 6b3b4083c5aa
-Create Date: 2026-02-28 00:00:00.000000
-
-"""
-
-import sqlalchemy as sa
-from alembic import op
-
-revision = "4a1e4b1c89d2"
-down_revision = "6b3b4083c5aa"
-branch_labels = None
-depends_on = None
-
-TABLE = "user_file"
-COLUMN = "status"
-CONSTRAINT_NAME = "ck_user_file_status"
-
-OLD_VALUES = ("PROCESSING", "COMPLETED", "FAILED", "CANCELED", "DELETING")
-NEW_VALUES = ("PROCESSING", "INDEXING", "COMPLETED", "FAILED", "CANCELED", "DELETING")
-
-
-def _drop_status_check_constraint() -> None:
-    """Drop the existing CHECK constraint on user_file.status.
-
-    The constraint name is auto-generated by SQLAlchemy and unknown,
-    so we look it up via the inspector.
-    """
-    inspector = sa.inspect(op.get_bind())
-    for constraint in inspector.get_check_constraints(TABLE):
-        if COLUMN in constraint.get("sqltext", ""):
-            constraint_name = constraint["name"]
-            if constraint_name is not None:
-                op.drop_constraint(constraint_name, TABLE, type_="check")
-
-
-def upgrade() -> None:
-    _drop_status_check_constraint()
-    in_clause = ", ".join(f"'{v}'" for v in NEW_VALUES)
-    op.create_check_constraint(CONSTRAINT_NAME, TABLE, f"{COLUMN} IN ({in_clause})")
-
-
-def downgrade() -> None:
-    op.execute(
-        f"UPDATE {TABLE} SET {COLUMN} = 'PROCESSING' WHERE {COLUMN} = 'INDEXING'"
-    )
-    op.drop_constraint(CONSTRAINT_NAME, TABLE, type_="check")
-    in_clause = ", ".join(f"'{v}'" for v in OLD_VALUES)
-    op.create_check_constraint(CONSTRAINT_NAME, TABLE, f"{COLUMN} IN ({in_clause})")

backend/alembic/versions/6b3b4083c5aa_persona_cleanup_and_featured.py

@@ -1,112 +0,0 @@
-"""persona cleanup and featured
-
-Revision ID: 6b3b4083c5aa
-Revises: 57122d037335
-Create Date: 2026-02-26 12:00:00.000000
-
-"""
-
-from alembic import op
-import sqlalchemy as sa
-
-
-# revision identifiers, used by Alembic.
-revision = "6b3b4083c5aa"
-down_revision = "57122d037335"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    # Add featured column with nullable=True first
-    op.add_column("persona", sa.Column("featured", sa.Boolean(), nullable=True))
-
-    # Migrate data from is_default_persona to featured
-    op.execute("UPDATE persona SET featured = is_default_persona")
-
-    # Make featured non-nullable with default=False
-    op.alter_column(
-        "persona",
-        "featured",
-        existing_type=sa.Boolean(),
-        nullable=False,
-        server_default=sa.false(),
-    )
-
-    # Drop is_default_persona column
-    op.drop_column("persona", "is_default_persona")
-
-    # Drop unused columns
-    op.drop_column("persona", "num_chunks")
-    op.drop_column("persona", "chunks_above")
-    op.drop_column("persona", "chunks_below")
-    op.drop_column("persona", "llm_relevance_filter")
-    op.drop_column("persona", "llm_filter_extraction")
-    op.drop_column("persona", "recency_bias")
-
-
-def downgrade() -> None:
-    # Add back recency_bias column
-    op.add_column(
-        "persona",
-        sa.Column(
-            "recency_bias",
-            sa.VARCHAR(),
-            nullable=False,
-            server_default="base_decay",
-        ),
-    )
-
-    # Add back llm_filter_extraction column
-    op.add_column(
-        "persona",
-        sa.Column(
-            "llm_filter_extraction",
-            sa.Boolean(),
-            nullable=False,
-            server_default=sa.false(),
-        ),
-    )
-
-    # Add back llm_relevance_filter column
-    op.add_column(
-        "persona",
-        sa.Column(
-            "llm_relevance_filter",
-            sa.Boolean(),
-            nullable=False,
-            server_default=sa.false(),
-        ),
-    )
-
-    # Add back chunks_below column
-    op.add_column(
-        "persona",
-        sa.Column("chunks_below", sa.Integer(), nullable=False, server_default="0"),
-    )
-
-    # Add back chunks_above column
-    op.add_column(
-        "persona",
-        sa.Column("chunks_above", sa.Integer(), nullable=False, server_default="0"),
-    )
-
-    # Add back num_chunks column
-    op.add_column("persona", sa.Column("num_chunks", sa.Float(), nullable=True))
-
-    # Add back is_default_persona column
-    op.add_column(
-        "persona",
-        sa.Column(
-            "is_default_persona",
-            sa.Boolean(),
-            nullable=False,
-            server_default=sa.false(),
-        ),
-    )
-
-    # Migrate data from featured to is_default_persona
-    op.execute("UPDATE persona SET is_default_persona = featured")
-
-    # Drop featured column
-    op.drop_column("persona", "featured")

backend/alembic/versions/a3b8d9e2f1c4_make_scim_external_id_nullable.py

@@ -1,34 +0,0 @@
-"""make scim_user_mapping.external_id nullable
-
-Revision ID: a3b8d9e2f1c4
-Revises: 2664261bfaab
-Create Date: 2026-03-02
-
-"""
-
-from alembic import op
-
-
-# revision identifiers, used by Alembic.
-revision = "a3b8d9e2f1c4"
-down_revision = "2664261bfaab"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    op.alter_column(
-        "scim_user_mapping",
-        "external_id",
-        nullable=True,
-    )
-
-
-def downgrade() -> None:
-    # Delete any rows where external_id is NULL before re-applying NOT NULL
-    op.execute("DELETE FROM scim_user_mapping WHERE external_id IS NULL")
-    op.alter_column(
-        "scim_user_mapping",
-        "external_id",
-        nullable=False,
-    )

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

@@ -0,0 +1,15 @@
+from onyx.background.celery.apps import app_base
+from onyx.background.celery.apps.background import celery_app
+
+
+celery_app.autodiscover_tasks(
+    app_base.filter_task_modules(
+        [
+            "ee.onyx.background.celery.tasks.doc_permission_syncing",
+            "ee.onyx.background.celery.tasks.external_group_syncing",
+            "ee.onyx.background.celery.tasks.cleanup",
+            "ee.onyx.background.celery.tasks.tenant_provisioning",
+            "ee.onyx.background.celery.tasks.query_history",
+        ]
+    )
+)

backend/ee/onyx/db/license.py

@@ -11,10 +11,11 @@ from ee.onyx.server.license.models import LicenseMetadata
 from ee.onyx.server.license.models import LicensePayload
 from ee.onyx.server.license.models import LicenseSource
 from onyx.auth.schemas import UserRole
-from onyx.cache.factory import get_cache_backend
 from onyx.configs.constants import ANONYMOUS_USER_EMAIL
 from onyx.db.models import License
 from onyx.db.models import User
+from onyx.redis.redis_pool import get_redis_client
+from onyx.redis.redis_pool import get_redis_replica_client
 from onyx.utils.logger import setup_logger
 from shared_configs.configs import MULTI_TENANT
 from shared_configs.contextvars import get_current_tenant_id
@@ -141,7 +142,7 @@ def get_used_seats(tenant_id: str | None = None) -> int:
 
 def get_cached_license_metadata(tenant_id: str | None = None) -> LicenseMetadata | None:
     """
-    Get license metadata from cache.
+    Get license metadata from Redis cache.
 
     Args:
         tenant_id: Tenant ID (for multi-tenant deployments)
@@ -149,34 +150,38 @@ def get_cached_license_metadata(tenant_id: str | None = None) -> LicenseMetadata
     Returns:
         LicenseMetadata if cached, None otherwise
     """
-    cache = get_cache_backend(tenant_id=tenant_id)
-    cached = cache.get(LICENSE_METADATA_KEY)
-    if not cached:
-        return None
+    tenant = tenant_id or get_current_tenant_id()
+    redis_client = get_redis_replica_client(tenant_id=tenant)
 
-    try:
-        cached_str = (
-            cached.decode("utf-8") if isinstance(cached, bytes) else str(cached)
-        )
-        return LicenseMetadata.model_validate_json(cached_str)
-    except Exception as e:
-        logger.warning(f"Failed to parse cached license metadata: {e}")
-        return None
+    cached = redis_client.get(LICENSE_METADATA_KEY)
+    if cached:
+        try:
+            cached_str: str
+            if isinstance(cached, bytes):
+                cached_str = cached.decode("utf-8")
+            else:
+                cached_str = str(cached)
+            return LicenseMetadata.model_validate_json(cached_str)
+        except Exception as e:
+            logger.warning(f"Failed to parse cached license metadata: {e}")
+            return None
+    return None
 
 
 def invalidate_license_cache(tenant_id: str | None = None) -> None:
     """
     Invalidate the license metadata cache (not the license itself).
 
-    Deletes the cached LicenseMetadata. The actual license in the database
-    is not affected. Delete is idempotent — if the key doesn't exist, this
-    is a no-op.
+    This deletes the cached LicenseMetadata from Redis. The actual license
+    in the database is not affected. Redis delete is idempotent - if the
+    key doesn't exist, this is a no-op.
 
     Args:
         tenant_id: Tenant ID (for multi-tenant deployments)
     """
-    cache = get_cache_backend(tenant_id=tenant_id)
-    cache.delete(LICENSE_METADATA_KEY)
+    tenant = tenant_id or get_current_tenant_id()
+    redis_client = get_redis_client(tenant_id=tenant)
+    redis_client.delete(LICENSE_METADATA_KEY)
     logger.info("License cache invalidated")
 
 
@@ -187,7 +192,7 @@ def update_license_cache(
     tenant_id: str | None = None,
 ) -> LicenseMetadata:
     """
-    Update the cache with license metadata.
+    Update the Redis cache with license metadata.
 
     We cache all license statuses (ACTIVE, GRACE_PERIOD, GATED_ACCESS) because:
     1. Frontend needs status to show appropriate UI/banners
@@ -206,7 +211,7 @@ def update_license_cache(
     from ee.onyx.utils.license import get_license_status
 
     tenant = tenant_id or get_current_tenant_id()
-    cache = get_cache_backend(tenant_id=tenant_id)
+    redis_client = get_redis_client(tenant_id=tenant)
 
     used_seats = get_used_seats(tenant)
     status = get_license_status(payload, grace_period_end)
@@ -225,7 +230,7 @@ def update_license_cache(
         stripe_subscription_id=payload.stripe_subscription_id,
     )
 
-    cache.set(
+    redis_client.set(
         LICENSE_METADATA_KEY,
         metadata.model_dump_json(),
         ex=LICENSE_CACHE_TTL_SECONDS,

backend/ee/onyx/db/scim.py

@@ -126,16 +126,12 @@ class ScimDAL(DAL):
 
     def create_user_mapping(
         self,
-        external_id: str | None,
+        external_id: str,
         user_id: UUID,
         scim_username: str | None = None,
         fields: ScimMappingFields | None = None,
     ) -> ScimUserMapping:
-        """Create a SCIM mapping for a user.
-
-        ``external_id`` may be ``None`` when the IdP omits it (RFC 7643
-        allows this). The mapping still marks the user as SCIM-managed.
-        """
+        """Create a mapping between a SCIM externalId and an Onyx user."""
         f = fields or ScimMappingFields()
         mapping = ScimUserMapping(
             external_id=external_id,
@@ -274,13 +270,8 @@ class ScimDAL(DAL):
         Raises:
             ValueError: If the filter uses an unsupported attribute.
         """
-        # Inner-join with ScimUserMapping so only SCIM-managed users appear.
-        # Pre-existing system accounts (anonymous, admin, etc.) are excluded
-        # unless they were explicitly linked via SCIM provisioning.
-        query = (
-            select(User)
-            .join(ScimUserMapping, ScimUserMapping.user_id == User.id)
-            .where(User.role.notin_([UserRole.SLACK_USER, UserRole.EXT_PERM_USER]))
+        query = select(User).where(
+            User.role.notin_([UserRole.SLACK_USER, UserRole.EXT_PERM_USER])
         )
 
         if scim_filter:
@@ -330,37 +321,34 @@ class ScimDAL(DAL):
         scim_username: str | None = None,
         fields: ScimMappingFields | None = None,
     ) -> None:
-        """Sync the SCIM mapping for a user.
-
-        If a mapping already exists, its fields are updated (including
-        setting ``external_id`` to ``None`` when the IdP omits it).
-        If no mapping exists and ``new_external_id`` is provided, a new
-        mapping is created.  A mapping is never deleted here — SCIM-managed
-        users must retain their mapping to remain visible in ``GET /Users``.
+        """Create, update, or delete the external ID mapping for a user.
 
         When *fields* is provided, all mapping fields are written
         unconditionally — including ``None`` values — so that a caller can
         clear a previously-set field (e.g. removing a department).
         """
         mapping = self.get_user_mapping_by_user_id(user_id)
-        if mapping:
-            if mapping.external_id != new_external_id:
-                mapping.external_id = new_external_id
-            if scim_username is not None:
-                mapping.scim_username = scim_username
-            if fields is not None:
-                mapping.department = fields.department
-                mapping.manager = fields.manager
-                mapping.given_name = fields.given_name
-                mapping.family_name = fields.family_name
-                mapping.scim_emails_json = fields.scim_emails_json
-        elif new_external_id:
-            self.create_user_mapping(
-                external_id=new_external_id,
-                user_id=user_id,
-                scim_username=scim_username,
-                fields=fields,
-            )
+        if new_external_id:
+            if mapping:
+                if mapping.external_id != new_external_id:
+                    mapping.external_id = new_external_id
+                if scim_username is not None:
+                    mapping.scim_username = scim_username
+                if fields is not None:
+                    mapping.department = fields.department
+                    mapping.manager = fields.manager
+                    mapping.given_name = fields.given_name
+                    mapping.family_name = fields.family_name
+                    mapping.scim_emails_json = fields.scim_emails_json
+            else:
+                self.create_user_mapping(
+                    external_id=new_external_id,
+                    user_id=user_id,
+                    scim_username=scim_username,
+                    fields=fields,
+                )
+        elif mapping:
+            self.delete_user_mapping(mapping.id)
 
     def _get_user_mappings_batch(
         self, user_ids: list[UUID]

backend/ee/onyx/main.py

@@ -31,7 +31,6 @@ from ee.onyx.server.query_and_chat.query_backend import (
 from ee.onyx.server.query_and_chat.search_backend import router as search_router
 from ee.onyx.server.query_history.api import router as query_history_router
 from ee.onyx.server.reporting.usage_export_api import router as usage_export_router
-from ee.onyx.server.scim.api import register_scim_exception_handlers
 from ee.onyx.server.scim.api import scim_router
 from ee.onyx.server.seeding import seed_db
 from ee.onyx.server.tenants.api import router as tenants_router
@@ -168,7 +167,6 @@ def get_application() -> FastAPI:
     # they use their own SCIM bearer token auth).
     # Not behind APP_API_PREFIX because IdPs expect /scim/v2/... directly.
     application.include_router(scim_router)
-    register_scim_exception_handlers(application)
 
     # Ensure all routes have auth enabled or are explicitly marked as public
     check_ee_router_auth(application)

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

@@ -26,6 +26,7 @@ import asyncio
 import httpx
 from fastapi import APIRouter
 from fastapi import Depends
+from fastapi import HTTPException
 from pydantic import BaseModel
 from sqlalchemy.orm import Session
 
@@ -41,6 +42,7 @@ from ee.onyx.server.billing.models import SeatUpdateRequest
 from ee.onyx.server.billing.models import SeatUpdateResponse
 from ee.onyx.server.billing.models import StripePublishableKeyResponse
 from ee.onyx.server.billing.models import SubscriptionStatusResponse
+from ee.onyx.server.billing.service import BillingServiceError
 from ee.onyx.server.billing.service import (
     create_checkout_session as create_checkout_service,
 )
@@ -56,8 +58,6 @@ from onyx.configs.app_configs import STRIPE_PUBLISHABLE_KEY_OVERRIDE
 from onyx.configs.app_configs import STRIPE_PUBLISHABLE_KEY_URL
 from onyx.configs.app_configs import WEB_DOMAIN
 from onyx.db.engine.sql_engine import get_session
-from onyx.error_handling.error_codes import OnyxErrorCode
-from onyx.error_handling.exceptions import OnyxError
 from onyx.redis.redis_pool import get_shared_redis_client
 from onyx.utils.logger import setup_logger
 from shared_configs.configs import MULTI_TENANT
@@ -169,23 +169,26 @@ async def create_checkout_session(
     if seats is not None:
         used_seats = get_used_seats(tenant_id)
         if seats < used_seats:
-            raise OnyxError(
-                OnyxErrorCode.VALIDATION_ERROR,
-                f"Cannot subscribe with fewer seats than current usage. "
+            raise HTTPException(
+                status_code=400,
+                detail=f"Cannot subscribe with fewer seats than current usage. "
                 f"You have {used_seats} active users/integrations but requested {seats} seats.",
             )
 
     # Build redirect URL for after checkout completion
     redirect_url = f"{WEB_DOMAIN}/admin/billing?checkout=success"
 
-    return await create_checkout_service(
-        billing_period=billing_period,
-        seats=seats,
-        email=email,
-        license_data=license_data,
-        redirect_url=redirect_url,
-        tenant_id=tenant_id,
-    )
+    try:
+        return await create_checkout_service(
+            billing_period=billing_period,
+            seats=seats,
+            email=email,
+            license_data=license_data,
+            redirect_url=redirect_url,
+            tenant_id=tenant_id,
+        )
+    except BillingServiceError as e:
+        raise HTTPException(status_code=e.status_code, detail=e.message)
 
 
 @router.post("/create-customer-portal-session")
@@ -203,15 +206,18 @@ async def create_customer_portal_session(
 
     # Self-hosted requires license
     if not MULTI_TENANT and not license_data:
-        raise OnyxError(OnyxErrorCode.VALIDATION_ERROR, "No license found")
+        raise HTTPException(status_code=400, detail="No license found")
 
     return_url = request.return_url if request else f"{WEB_DOMAIN}/admin/billing"
 
-    return await create_portal_service(
-        license_data=license_data,
-        return_url=return_url,
-        tenant_id=tenant_id,
-    )
+    try:
+        return await create_portal_service(
+            license_data=license_data,
+            return_url=return_url,
+            tenant_id=tenant_id,
+        )
+    except BillingServiceError as e:
+        raise HTTPException(status_code=e.status_code, detail=e.message)
 
 
 @router.get("/billing-information")
@@ -234,9 +240,9 @@ async def get_billing_information(
 
     # Check circuit breaker (self-hosted only)
     if _is_billing_circuit_open():
-        raise OnyxError(
-            OnyxErrorCode.SERVICE_UNAVAILABLE,
-            "Stripe connection temporarily disabled. Click 'Connect to Stripe' to retry.",
+        raise HTTPException(
+            status_code=503,
+            detail="Stripe connection temporarily disabled. Click 'Connect to Stripe' to retry.",
         )
 
     try:
@@ -244,11 +250,11 @@ async def get_billing_information(
             license_data=license_data,
             tenant_id=tenant_id,
         )
-    except OnyxError as e:
+    except BillingServiceError as e:
         # Open circuit breaker on connection failures (self-hosted only)
         if e.status_code in (502, 503, 504):
             _open_billing_circuit()
-        raise
+        raise HTTPException(status_code=e.status_code, detail=e.message)
 
 
 @router.post("/seats/update")
@@ -268,25 +274,31 @@ async def update_seats(
 
     # Self-hosted requires license
     if not MULTI_TENANT and not license_data:
-        raise OnyxError(OnyxErrorCode.VALIDATION_ERROR, "No license found")
+        raise HTTPException(status_code=400, detail="No license found")
 
     # Validate that new seat count is not less than current used seats
     used_seats = get_used_seats(tenant_id)
     if request.new_seat_count < used_seats:
-        raise OnyxError(
-            OnyxErrorCode.VALIDATION_ERROR,
-            f"Cannot reduce seats below current usage. "
+        raise HTTPException(
+            status_code=400,
+            detail=f"Cannot reduce seats below current usage. "
             f"You have {used_seats} active users/integrations but requested {request.new_seat_count} seats.",
         )
 
-    # Note: Don't store license here - the control plane may still be processing
-    # the subscription update. The frontend should call /license/claim after a
-    # short delay to get the freshly generated license.
-    return await update_seat_service(
-        new_seat_count=request.new_seat_count,
-        license_data=license_data,
-        tenant_id=tenant_id,
-    )
+    try:
+        result = await update_seat_service(
+            new_seat_count=request.new_seat_count,
+            license_data=license_data,
+            tenant_id=tenant_id,
+        )
+
+        # Note: Don't store license here - the control plane may still be processing
+        # the subscription update. The frontend should call /license/claim after a
+        # short delay to get the freshly generated license.
+
+        return result
+    except BillingServiceError as e:
+        raise HTTPException(status_code=e.status_code, detail=e.message)
 
 
 @router.get("/stripe-publishable-key")
@@ -317,18 +329,18 @@ async def get_stripe_publishable_key() -> StripePublishableKeyResponse:
         if STRIPE_PUBLISHABLE_KEY_OVERRIDE:
             key = STRIPE_PUBLISHABLE_KEY_OVERRIDE.strip()
             if not key.startswith("pk_"):
-                raise OnyxError(
-                    OnyxErrorCode.INTERNAL_ERROR,
-                    "Invalid Stripe publishable key format",
+                raise HTTPException(
+                    status_code=500,
+                    detail="Invalid Stripe publishable key format",
                 )
             _stripe_publishable_key_cache = key
             return StripePublishableKeyResponse(publishable_key=key)
 
         # Fall back to S3 bucket
         if not STRIPE_PUBLISHABLE_KEY_URL:
-            raise OnyxError(
-                OnyxErrorCode.INTERNAL_ERROR,
-                "Stripe publishable key is not configured",
+            raise HTTPException(
+                status_code=500,
+                detail="Stripe publishable key is not configured",
             )
 
         try:
@@ -339,17 +351,17 @@ async def get_stripe_publishable_key() -> StripePublishableKeyResponse:
 
                 # Validate key format
                 if not key.startswith("pk_"):
-                    raise OnyxError(
-                        OnyxErrorCode.INTERNAL_ERROR,
-                        "Invalid Stripe publishable key format",
+                    raise HTTPException(
+                        status_code=500,
+                        detail="Invalid Stripe publishable key format",
                     )
 
                 _stripe_publishable_key_cache = key
                 return StripePublishableKeyResponse(publishable_key=key)
         except httpx.HTTPError:
-            raise OnyxError(
-                OnyxErrorCode.INTERNAL_ERROR,
-                "Failed to fetch Stripe publishable key",
+            raise HTTPException(
+                status_code=500,
+                detail="Failed to fetch Stripe publishable key",
             )
 
 

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

@@ -22,8 +22,6 @@ from ee.onyx.server.billing.models import SeatUpdateResponse
 from ee.onyx.server.billing.models import SubscriptionStatusResponse
 from ee.onyx.server.tenants.access import generate_data_plane_token
 from onyx.configs.app_configs import CONTROL_PLANE_API_BASE_URL
-from onyx.error_handling.error_codes import OnyxErrorCode
-from onyx.error_handling.exceptions import OnyxError
 from onyx.utils.logger import setup_logger
 from shared_configs.configs import MULTI_TENANT
 
@@ -33,6 +31,15 @@ logger = setup_logger()
 _REQUEST_TIMEOUT = 30.0
 
 
+class BillingServiceError(Exception):
+    """Exception raised for billing service errors."""
+
+    def __init__(self, message: str, status_code: int = 500):
+        self.message = message
+        self.status_code = status_code
+        super().__init__(self.message)
+
+
 def _get_proxy_headers(license_data: str | None) -> dict[str, str]:
     """Build headers for proxy requests (self-hosted).
 
@@ -94,7 +101,7 @@ async def _make_billing_request(
         Response JSON as dict
 
     Raises:
-        OnyxError: If request fails
+        BillingServiceError: If request fails
     """
 
     base_url = _get_base_url()
@@ -121,17 +128,11 @@ async def _make_billing_request(
         except Exception:
             pass
         logger.error(f"{error_message}: {e.response.status_code} - {detail}")
-        raise OnyxError(
-            OnyxErrorCode.BAD_GATEWAY,
-            detail,
-            status_code_override=e.response.status_code,
-        )
+        raise BillingServiceError(detail, e.response.status_code)
 
     except httpx.RequestError:
         logger.exception("Failed to connect to billing service")
-        raise OnyxError(
-            OnyxErrorCode.BAD_GATEWAY, "Failed to connect to billing service"
-        )
+        raise BillingServiceError("Failed to connect to billing service", 502)
 
 
 async def create_checkout_session(

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

@@ -14,6 +14,7 @@ import requests
 from fastapi import APIRouter
 from fastapi import Depends
 from fastapi import File
+from fastapi import HTTPException
 from fastapi import UploadFile
 from sqlalchemy.orm import Session
 
@@ -34,8 +35,6 @@ from ee.onyx.server.license.models import SeatUsageResponse
 from ee.onyx.utils.license import verify_license_signature
 from onyx.auth.users import User
 from onyx.db.engine.sql_engine import get_session
-from onyx.error_handling.error_codes import OnyxErrorCode
-from onyx.error_handling.exceptions import OnyxError
 from onyx.utils.logger import setup_logger
 from shared_configs.configs import MULTI_TENANT
 
@@ -128,9 +127,9 @@ async def claim_license(
     2. Without session_id: Re-claim using existing license for auth
     """
     if MULTI_TENANT:
-        raise OnyxError(
-            OnyxErrorCode.VALIDATION_ERROR,
-            "License claiming is only available for self-hosted deployments",
+        raise HTTPException(
+            status_code=400,
+            detail="License claiming is only available for self-hosted deployments",
         )
 
     try:
@@ -147,16 +146,15 @@ async def claim_license(
             # Re-claim using existing license for auth
             metadata = get_license_metadata(db_session)
             if not metadata or not metadata.tenant_id:
-                raise OnyxError(
-                    OnyxErrorCode.VALIDATION_ERROR,
-                    "No license found. Provide session_id after checkout.",
+                raise HTTPException(
+                    status_code=400,
+                    detail="No license found. Provide session_id after checkout.",
                 )
 
             license_row = get_license(db_session)
             if not license_row or not license_row.license_data:
-                raise OnyxError(
-                    OnyxErrorCode.VALIDATION_ERROR,
-                    "No license found in database",
+                raise HTTPException(
+                    status_code=400, detail="No license found in database"
                 )
 
             url = f"{CLOUD_DATA_PLANE_URL}/proxy/license/{metadata.tenant_id}"
@@ -175,7 +173,7 @@ async def claim_license(
         license_data = data.get("license")
 
         if not license_data:
-            raise OnyxError(OnyxErrorCode.NOT_FOUND, "No license in response")
+            raise HTTPException(status_code=404, detail="No license in response")
 
         # Verify signature before persisting
         payload = verify_license_signature(license_data)
@@ -201,14 +199,12 @@ async def claim_license(
             detail = error_data.get("detail", detail)
         except Exception:
             pass
-        raise OnyxError(
-            OnyxErrorCode.BAD_GATEWAY, detail, status_code_override=status_code
-        )
+        raise HTTPException(status_code=status_code, detail=detail)
     except ValueError as e:
-        raise OnyxError(OnyxErrorCode.VALIDATION_ERROR, str(e))
+        raise HTTPException(status_code=400, detail=str(e))
     except requests.RequestException:
-        raise OnyxError(
-            OnyxErrorCode.BAD_GATEWAY, "Failed to connect to license server"
+        raise HTTPException(
+            status_code=502, detail="Failed to connect to license server"
         )
 
 
@@ -225,9 +221,9 @@ async def upload_license(
     The license file must be cryptographically signed by Onyx.
     """
     if MULTI_TENANT:
-        raise OnyxError(
-            OnyxErrorCode.VALIDATION_ERROR,
-            "License upload is only available for self-hosted deployments",
+        raise HTTPException(
+            status_code=400,
+            detail="License upload is only available for self-hosted deployments",
         )
 
     try:
@@ -238,14 +234,14 @@ async def upload_license(
         # Remove any stray whitespace/newlines from user input
         license_data = license_data.strip()
     except UnicodeDecodeError:
-        raise OnyxError(OnyxErrorCode.INVALID_INPUT, "Invalid license file format")
+        raise HTTPException(status_code=400, detail="Invalid license file format")
 
     # Verify cryptographic signature - this is the only validation needed
     # The license's tenant_id identifies the customer in control plane, not locally
     try:
         payload = verify_license_signature(license_data)
     except ValueError as e:
-        raise OnyxError(OnyxErrorCode.VALIDATION_ERROR, str(e))
+        raise HTTPException(status_code=400, detail=str(e))
 
     # Persist to DB and update cache
     upsert_license(db_session, license_data)
@@ -301,9 +297,9 @@ async def delete_license(
     Admin only - removes license from database and invalidates cache.
     """
     if MULTI_TENANT:
-        raise OnyxError(
-            OnyxErrorCode.VALIDATION_ERROR,
-            "License deletion is only available for self-hosted deployments",
+        raise HTTPException(
+            status_code=400,
+            detail="License deletion is only available for self-hosted deployments",
         )
 
     try:

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

@@ -46,6 +46,7 @@ from fastapi import FastAPI
 from fastapi import Request
 from fastapi import Response
 from fastapi.responses import JSONResponse
+from redis.exceptions import RedisError
 from sqlalchemy.exc import SQLAlchemyError
 
 from ee.onyx.configs.app_configs import LICENSE_ENFORCEMENT_ENABLED
@@ -55,7 +56,6 @@ from ee.onyx.configs.license_enforcement_config import (
 )
 from ee.onyx.db.license import get_cached_license_metadata
 from ee.onyx.db.license import refresh_license_cache
-from onyx.cache.interface import CACHE_TRANSIENT_ERRORS
 from onyx.db.engine.sql_engine import get_session_with_current_tenant
 from onyx.server.settings.models import ApplicationStatus
 from shared_configs.contextvars import get_current_tenant_id
@@ -164,9 +164,9 @@ def add_license_enforcement_middleware(
                     "[license_enforcement] No license, allowing community features"
                 )
                 is_gated = False
-        except CACHE_TRANSIENT_ERRORS as e:
+        except RedisError as e:
             logger.warning(f"Failed to check license metadata: {e}")
-            # Fail open - don't block users due to cache connectivity issues
+            # Fail open - don't block users due to Redis connectivity issues
             is_gated = False
 
         if is_gated:

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

@@ -15,9 +15,7 @@ from uuid import UUID
 
 from fastapi import APIRouter
 from fastapi import Depends
-from fastapi import FastAPI
 from fastapi import Query
-from fastapi import Request
 from fastapi import Response
 from fastapi.responses import JSONResponse
 from fastapi_users.password import PasswordHelper
@@ -26,7 +24,6 @@ from sqlalchemy.exc import IntegrityError
 from sqlalchemy.orm import Session
 
 from ee.onyx.db.scim import ScimDAL
-from ee.onyx.server.scim.auth import ScimAuthError
 from ee.onyx.server.scim.auth import verify_scim_token
 from ee.onyx.server.scim.filtering import parse_scim_filter
 from ee.onyx.server.scim.models import SCIM_LIST_RESPONSE_SCHEMA
@@ -80,22 +77,6 @@ scim_router = APIRouter(prefix="/scim/v2", tags=["SCIM"])
 _pw_helper = PasswordHelper()
 
 
-def register_scim_exception_handlers(app: FastAPI) -> None:
-    """Register SCIM-specific exception handlers on the FastAPI app.
-
-    Call this after ``app.include_router(scim_router)`` so that auth
-    failures from ``verify_scim_token`` return RFC 7644 §3.12 error
-    envelopes (with ``schemas`` and ``status`` fields) instead of
-    FastAPI's default ``{"detail": "..."}`` format.
-    """
-
-    @app.exception_handler(ScimAuthError)
-    async def _handle_scim_auth_error(
-        _request: Request, exc: ScimAuthError
-    ) -> ScimJSONResponse:
-        return _scim_error_response(exc.status_code, exc.detail)
-
-
 def _get_provider(
     _token: ScimToken = Depends(verify_scim_token),
 ) -> ScimProvider:
@@ -423,63 +404,21 @@ def create_user(
 
     email = user_resource.userName.strip()
 
-    # Check for existing user — if they exist but aren't SCIM-managed yet,
-    # link them to the IdP rather than rejecting with 409.
-    external_id: str | None = user_resource.externalId
-    scim_username: str = user_resource.userName.strip()
-    fields: ScimMappingFields = _fields_from_resource(user_resource)
+    # externalId is how the IdP correlates this user on subsequent requests.
+    # Without it, the IdP can't find the user and will try to re-create,
+    # hitting a 409 conflict — so we require it up front.
+    if not user_resource.externalId:
+        return _scim_error_response(400, "externalId is required")
 
-    existing_user = dal.get_user_by_email(email)
-    if existing_user:
-        existing_mapping = dal.get_user_mapping_by_user_id(existing_user.id)
-        if existing_mapping:
-            return _scim_error_response(409, f"User with email {email} already exists")
-
-        # Adopt pre-existing user into SCIM management.
-        # Reactivating a deactivated user consumes a seat, so enforce the
-        # seat limit the same way replace_user does.
-        if user_resource.active and not existing_user.is_active:
-            seat_error = _check_seat_availability(dal)
-            if seat_error:
-                return _scim_error_response(403, seat_error)
-
-        personal_name = _scim_name_to_str(user_resource.name)
-        dal.update_user(
-            existing_user,
-            is_active=user_resource.active,
-            **({"personal_name": personal_name} if personal_name else {}),
-        )
-
-        try:
-            dal.create_user_mapping(
-                external_id=external_id,
-                user_id=existing_user.id,
-                scim_username=scim_username,
-                fields=fields,
-            )
-            dal.commit()
-        except IntegrityError:
-            dal.rollback()
-            return _scim_error_response(
-                409, f"User with email {email} already has a SCIM mapping"
-            )
-
-        return _scim_resource_response(
-            provider.build_user_resource(
-                existing_user,
-                external_id,
-                scim_username=scim_username,
-                fields=fields,
-            ),
-            status_code=201,
-        )
-
-    # Only enforce seat limit for net-new users — adopting a pre-existing
-    # user doesn't consume a new seat.
+    # Enforce seat limit
     seat_error = _check_seat_availability(dal)
     if seat_error:
         return _scim_error_response(403, seat_error)
 
+    # Check for existing user
+    if dal.get_user_by_email(email):
+        return _scim_error_response(409, f"User with email {email} already exists")
+
     # Create user with a random password (SCIM users authenticate via IdP)
     personal_name = _scim_name_to_str(user_resource.name)
     user = User(
@@ -497,21 +436,18 @@ def create_user(
         dal.rollback()
         return _scim_error_response(409, f"User with email {email} already exists")
 
-    # Always create a SCIM mapping so that the user is marked as
-    # SCIM-managed. externalId may be None (RFC 7643 says it's optional).
-    try:
-        dal.create_user_mapping(
-            external_id=external_id,
-            user_id=user.id,
-            scim_username=scim_username,
-            fields=fields,
-        )
-        dal.commit()
-    except IntegrityError:
-        dal.rollback()
-        return _scim_error_response(
-            409, f"User with email {email} already has a SCIM mapping"
-        )
+    # Create SCIM mapping (externalId is validated above, always present)
+    external_id = user_resource.externalId
+    scim_username = user_resource.userName.strip()
+    fields = _fields_from_resource(user_resource)
+    dal.create_user_mapping(
+        external_id=external_id,
+        user_id=user.id,
+        scim_username=scim_username,
+        fields=fields,
+    )
+
+    dal.commit()
 
     return _scim_resource_response(
         provider.build_user_resource(

backend/ee/onyx/server/scim/auth.py

@@ -19,6 +19,7 @@ import hashlib
 import secrets
 
 from fastapi import Depends
+from fastapi import HTTPException
 from fastapi import Request
 from sqlalchemy.orm import Session
 
@@ -27,21 +28,6 @@ from onyx.auth.utils import get_hashed_bearer_token_from_request
 from onyx.db.engine.sql_engine import get_session
 from onyx.db.models import ScimToken
 
-
-class ScimAuthError(Exception):
-    """Raised when SCIM bearer token authentication fails.
-
-    Unlike HTTPException, this carries the status and detail so the SCIM
-    exception handler can wrap them in an RFC 7644 §3.12 error envelope
-    with ``schemas`` and ``status`` fields.
-    """
-
-    def __init__(self, status_code: int, detail: str) -> None:
-        self.status_code = status_code
-        self.detail = detail
-        super().__init__(detail)
-
-
 SCIM_TOKEN_PREFIX = "onyx_scim_"
 SCIM_TOKEN_LENGTH = 48
 
@@ -96,14 +82,23 @@ def verify_scim_token(
     """
     hashed = _get_hashed_scim_token_from_request(request)
     if not hashed:
-        raise ScimAuthError(401, "Missing or invalid SCIM bearer token")
+        raise HTTPException(
+            status_code=401,
+            detail="Missing or invalid SCIM bearer token",
+        )
 
     token = dal.get_token_by_hash(hashed)
 
     if not token:
-        raise ScimAuthError(401, "Invalid SCIM bearer token")
+        raise HTTPException(
+            status_code=401,
+            detail="Invalid SCIM bearer token",
+        )
 
     if not token.is_active:
-        raise ScimAuthError(401, "SCIM token has been revoked")
+        raise HTTPException(
+            status_code=401,
+            detail="SCIM token has been revoked",
+        )
 
     return token

backend/ee/onyx/server/scim/providers/base.py

@@ -153,31 +153,26 @@ class ScimProvider(ABC):
         self,
         user: User,
         fields: ScimMappingFields,
-    ) -> ScimName:
+    ) -> ScimName | None:
         """Build SCIM name components for the response.
 
         Round-trips stored ``given_name``/``family_name`` when available (so
         the IdP gets back what it sent). Falls back to splitting
         ``personal_name`` for users provisioned before we stored components.
-        Always returns a ScimName — Okta's spec tests expect ``name``
-        (with ``givenName``/``familyName``) on every user resource.
         Providers may override for custom behavior.
         """
         if fields.given_name is not None or fields.family_name is not None:
             return ScimName(
-                givenName=fields.given_name or "",
-                familyName=fields.family_name or "",
-                formatted=user.personal_name or "",
+                givenName=fields.given_name,
+                familyName=fields.family_name,
+                formatted=user.personal_name,
             )
         if not user.personal_name:
-            # Derive a reasonable name from the email so that SCIM spec tests
-            # see non-empty givenName / familyName for every user resource.
-            local = user.email.split("@")[0] if user.email else ""
-            return ScimName(givenName=local, familyName="", formatted=local)
+            return None
         parts = user.personal_name.split(" ", 1)
         return ScimName(
             givenName=parts[0],
-            familyName=parts[1] if len(parts) > 1 else "",
+            familyName=parts[1] if len(parts) > 1 else None,
             formatted=user.personal_name,
         )
 

backend/ee/onyx/server/seeding.py

@@ -18,8 +18,8 @@ from ee.onyx.server.enterprise_settings.store import (
     store_settings as store_ee_settings,
 )
 from ee.onyx.server.enterprise_settings.store import upload_logo
+from onyx.context.search.enums import RecencyBiasSetting
 from onyx.db.engine.sql_engine import get_session_with_current_tenant
-from onyx.db.llm import fetch_existing_llm_provider
 from onyx.db.llm import update_default_provider
 from onyx.db.llm import upsert_llm_provider
 from onyx.db.models import Tool
@@ -117,38 +117,15 @@ def _seed_custom_tools(db_session: Session, tools: List[CustomToolSeed]) -> None
 def _seed_llms(
     db_session: Session, llm_upsert_requests: list[LLMProviderUpsertRequest]
 ) -> None:
-    if not llm_upsert_requests:
-        return
-
-    logger.notice("Seeding LLMs")
-    for request in llm_upsert_requests:
-        existing = fetch_existing_llm_provider(name=request.name, db_session=db_session)
-        if existing:
-            request.id = existing.id
-    seeded_providers = [
-        upsert_llm_provider(llm_upsert_request, db_session)
-        for llm_upsert_request in llm_upsert_requests
-    ]
-
-    default_provider = next(
-        (p for p in seeded_providers if p.model_configurations), None
-    )
-    if not default_provider:
-        return
-
-    visible_configs = [
-        mc for mc in default_provider.model_configurations if mc.is_visible
-    ]
-    default_config = (
-        visible_configs[0]
-        if visible_configs
-        else default_provider.model_configurations[0]
-    )
-    update_default_provider(
-        provider_id=default_provider.id,
-        model_name=default_config.name,
-        db_session=db_session,
-    )
+    if llm_upsert_requests:
+        logger.notice("Seeding LLMs")
+        seeded_providers = [
+            upsert_llm_provider(llm_upsert_request, db_session)
+            for llm_upsert_request in llm_upsert_requests
+        ]
+        update_default_provider(
+            provider_id=seeded_providers[0].id, db_session=db_session
+        )
 
 
 def _seed_personas(db_session: Session, personas: list[PersonaUpsertRequest]) -> None:
@@ -160,6 +137,12 @@ def _seed_personas(db_session: Session, personas: list[PersonaUpsertRequest]) ->
                     user=None,  # Seeding is done as admin
                     name=persona.name,
                     description=persona.description,
+                    num_chunks=(
+                        persona.num_chunks if persona.num_chunks is not None else 0.0
+                    ),
+                    llm_relevance_filter=persona.llm_relevance_filter,
+                    llm_filter_extraction=persona.llm_filter_extraction,
+                    recency_bias=RecencyBiasSetting.AUTO,
                     document_set_ids=persona.document_set_ids,
                     llm_model_provider_override=persona.llm_model_provider_override,
                     llm_model_version_override=persona.llm_model_version_override,
@@ -171,7 +154,6 @@ def _seed_personas(db_session: Session, personas: list[PersonaUpsertRequest]) ->
                     system_prompt=persona.system_prompt,
                     task_prompt=persona.task_prompt,
                     datetime_aware=persona.datetime_aware,
-                    featured=persona.featured,
                     commit=False,
                 )
             db_session.commit()

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

@@ -6,7 +6,6 @@ from sqlalchemy.exc import SQLAlchemyError
 from ee.onyx.configs.app_configs import LICENSE_ENFORCEMENT_ENABLED
 from ee.onyx.db.license import get_cached_license_metadata
 from ee.onyx.db.license import refresh_license_cache
-from onyx.cache.interface import CACHE_TRANSIENT_ERRORS
 from onyx.configs.app_configs import ENTERPRISE_EDITION_ENABLED
 from onyx.db.engine.sql_engine import get_session_with_current_tenant
 from onyx.server.settings.models import ApplicationStatus
@@ -110,12 +109,6 @@ def apply_license_status_to_settings(settings: Settings) -> Settings:
             if metadata.status == _BLOCKING_STATUS:
                 settings.application_status = metadata.status
                 settings.ee_features_enabled = False
-            elif metadata.used_seats > metadata.seats:
-                # License is valid but seat limit exceeded
-                settings.application_status = ApplicationStatus.SEAT_LIMIT_EXCEEDED
-                settings.seat_count = metadata.seats
-                settings.used_seats = metadata.used_seats
-                settings.ee_features_enabled = True
             else:
                 # Has a valid license (GRACE_PERIOD/PAYMENT_REMINDER still allow EE features)
                 settings.ee_features_enabled = True
@@ -126,7 +119,7 @@ def apply_license_status_to_settings(settings: Settings) -> Settings:
                 # syncing) means indexed data may need protection.
                 settings.application_status = _BLOCKING_STATUS
             settings.ee_features_enabled = False
-    except CACHE_TRANSIENT_ERRORS as e:
+    except RedisError as e:
         logger.warning(f"Failed to check license metadata for settings: {e}")
         # Fail closed - disable EE features if we can't verify license
         settings.ee_features_enabled = False

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

@@ -21,6 +21,7 @@ import asyncio
 import httpx
 from fastapi import APIRouter
 from fastapi import Depends
+from fastapi import HTTPException
 
 from ee.onyx.auth.users import current_admin_user
 from ee.onyx.server.tenants.access import control_plane_dep
@@ -42,8 +43,6 @@ from onyx.auth.users import User
 from onyx.configs.app_configs import STRIPE_PUBLISHABLE_KEY_OVERRIDE
 from onyx.configs.app_configs import STRIPE_PUBLISHABLE_KEY_URL
 from onyx.configs.app_configs import WEB_DOMAIN
-from onyx.error_handling.error_codes import OnyxErrorCode
-from onyx.error_handling.exceptions import OnyxError
 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
@@ -117,14 +116,9 @@ async def create_customer_portal_session(
     try:
         portal_url = fetch_customer_portal_session(tenant_id, return_url)
         return {"stripe_customer_portal_url": portal_url}
-    except OnyxError:
-        raise
-    except Exception:
+    except Exception as e:
         logger.exception("Failed to create customer portal session")
-        raise OnyxError(
-            OnyxErrorCode.INTERNAL_ERROR,
-            "Failed to create customer portal session",
-        )
+        raise HTTPException(status_code=500, detail=str(e))
 
 
 @router.post("/create-checkout-session")
@@ -140,14 +134,9 @@ async def create_checkout_session(
     try:
         checkout_url = fetch_stripe_checkout_session(tenant_id, billing_period, seats)
         return {"stripe_checkout_url": checkout_url}
-    except OnyxError:
-        raise
-    except Exception:
+    except Exception as e:
         logger.exception("Failed to create checkout session")
-        raise OnyxError(
-            OnyxErrorCode.INTERNAL_ERROR,
-            "Failed to create checkout session",
-        )
+        raise HTTPException(status_code=500, detail=str(e))
 
 
 @router.post("/create-subscription-session")
@@ -158,20 +147,15 @@ async def create_subscription_session(
     try:
         tenant_id = CURRENT_TENANT_ID_CONTEXTVAR.get()
         if not tenant_id:
-            raise OnyxError(OnyxErrorCode.VALIDATION_ERROR, "Tenant ID not found")
+            raise HTTPException(status_code=400, detail="Tenant ID not found")
 
         billing_period = request.billing_period if request else "monthly"
         session_id = fetch_stripe_checkout_session(tenant_id, billing_period)
         return SubscriptionSessionResponse(sessionId=session_id)
 
-    except OnyxError:
-        raise
-    except Exception:
+    except Exception as e:
         logger.exception("Failed to create subscription session")
-        raise OnyxError(
-            OnyxErrorCode.INTERNAL_ERROR,
-            "Failed to create subscription session",
-        )
+        raise HTTPException(status_code=500, detail=str(e))
 
 
 @router.get("/stripe-publishable-key")
@@ -202,18 +186,18 @@ async def get_stripe_publishable_key() -> StripePublishableKeyResponse:
         if STRIPE_PUBLISHABLE_KEY_OVERRIDE:
             key = STRIPE_PUBLISHABLE_KEY_OVERRIDE.strip()
             if not key.startswith("pk_"):
-                raise OnyxError(
-                    OnyxErrorCode.INTERNAL_ERROR,
-                    "Invalid Stripe publishable key format",
+                raise HTTPException(
+                    status_code=500,
+                    detail="Invalid Stripe publishable key format",
                 )
             _stripe_publishable_key_cache = key
             return StripePublishableKeyResponse(publishable_key=key)
 
         # Fall back to S3 bucket
         if not STRIPE_PUBLISHABLE_KEY_URL:
-            raise OnyxError(
-                OnyxErrorCode.INTERNAL_ERROR,
-                "Stripe publishable key is not configured",
+            raise HTTPException(
+                status_code=500,
+                detail="Stripe publishable key is not configured",
             )
 
         try:
@@ -224,15 +208,15 @@ async def get_stripe_publishable_key() -> StripePublishableKeyResponse:
 
                 # Validate key format
                 if not key.startswith("pk_"):
-                    raise OnyxError(
-                        OnyxErrorCode.INTERNAL_ERROR,
-                        "Invalid Stripe publishable key format",
+                    raise HTTPException(
+                        status_code=500,
+                        detail="Invalid Stripe publishable key format",
                     )
 
                 _stripe_publishable_key_cache = key
                 return StripePublishableKeyResponse(publishable_key=key)
         except httpx.HTTPError:
-            raise OnyxError(
-                OnyxErrorCode.INTERNAL_ERROR,
-                "Failed to fetch Stripe publishable key",
+            raise HTTPException(
+                status_code=500,
+                detail="Failed to fetch Stripe publishable key",
             )

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

@@ -33,7 +33,6 @@ from onyx.configs.constants import MilestoneRecordType
 from onyx.db.engine.sql_engine import get_session_with_shared_schema
 from onyx.db.engine.sql_engine import get_session_with_tenant
 from onyx.db.image_generation import create_default_image_gen_config_from_api_key
-from onyx.db.llm import fetch_existing_llm_provider
 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
@@ -303,17 +302,12 @@ def configure_default_api_keys(db_session: Session) -> None:
 
     has_set_default_provider = False
 
-    def _upsert(request: LLMProviderUpsertRequest, default_model: str) -> None:
+    def _upsert(request: LLMProviderUpsertRequest) -> None:
         nonlocal has_set_default_provider
         try:
-            existing = fetch_existing_llm_provider(
-                name=request.name, db_session=db_session
-            )
-            if existing:
-                request.id = existing.id
             provider = upsert_llm_provider(request, db_session)
             if not has_set_default_provider:
-                update_default_provider(provider.id, default_model, db_session)
+                update_default_provider(provider.id, db_session)
                 has_set_default_provider = True
         except Exception as e:
             logger.error(f"Failed to configure {request.provider} provider: {e}")
@@ -331,13 +325,14 @@ def configure_default_api_keys(db_session: Session) -> None:
             name="OpenAI",
             provider=OPENAI_PROVIDER_NAME,
             api_key=OPENAI_DEFAULT_API_KEY,
+            default_model_name=default_model_name,
             model_configurations=_build_model_configuration_upsert_requests(
                 OPENAI_PROVIDER_NAME, recommendations
             ),
             api_key_changed=True,
             is_auto_mode=True,
         )
-        _upsert(openai_provider, default_model_name)
+        _upsert(openai_provider)
 
         # Create default image generation config using the OpenAI API key
         try:
@@ -366,13 +361,14 @@ def configure_default_api_keys(db_session: Session) -> None:
             name="Anthropic",
             provider=ANTHROPIC_PROVIDER_NAME,
             api_key=ANTHROPIC_DEFAULT_API_KEY,
+            default_model_name=default_model_name,
             model_configurations=_build_model_configuration_upsert_requests(
                 ANTHROPIC_PROVIDER_NAME, recommendations
             ),
             api_key_changed=True,
             is_auto_mode=True,
         )
-        _upsert(anthropic_provider, default_model_name)
+        _upsert(anthropic_provider)
     else:
         logger.info(
             "ANTHROPIC_DEFAULT_API_KEY not set, skipping Anthropic provider configuration"
@@ -397,13 +393,14 @@ def configure_default_api_keys(db_session: Session) -> None:
             name="Google Vertex AI",
             provider=VERTEXAI_PROVIDER_NAME,
             custom_config=custom_config,
+            default_model_name=default_model_name,
             model_configurations=_build_model_configuration_upsert_requests(
                 VERTEXAI_PROVIDER_NAME, recommendations
             ),
             api_key_changed=True,
             is_auto_mode=True,
         )
-        _upsert(vertexai_provider, default_model_name)
+        _upsert(vertexai_provider)
     else:
         logger.info(
             "VERTEXAI_DEFAULT_CREDENTIALS not set, skipping Vertex AI provider configuration"
@@ -435,11 +432,12 @@ def configure_default_api_keys(db_session: Session) -> None:
             name="OpenRouter",
             provider=OPENROUTER_PROVIDER_NAME,
             api_key=OPENROUTER_DEFAULT_API_KEY,
+            default_model_name=default_model_name,
             model_configurations=model_configurations,
             api_key_changed=True,
             is_auto_mode=True,
         )
-        _upsert(openrouter_provider, default_model_name)
+        _upsert(openrouter_provider)
     else:
         logger.info(
             "OPENROUTER_DEFAULT_API_KEY not set, skipping OpenRouter provider configuration"

backend/onyx/auth/users.py

@@ -120,6 +120,7 @@ from onyx.db.models import User
 from onyx.db.pat import fetch_user_for_pat
 from onyx.db.users import get_user_by_email
 from onyx.redis.redis_pool import get_async_redis_connection
+from onyx.redis.redis_pool import get_redis_client
 from onyx.server.settings.store import load_settings
 from onyx.server.utils import BasicAuthenticationError
 from onyx.utils.logger import setup_logger
@@ -200,14 +201,13 @@ def user_needs_to_be_verified() -> bool:
 
 
 def anonymous_user_enabled(*, tenant_id: str | None = None) -> bool:
-    from onyx.cache.factory import get_cache_backend
-
-    cache = get_cache_backend(tenant_id=tenant_id)
-    value = cache.get(OnyxRedisLocks.ANONYMOUS_USER_ENABLED)
+    redis_client = get_redis_client(tenant_id=tenant_id)
+    value = redis_client.get(OnyxRedisLocks.ANONYMOUS_USER_ENABLED)
 
     if value is None:
         return False
 
+    assert isinstance(value, bytes)
     return int(value.decode("utf-8")) == 1
 
 
@@ -543,7 +543,7 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
         result = await db_session.execute(
             select(Persona.id)
             .where(
-                Persona.featured.is_(True),
+                Persona.is_default_persona.is_(True),
                 Persona.is_public.is_(True),
                 Persona.is_visible.is_(True),
                 Persona.deleted.is_(False),
@@ -725,19 +725,11 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
                         if user_by_session:
                             user = user_by_session
 
-                # If the user is inactive, check seat availability before
-                # upgrading role — otherwise they'd become an inactive BASIC
-                # user who still can't log in.
-                if not user.is_active:
-                    with get_session_with_current_tenant() as sync_db:
-                        enforce_seat_limit(sync_db)
-
                 await self.user_db.update(
                     user,
                     {
                         "is_verified": is_verified_by_default,
                         "role": UserRole.BASIC,
-                        **({"is_active": True} if not user.is_active else {}),
                     },
                 )
 

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

@@ -0,0 +1,142 @@
+from typing import Any
+from typing import cast
+
+from celery import Celery
+from celery import signals
+from celery import Task
+from celery.apps.worker import Worker
+from celery.signals import celeryd_init
+from celery.signals import worker_init
+from celery.signals import worker_process_init
+from celery.signals import worker_ready
+from celery.signals import worker_shutdown
+
+import onyx.background.celery.apps.app_base as app_base
+from onyx.background.celery.celery_utils import httpx_init_vespa_pool
+from onyx.configs.app_configs import MANAGED_VESPA
+from onyx.configs.app_configs import VESPA_CLOUD_CERT_PATH
+from onyx.configs.app_configs import VESPA_CLOUD_KEY_PATH
+from onyx.configs.constants import POSTGRES_CELERY_WORKER_BACKGROUND_APP_NAME
+from onyx.db.engine.sql_engine import SqlEngine
+from onyx.utils.logger import setup_logger
+from shared_configs.configs import MULTI_TENANT
+
+
+logger = setup_logger()
+
+celery_app = Celery(__name__)
+celery_app.config_from_object("onyx.background.celery.configs.background")
+celery_app.Task = app_base.TenantAwareTask  # type: ignore [misc]
+
+
+@signals.task_prerun.connect
+def on_task_prerun(
+    sender: Any | None = None,
+    task_id: str | None = None,
+    task: Task | None = None,
+    args: tuple | None = None,
+    kwargs: dict | None = None,
+    **kwds: Any,
+) -> None:
+    app_base.on_task_prerun(sender, task_id, task, args, kwargs, **kwds)
+
+
+@signals.task_postrun.connect
+def on_task_postrun(
+    sender: Any | None = None,
+    task_id: str | None = None,
+    task: Task | None = None,
+    args: tuple | None = None,
+    kwargs: dict | None = None,
+    retval: Any | None = None,
+    state: str | None = None,
+    **kwds: Any,
+) -> None:
+    app_base.on_task_postrun(sender, task_id, task, args, kwargs, retval, state, **kwds)
+
+
+@celeryd_init.connect
+def on_celeryd_init(sender: str, conf: Any = None, **kwargs: Any) -> None:
+    app_base.on_celeryd_init(sender, conf, **kwargs)
+
+
+@worker_init.connect
+def on_worker_init(sender: Worker, **kwargs: Any) -> None:
+    EXTRA_CONCURRENCY = 8  # small extra fudge factor for connection limits
+
+    logger.info("worker_init signal received for consolidated background worker.")
+
+    SqlEngine.set_app_name(POSTGRES_CELERY_WORKER_BACKGROUND_APP_NAME)
+    pool_size = cast(int, sender.concurrency)  # type: ignore
+    SqlEngine.init_engine(pool_size=pool_size, max_overflow=EXTRA_CONCURRENCY)
+
+    # Initialize Vespa httpx pool (needed for light worker tasks)
+    if MANAGED_VESPA:
+        httpx_init_vespa_pool(
+            sender.concurrency + EXTRA_CONCURRENCY,  # type: ignore
+            ssl_cert=VESPA_CLOUD_CERT_PATH,
+            ssl_key=VESPA_CLOUD_KEY_PATH,
+        )
+    else:
+        httpx_init_vespa_pool(sender.concurrency + EXTRA_CONCURRENCY)  # type: ignore
+
+    app_base.wait_for_redis(sender, **kwargs)
+    app_base.wait_for_db(sender, **kwargs)
+    app_base.wait_for_vespa_or_shutdown(sender, **kwargs)
+
+    # Less startup checks in multi-tenant case
+    if MULTI_TENANT:
+        return
+
+    app_base.on_secondary_worker_init(sender, **kwargs)
+
+
+@worker_ready.connect
+def on_worker_ready(sender: Any, **kwargs: Any) -> None:
+    app_base.on_worker_ready(sender, **kwargs)
+
+
+@worker_shutdown.connect
+def on_worker_shutdown(sender: Any, **kwargs: Any) -> None:
+    app_base.on_worker_shutdown(sender, **kwargs)
+
+
+@worker_process_init.connect
+def init_worker(**kwargs: Any) -> None:  # noqa: ARG001
+    SqlEngine.reset_engine()
+
+
+@signals.setup_logging.connect
+def on_setup_logging(
+    loglevel: Any, logfile: Any, format: Any, colorize: Any, **kwargs: Any
+) -> None:
+    app_base.on_setup_logging(loglevel, logfile, format, colorize, **kwargs)
+
+
+base_bootsteps = app_base.get_bootsteps()
+for bootstep in base_bootsteps:
+    celery_app.steps["worker"].add(bootstep)
+
+celery_app.autodiscover_tasks(
+    app_base.filter_task_modules(
+        [
+            # Original background worker tasks
+            "onyx.background.celery.tasks.pruning",
+            "onyx.background.celery.tasks.monitoring",
+            "onyx.background.celery.tasks.user_file_processing",
+            "onyx.background.celery.tasks.llm_model_update",
+            # Light worker tasks
+            "onyx.background.celery.tasks.shared",
+            "onyx.background.celery.tasks.vespa",
+            "onyx.background.celery.tasks.connector_deletion",
+            "onyx.background.celery.tasks.doc_permission_syncing",
+            "onyx.background.celery.tasks.opensearch_migration",
+            # Docprocessing worker tasks
+            "onyx.background.celery.tasks.docprocessing",
+            # Docfetching worker tasks
+            "onyx.background.celery.tasks.docfetching",
+            # Sandbox cleanup tasks (isolated in build feature)
+            "onyx.server.features.build.sandbox.tasks",
+        ]
+    )
+)

backend/onyx/background/celery/configs/background.py

@@ -0,0 +1,23 @@
+import onyx.background.celery.configs.base as shared_config
+from onyx.configs.app_configs import CELERY_WORKER_BACKGROUND_CONCURRENCY
+
+broker_url = shared_config.broker_url
+broker_connection_retry_on_startup = shared_config.broker_connection_retry_on_startup
+broker_pool_limit = shared_config.broker_pool_limit
+broker_transport_options = shared_config.broker_transport_options
+
+redis_socket_keepalive = shared_config.redis_socket_keepalive
+redis_retry_on_timeout = shared_config.redis_retry_on_timeout
+redis_backend_health_check_interval = shared_config.redis_backend_health_check_interval
+
+result_backend = shared_config.result_backend
+result_expires = shared_config.result_expires  # 86400 seconds is the default
+
+task_default_priority = shared_config.task_default_priority
+task_acks_late = shared_config.task_acks_late
+
+worker_concurrency = CELERY_WORKER_BACKGROUND_CONCURRENCY
+worker_pool = "threads"
+# Increased from 1 to 4 to handle fast light worker tasks more efficiently
+# This allows the worker to prefetch multiple tasks per thread
+worker_prefetch_multiplier = 4

backend/onyx/background/celery/tasks/beat_schedule.py

@@ -241,7 +241,8 @@ _VECTOR_DB_BEAT_TASK_NAMES: set[str] = {
     "check-for-index-attempt-cleanup",
     "check-for-doc-permissions-sync",
     "check-for-external-group-sync",
-    "migrate-chunks-from-vespa-to-opensearch",
+    "check-for-documents-for-opensearch-migration",
+    "migrate-documents-from-vespa-to-opensearch",
 }
 
 if DISABLE_VECTOR_DB:

backend/onyx/background/celery/tasks/opensearch_migration/tasks.py

@@ -30,7 +30,6 @@ from onyx.background.celery.tasks.opensearch_migration.transformer import (
     transform_vespa_chunks_to_opensearch_chunks,
 )
 from onyx.configs.app_configs import ENABLE_OPENSEARCH_INDEXING_FOR_ONYX
-from onyx.configs.app_configs import VESPA_MIGRATION_REQUEST_TIMEOUT_S
 from onyx.configs.constants import OnyxCeleryTask
 from onyx.configs.constants import OnyxRedisLocks
 from onyx.db.engine.sql_engine import get_session_with_current_tenant
@@ -48,7 +47,6 @@ from onyx.document_index.interfaces_new import TenantState
 from onyx.document_index.opensearch.opensearch_document_index import (
     OpenSearchDocumentIndex,
 )
-from onyx.document_index.vespa.shared_utils.utils import get_vespa_http_client
 from onyx.document_index.vespa.vespa_document_index import VespaDocumentIndex
 from onyx.indexing.models import IndexingSetting
 from onyx.redis.redis_pool import get_redis_client
@@ -148,12 +146,7 @@ def migrate_chunks_from_vespa_to_opensearch_task(
             task_logger.error(err_str)
             return False
 
-        with (
-            get_session_with_current_tenant() as db_session,
-            get_vespa_http_client(
-                timeout=VESPA_MIGRATION_REQUEST_TIMEOUT_S
-            ) as vespa_client,
-        ):
+        with get_session_with_current_tenant() as db_session:
             try_insert_opensearch_tenant_migration_record_with_commit(db_session)
             search_settings = get_current_search_settings(db_session)
             tenant_state = TenantState(tenant_id=tenant_id, multitenant=MULTI_TENANT)
@@ -168,7 +161,6 @@ def migrate_chunks_from_vespa_to_opensearch_task(
                 index_name=search_settings.index_name,
                 tenant_state=tenant_state,
                 large_chunks_enabled=False,
-                httpx_client=vespa_client,
             )
 
             sanitized_doc_start_time = time.monotonic()

backend/onyx/background/celery/tasks/user_file_processing/tasks.py

@@ -414,31 +414,34 @@ def _process_user_file_with_indexing(
         raise RuntimeError(f"Indexing pipeline failed for user file {user_file_id}")
 
 
-def process_user_file_impl(
-    *, user_file_id: str, tenant_id: str, redis_locking: bool
+@shared_task(
+    name=OnyxCeleryTask.PROCESS_SINGLE_USER_FILE,
+    bind=True,
+    ignore_result=True,
+)
+def process_single_user_file(
+    self: Task, *, user_file_id: str, tenant_id: str  # noqa: ARG001
 ) -> None:
-    """Core implementation for processing a single user file.
-
-    When redis_locking=True, acquires a per-file Redis lock and clears the
-    queued-key guard (Celery path).  When redis_locking=False, skips all Redis
-    operations (BackgroundTask path).
-    """
-    task_logger.info(f"process_user_file_impl - Starting id={user_file_id}")
+    task_logger.info(f"process_single_user_file - Starting id={user_file_id}")
     start = time.monotonic()
 
-    file_lock: RedisLock | None = None
-    if redis_locking:
-        redis_client = get_redis_client(tenant_id=tenant_id)
-        redis_client.delete(_user_file_queued_key(user_file_id))
-        file_lock = redis_client.lock(
-            _user_file_lock_key(user_file_id),
-            timeout=CELERY_USER_FILE_PROCESSING_LOCK_TIMEOUT,
+    redis_client = get_redis_client(tenant_id=tenant_id)
+
+    # Clear the "queued" guard set by the beat generator so that the next beat
+    # cycle can re-enqueue this file if it is still in PROCESSING state after
+    # this task completes or fails.
+    redis_client.delete(_user_file_queued_key(user_file_id))
+
+    file_lock: RedisLock = redis_client.lock(
+        _user_file_lock_key(user_file_id),
+        timeout=CELERY_USER_FILE_PROCESSING_LOCK_TIMEOUT,
+    )
+
+    if not file_lock.acquire(blocking=False):
+        task_logger.info(
+            f"process_single_user_file - Lock held, skipping user_file_id={user_file_id}"
         )
-        if file_lock is not None and not file_lock.acquire(blocking=False):
-            task_logger.info(
-                f"process_user_file_impl - Lock held, skipping user_file_id={user_file_id}"
-            )
-            return
+        return None
 
     documents: list[Document] = []
     try:
@@ -446,18 +449,15 @@ def process_user_file_impl(
             uf = db_session.get(UserFile, _as_uuid(user_file_id))
             if not uf:
                 task_logger.warning(
-                    f"process_user_file_impl - UserFile not found id={user_file_id}"
+                    f"process_single_user_file - UserFile not found id={user_file_id}"
                 )
-                return
+                return None
 
-            if uf.status not in (
-                UserFileStatus.PROCESSING,
-                UserFileStatus.INDEXING,
-            ):
+            if uf.status != UserFileStatus.PROCESSING:
                 task_logger.info(
-                    f"process_user_file_impl - Skipping id={user_file_id} status={uf.status}"
+                    f"process_single_user_file - Skipping id={user_file_id} status={uf.status}"
                 )
-                return
+                return None
 
             connector = LocalFileConnector(
                 file_locations=[uf.file_id],
@@ -471,6 +471,7 @@ def process_user_file_impl(
                         [doc for doc in batch if not isinstance(doc, HierarchyNode)]
                     )
 
+                # update the document id to userfile id in the documents
                 for document in documents:
                     document.id = str(user_file_id)
                     document.source = DocumentSource.USER_FILE
@@ -492,8 +493,9 @@ def process_user_file_impl(
 
             except Exception as e:
                 task_logger.exception(
-                    f"process_user_file_impl - Error processing file id={user_file_id} - {e.__class__.__name__}"
+                    f"process_single_user_file - Error processing file id={user_file_id} - {e.__class__.__name__}"
                 )
+                # don't update the status if the user file is being deleted
                 current_user_file = db_session.get(UserFile, _as_uuid(user_file_id))
                 if (
                     current_user_file
@@ -502,43 +504,33 @@ def process_user_file_impl(
                     uf.status = UserFileStatus.FAILED
                     db_session.add(uf)
                     db_session.commit()
-                return
+                return None
 
         elapsed = time.monotonic() - start
         task_logger.info(
-            f"process_user_file_impl - Finished id={user_file_id} docs={len(documents)} elapsed={elapsed:.2f}s"
+            f"process_single_user_file - Finished id={user_file_id} docs={len(documents)} elapsed={elapsed:.2f}s"
         )
+        return None
     except Exception as e:
+        # Attempt to mark the file as failed
         with get_session_with_current_tenant() as db_session:
             uf = db_session.get(UserFile, _as_uuid(user_file_id))
             if uf:
+                # don't update the status if the user file is being deleted
                 if uf.status != UserFileStatus.DELETING:
                     uf.status = UserFileStatus.FAILED
                 db_session.add(uf)
                 db_session.commit()
 
         task_logger.exception(
-            f"process_user_file_impl - Error processing file id={user_file_id} - {e.__class__.__name__}"
+            f"process_single_user_file - Error processing file id={user_file_id} - {e.__class__.__name__}"
         )
-        raise
+        return None
     finally:
-        if file_lock is not None and file_lock.owned():
+        if file_lock.owned():
             file_lock.release()
 
 
-@shared_task(
-    name=OnyxCeleryTask.PROCESS_SINGLE_USER_FILE,
-    bind=True,
-    ignore_result=True,
-)
-def process_single_user_file(
-    self: Task, *, user_file_id: str, tenant_id: str  # noqa: ARG001
-) -> None:
-    process_user_file_impl(
-        user_file_id=user_file_id, tenant_id=tenant_id, redis_locking=True
-    )
-
-
 @shared_task(
     name=OnyxCeleryTask.CHECK_FOR_USER_FILE_DELETE,
     soft_time_limit=300,
@@ -589,38 +581,36 @@ def check_for_user_file_delete(self: Task, *, tenant_id: str) -> None:
     return None
 
 
-def delete_user_file_impl(
-    *, user_file_id: str, tenant_id: str, redis_locking: bool
+@shared_task(
+    name=OnyxCeleryTask.DELETE_SINGLE_USER_FILE,
+    bind=True,
+    ignore_result=True,
+)
+def process_single_user_file_delete(
+    self: Task, *, user_file_id: str, tenant_id: str  # noqa: ARG001
 ) -> None:
-    """Core implementation for deleting a single user file.
-
-    When redis_locking=True, acquires a per-file Redis lock (Celery path).
-    When redis_locking=False, skips Redis operations (BackgroundTask path).
-    """
-    task_logger.info(f"delete_user_file_impl - Starting id={user_file_id}")
-
-    file_lock: RedisLock | None = None
-    if redis_locking:
-        redis_client = get_redis_client(tenant_id=tenant_id)
-        file_lock = redis_client.lock(
-            _user_file_delete_lock_key(user_file_id),
-            timeout=CELERY_GENERIC_BEAT_LOCK_TIMEOUT,
+    """Process a single user file delete."""
+    task_logger.info(f"process_single_user_file_delete - Starting id={user_file_id}")
+    redis_client = get_redis_client(tenant_id=tenant_id)
+    file_lock: RedisLock = redis_client.lock(
+        _user_file_delete_lock_key(user_file_id),
+        timeout=CELERY_GENERIC_BEAT_LOCK_TIMEOUT,
+    )
+    if not file_lock.acquire(blocking=False):
+        task_logger.info(
+            f"process_single_user_file_delete - Lock held, skipping user_file_id={user_file_id}"
         )
-        if file_lock is not None and not file_lock.acquire(blocking=False):
-            task_logger.info(
-                f"delete_user_file_impl - Lock held, skipping user_file_id={user_file_id}"
-            )
-            return
-
+        return None
     try:
         with get_session_with_current_tenant() as db_session:
             user_file = db_session.get(UserFile, _as_uuid(user_file_id))
             if not user_file:
                 task_logger.info(
-                    f"delete_user_file_impl - User file not found id={user_file_id}"
+                    f"process_single_user_file_delete - User file not found id={user_file_id}"
                 )
-                return
+                return None
 
+            # 1) Delete vector DB chunks (skip when disabled)
             if not DISABLE_VECTOR_DB:
                 if MANAGED_VESPA:
                     httpx_init_vespa_pool(
@@ -658,6 +648,7 @@ def delete_user_file_impl(
                         chunk_count=chunk_count,
                     )
 
+            # 2) Delete the user-uploaded file content from filestore (blob + metadata)
             file_store = get_default_file_store()
             try:
                 file_store.delete_file(user_file.file_id)
@@ -665,34 +656,26 @@ def delete_user_file_impl(
                     user_file_id_to_plaintext_file_name(user_file.id)
                 )
             except Exception as e:
+                # This block executed only if the file is not found in the filestore
                 task_logger.exception(
-                    f"delete_user_file_impl - Error deleting file id={user_file.id} - {e.__class__.__name__}"
+                    f"process_single_user_file_delete - Error deleting file id={user_file.id} - {e.__class__.__name__}"
                 )
 
+            # 3) Finally, delete the UserFile row
             db_session.delete(user_file)
             db_session.commit()
-            task_logger.info(f"delete_user_file_impl - Completed id={user_file_id}")
+            task_logger.info(
+                f"process_single_user_file_delete - Completed id={user_file_id}"
+            )
     except Exception as e:
         task_logger.exception(
-            f"delete_user_file_impl - Error processing file id={user_file_id} - {e.__class__.__name__}"
+            f"process_single_user_file_delete - Error processing file id={user_file_id} - {e.__class__.__name__}"
         )
-        raise
+        return None
     finally:
-        if file_lock is not None and file_lock.owned():
+        if file_lock.owned():
             file_lock.release()
-
-
-@shared_task(
-    name=OnyxCeleryTask.DELETE_SINGLE_USER_FILE,
-    bind=True,
-    ignore_result=True,
-)
-def process_single_user_file_delete(
-    self: Task, *, user_file_id: str, tenant_id: str  # noqa: ARG001
-) -> None:
-    delete_user_file_impl(
-        user_file_id=user_file_id, tenant_id=tenant_id, redis_locking=True
-    )
+    return None
 
 
 @shared_task(
@@ -764,30 +747,32 @@ def check_for_user_file_project_sync(self: Task, *, tenant_id: str) -> None:
     return None
 
 
-def project_sync_user_file_impl(
-    *, user_file_id: str, tenant_id: str, redis_locking: bool
+@shared_task(
+    name=OnyxCeleryTask.PROCESS_SINGLE_USER_FILE_PROJECT_SYNC,
+    bind=True,
+    ignore_result=True,
+)
+def process_single_user_file_project_sync(
+    self: Task, *, user_file_id: str, tenant_id: str  # noqa: ARG001
 ) -> None:
-    """Core implementation for syncing a user file's project/persona metadata.
+    """Process a single user file project sync."""
+    task_logger.info(
+        f"process_single_user_file_project_sync - Starting id={user_file_id}"
+    )
 
-    When redis_locking=True, acquires a per-file Redis lock and clears the
-    queued-key guard (Celery path).  When redis_locking=False, skips Redis
-    operations (BackgroundTask path).
-    """
-    task_logger.info(f"project_sync_user_file_impl - Starting id={user_file_id}")
+    redis_client = get_redis_client(tenant_id=tenant_id)
+    redis_client.delete(_user_file_project_sync_queued_key(user_file_id))
 
-    file_lock: RedisLock | None = None
-    if redis_locking:
-        redis_client = get_redis_client(tenant_id=tenant_id)
-        redis_client.delete(_user_file_project_sync_queued_key(user_file_id))
-        file_lock = redis_client.lock(
-            user_file_project_sync_lock_key(user_file_id),
-            timeout=CELERY_USER_FILE_PROJECT_SYNC_LOCK_TIMEOUT,
+    file_lock: RedisLock = redis_client.lock(
+        user_file_project_sync_lock_key(user_file_id),
+        timeout=CELERY_USER_FILE_PROJECT_SYNC_LOCK_TIMEOUT,
+    )
+
+    if not file_lock.acquire(blocking=False):
+        task_logger.info(
+            f"process_single_user_file_project_sync - Lock held, skipping user_file_id={user_file_id}"
         )
-        if file_lock is not None and not file_lock.acquire(blocking=False):
-            task_logger.info(
-                f"project_sync_user_file_impl - Lock held, skipping user_file_id={user_file_id}"
-            )
-            return
+        return None
 
     try:
         with get_session_with_current_tenant() as db_session:
@@ -798,10 +783,11 @@ def project_sync_user_file_impl(
             ).scalar_one_or_none()
             if not user_file:
                 task_logger.info(
-                    f"project_sync_user_file_impl - User file not found id={user_file_id}"
+                    f"process_single_user_file_project_sync - User file not found id={user_file_id}"
                 )
-                return
+                return None
 
+            # Sync project metadata to vector DB (skip when disabled)
             if not DISABLE_VECTOR_DB:
                 if MANAGED_VESPA:
                     httpx_init_vespa_pool(
@@ -836,7 +822,7 @@ def project_sync_user_file_impl(
                     )
 
             task_logger.info(
-                f"project_sync_user_file_impl - User file id={user_file_id}"
+                f"process_single_user_file_project_sync - User file id={user_file_id}"
             )
 
             user_file.needs_project_sync = False
@@ -849,22 +835,11 @@ def project_sync_user_file_impl(
 
     except Exception as e:
         task_logger.exception(
-            f"project_sync_user_file_impl - Error syncing project for file id={user_file_id} - {e.__class__.__name__}"
+            f"process_single_user_file_project_sync - Error syncing project for file id={user_file_id} - {e.__class__.__name__}"
         )
-        raise
+        return None
     finally:
-        if file_lock is not None and file_lock.owned():
+        if file_lock.owned():
             file_lock.release()
 
-
-@shared_task(
-    name=OnyxCeleryTask.PROCESS_SINGLE_USER_FILE_PROJECT_SYNC,
-    bind=True,
-    ignore_result=True,
-)
-def process_single_user_file_project_sync(
-    self: Task, *, user_file_id: str, tenant_id: str  # noqa: ARG001
-) -> None:
-    project_sync_user_file_impl(
-        user_file_id=user_file_id, tenant_id=tenant_id, redis_locking=True
-    )
+    return None

backend/onyx/background/celery/versioned_apps/background.py

@@ -0,0 +1,10 @@
+from celery import Celery
+
+from onyx.utils.variable_functionality import fetch_versioned_implementation
+from onyx.utils.variable_functionality import set_is_ee_based_on_env_variable
+
+set_is_ee_based_on_env_variable()
+app: Celery = fetch_versioned_implementation(
+    "onyx.background.celery.apps.background",
+    "celery_app",
+)

backend/onyx/background/periodic_poller.py

@@ -1,307 +0,0 @@
-"""Periodic poller for NO_VECTOR_DB deployments.
-
-Replaces Celery Beat and background workers with a lightweight daemon thread
-that runs from the API server process.  Two responsibilities:
-
-1. Recovery polling (every 30 s): re-processes user files stuck in
-   PROCESSING / DELETING / needs_sync states via the drain loops defined
-   in ``task_utils.py``.
-
-2. Periodic task execution (configurable intervals): runs LLM model updates
-   and scheduled evals at their configured cadences, with Postgres advisory
-   lock deduplication across multiple API server instances.
-"""
-
-import threading
-import time
-from collections.abc import Callable
-from dataclasses import dataclass
-from dataclasses import field
-
-from onyx.utils.logger import setup_logger
-
-logger = setup_logger()
-
-RECOVERY_INTERVAL_SECONDS = 30
-PERIODIC_TASK_LOCK_BASE = 20_000
-PERIODIC_TASK_KV_PREFIX = "periodic_poller:last_claimed:"
-
-
-# ------------------------------------------------------------------
-# Periodic task definitions
-# ------------------------------------------------------------------
-
-
-_NEVER_RAN: float = -1e18
-
-
-@dataclass
-class _PeriodicTaskDef:
-    name: str
-    interval_seconds: float
-    lock_id: int
-    run_fn: Callable[[], None]
-    last_run_at: float = field(default=_NEVER_RAN)
-
-
-def _run_auto_llm_update() -> None:
-    from onyx.configs.app_configs import AUTO_LLM_CONFIG_URL
-
-    if not AUTO_LLM_CONFIG_URL:
-        return
-
-    from onyx.db.engine.sql_engine import get_session_with_current_tenant
-    from onyx.llm.well_known_providers.auto_update_service import (
-        sync_llm_models_from_github,
-    )
-
-    with get_session_with_current_tenant() as db_session:
-        sync_llm_models_from_github(db_session)
-
-
-def _run_cache_cleanup() -> None:
-    from onyx.cache.postgres_backend import cleanup_expired_cache_entries
-
-    cleanup_expired_cache_entries()
-
-
-def _run_scheduled_eval() -> None:
-    from onyx.configs.app_configs import BRAINTRUST_API_KEY
-    from onyx.configs.app_configs import SCHEDULED_EVAL_DATASET_NAMES
-    from onyx.configs.app_configs import SCHEDULED_EVAL_PERMISSIONS_EMAIL
-    from onyx.configs.app_configs import SCHEDULED_EVAL_PROJECT
-
-    if not all(
-        [
-            BRAINTRUST_API_KEY,
-            SCHEDULED_EVAL_PROJECT,
-            SCHEDULED_EVAL_DATASET_NAMES,
-            SCHEDULED_EVAL_PERMISSIONS_EMAIL,
-        ]
-    ):
-        return
-
-    from datetime import datetime
-    from datetime import timezone
-
-    from onyx.evals.eval import run_eval
-    from onyx.evals.models import EvalConfigurationOptions
-
-    run_timestamp = datetime.now(timezone.utc).strftime("%Y-%m-%d")
-    for dataset_name in SCHEDULED_EVAL_DATASET_NAMES:
-        try:
-            run_eval(
-                configuration=EvalConfigurationOptions(
-                    search_permissions_email=SCHEDULED_EVAL_PERMISSIONS_EMAIL,
-                    dataset_name=dataset_name,
-                    no_send_logs=False,
-                    braintrust_project=SCHEDULED_EVAL_PROJECT,
-                    experiment_name=f"{dataset_name} - {run_timestamp}",
-                ),
-                remote_dataset_name=dataset_name,
-            )
-        except Exception:
-            logger.exception(
-                f"Periodic poller - Failed scheduled eval for dataset {dataset_name}"
-            )
-
-
-_CACHE_CLEANUP_INTERVAL_SECONDS = 300
-
-
-def _build_periodic_tasks() -> list[_PeriodicTaskDef]:
-    from onyx.cache.interface import CacheBackendType
-    from onyx.configs.app_configs import AUTO_LLM_CONFIG_URL
-    from onyx.configs.app_configs import AUTO_LLM_UPDATE_INTERVAL_SECONDS
-    from onyx.configs.app_configs import CACHE_BACKEND
-    from onyx.configs.app_configs import SCHEDULED_EVAL_DATASET_NAMES
-
-    tasks: list[_PeriodicTaskDef] = []
-    if CACHE_BACKEND == CacheBackendType.POSTGRES:
-        tasks.append(
-            _PeriodicTaskDef(
-                name="cache-cleanup",
-                interval_seconds=_CACHE_CLEANUP_INTERVAL_SECONDS,
-                lock_id=PERIODIC_TASK_LOCK_BASE + 2,
-                run_fn=_run_cache_cleanup,
-            )
-        )
-    if AUTO_LLM_CONFIG_URL:
-        tasks.append(
-            _PeriodicTaskDef(
-                name="auto-llm-update",
-                interval_seconds=AUTO_LLM_UPDATE_INTERVAL_SECONDS,
-                lock_id=PERIODIC_TASK_LOCK_BASE,
-                run_fn=_run_auto_llm_update,
-            )
-        )
-    if SCHEDULED_EVAL_DATASET_NAMES:
-        tasks.append(
-            _PeriodicTaskDef(
-                name="scheduled-eval",
-                interval_seconds=7 * 24 * 3600,
-                lock_id=PERIODIC_TASK_LOCK_BASE + 1,
-                run_fn=_run_scheduled_eval,
-            )
-        )
-    return tasks
-
-
-# ------------------------------------------------------------------
-# Periodic task runner with advisory-lock-guarded claim
-# ------------------------------------------------------------------
-
-
-def _try_claim_task(task_def: _PeriodicTaskDef) -> bool:
-    """Atomically check whether *task_def* should run and record a claim.
-
-    Uses a transaction-scoped advisory lock for atomicity combined with a
-    ``KVStore`` timestamp for cross-instance dedup.  The DB session is held
-    only for this brief claim transaction, not during task execution.
-    """
-    from datetime import datetime
-    from datetime import timezone
-
-    from sqlalchemy import text
-
-    from onyx.db.engine.sql_engine import get_session_with_current_tenant
-    from onyx.db.models import KVStore
-
-    kv_key = PERIODIC_TASK_KV_PREFIX + task_def.name
-
-    with get_session_with_current_tenant() as db_session:
-        acquired = db_session.execute(
-            text("SELECT pg_try_advisory_xact_lock(:id)"),
-            {"id": task_def.lock_id},
-        ).scalar()
-        if not acquired:
-            return False
-
-        row = db_session.query(KVStore).filter_by(key=kv_key).first()
-        if row and row.value is not None:
-            last_claimed = datetime.fromisoformat(str(row.value))
-            elapsed = (datetime.now(timezone.utc) - last_claimed).total_seconds()
-            if elapsed < task_def.interval_seconds:
-                return False
-
-        now_ts = datetime.now(timezone.utc).isoformat()
-        if row:
-            row.value = now_ts
-        else:
-            db_session.add(KVStore(key=kv_key, value=now_ts))
-        db_session.commit()
-
-    return True
-
-
-def _try_run_periodic_task(task_def: _PeriodicTaskDef) -> None:
-    """Run *task_def* if its interval has elapsed and no peer holds the lock."""
-    now = time.monotonic()
-    if now - task_def.last_run_at < task_def.interval_seconds:
-        return
-
-    if not _try_claim_task(task_def):
-        return
-
-    try:
-        task_def.run_fn()
-        task_def.last_run_at = now
-    except Exception:
-        logger.exception(
-            f"Periodic poller - Error running periodic task {task_def.name}"
-        )
-
-
-# ------------------------------------------------------------------
-# Recovery / drain loop runner
-# ------------------------------------------------------------------
-
-
-def _run_drain_loops(tenant_id: str) -> None:
-    from onyx.background.task_utils import drain_delete_loop
-    from onyx.background.task_utils import drain_processing_loop
-    from onyx.background.task_utils import drain_project_sync_loop
-
-    drain_processing_loop(tenant_id)
-    drain_delete_loop(tenant_id)
-    drain_project_sync_loop(tenant_id)
-
-
-# ------------------------------------------------------------------
-# Startup recovery (10g)
-# ------------------------------------------------------------------
-
-
-def recover_stuck_user_files(tenant_id: str) -> None:
-    """Run all drain loops once to re-process files left in intermediate states.
-
-    Called from ``lifespan()`` on startup when ``DISABLE_VECTOR_DB`` is set.
-    """
-    logger.info("recover_stuck_user_files - Checking for stuck user files")
-    try:
-        _run_drain_loops(tenant_id)
-    except Exception:
-        logger.exception("recover_stuck_user_files - Error during recovery")
-
-
-# ------------------------------------------------------------------
-# Daemon thread (10f)
-# ------------------------------------------------------------------
-
-_shutdown_event = threading.Event()
-_poller_thread: threading.Thread | None = None
-
-
-def _poller_loop(tenant_id: str) -> None:
-    from shared_configs.contextvars import CURRENT_TENANT_ID_CONTEXTVAR
-
-    CURRENT_TENANT_ID_CONTEXTVAR.set(tenant_id)
-
-    periodic_tasks = _build_periodic_tasks()
-    logger.info(
-        f"Periodic poller started with {len(periodic_tasks)} periodic task(s): "
-        f"{[t.name for t in periodic_tasks]}"
-    )
-
-    while not _shutdown_event.is_set():
-        try:
-            _run_drain_loops(tenant_id)
-        except Exception:
-            logger.exception("Periodic poller - Error in recovery polling")
-
-        for task_def in periodic_tasks:
-            try:
-                _try_run_periodic_task(task_def)
-            except Exception:
-                logger.exception(
-                    f"Periodic poller - Unhandled error checking task {task_def.name}"
-                )
-
-        _shutdown_event.wait(RECOVERY_INTERVAL_SECONDS)
-
-
-def start_periodic_poller(tenant_id: str) -> None:
-    """Start the periodic poller daemon thread."""
-    global _poller_thread  # noqa: PLW0603
-    _shutdown_event.clear()
-    _poller_thread = threading.Thread(
-        target=_poller_loop,
-        args=(tenant_id,),
-        daemon=True,
-        name="no-vectordb-periodic-poller",
-    )
-    _poller_thread.start()
-    logger.info("Periodic poller thread started")
-
-
-def stop_periodic_poller() -> None:
-    """Signal the periodic poller to stop and wait for it to exit."""
-    global _poller_thread  # noqa: PLW0603
-    if _poller_thread is None:
-        return
-    _shutdown_event.set()
-    _poller_thread.join(timeout=10)
-    if _poller_thread.is_alive():
-        logger.warning("Periodic poller thread did not stop within timeout")
-    _poller_thread = None
-    logger.info("Periodic poller thread stopped")

backend/onyx/background/task_utils.py

@@ -1,33 +1,3 @@
-"""Background task utilities.
-
-Contains query-history report helpers (used by all deployment modes) and
-in-process background task execution helpers for NO_VECTOR_DB mode:
-
-- Atomic claim-and-mark helpers that prevent duplicate processing
-- Drain loops that process all pending user file work
-
-Each claim function runs a short-lived transaction: SELECT ... FOR UPDATE
-SKIP LOCKED, UPDATE the row to remove it from future queries, COMMIT.
-After the commit the row lock is released, but the row is no longer
-eligible for re-claiming.  No long-lived sessions or advisory locks.
-"""
-
-from uuid import UUID
-
-import sqlalchemy as sa
-from sqlalchemy import select
-from sqlalchemy.orm import Session
-
-from onyx.db.enums import UserFileStatus
-from onyx.db.models import UserFile
-from onyx.utils.logger import setup_logger
-
-logger = setup_logger()
-
-# ------------------------------------------------------------------
-# Query-history report helpers (pre-existing, used by all modes)
-# ------------------------------------------------------------------
-
 QUERY_REPORT_NAME_PREFIX = "query-history"
 
 
@@ -39,168 +9,3 @@ def construct_query_history_report_name(
 
 def extract_task_id_from_query_history_report_name(name: str) -> str:
     return name.removeprefix(f"{QUERY_REPORT_NAME_PREFIX}-").removesuffix(".csv")
-
-
-# ------------------------------------------------------------------
-# Atomic claim-and-mark helpers
-# ------------------------------------------------------------------
-# Each function runs inside a single short-lived session/transaction:
-#   1. SELECT ... FOR UPDATE SKIP LOCKED  (locks one eligible row)
-#   2. UPDATE the row so it is no longer eligible
-#   3. COMMIT  (releases the row lock)
-# After the commit, no other drain loop can claim the same row.
-
-
-def _claim_next_processing_file(db_session: Session) -> UUID | None:
-    """Claim the next PROCESSING file by transitioning it to INDEXING.
-
-    Returns the file id, or None when no eligible files remain.
-    """
-    file_id = db_session.execute(
-        select(UserFile.id)
-        .where(UserFile.status == UserFileStatus.PROCESSING)
-        .order_by(UserFile.created_at)
-        .limit(1)
-        .with_for_update(skip_locked=True)
-    ).scalar_one_or_none()
-    if file_id is None:
-        return None
-
-    db_session.execute(
-        sa.update(UserFile)
-        .where(UserFile.id == file_id)
-        .values(status=UserFileStatus.INDEXING)
-    )
-    db_session.commit()
-    return file_id
-
-
-def _claim_next_deleting_file(
-    db_session: Session,
-    exclude_ids: set[UUID] | None = None,
-) -> UUID | None:
-    """Claim the next DELETING file.
-
-    No status transition needed — the impl deletes the row on success.
-    The short-lived FOR UPDATE lock prevents concurrent claims.
-    *exclude_ids* prevents re-processing the same file if the impl fails.
-    """
-    stmt = (
-        select(UserFile.id)
-        .where(UserFile.status == UserFileStatus.DELETING)
-        .order_by(UserFile.created_at)
-        .limit(1)
-        .with_for_update(skip_locked=True)
-    )
-    if exclude_ids:
-        stmt = stmt.where(UserFile.id.notin_(exclude_ids))
-    file_id = db_session.execute(stmt).scalar_one_or_none()
-    db_session.commit()
-    return file_id
-
-
-def _claim_next_sync_file(
-    db_session: Session,
-    exclude_ids: set[UUID] | None = None,
-) -> UUID | None:
-    """Claim the next file needing project/persona sync.
-
-    No status transition needed — the impl clears the sync flags on
-    success.  The short-lived FOR UPDATE lock prevents concurrent claims.
-    *exclude_ids* prevents re-processing the same file if the impl fails.
-    """
-    stmt = (
-        select(UserFile.id)
-        .where(
-            sa.and_(
-                sa.or_(
-                    UserFile.needs_project_sync.is_(True),
-                    UserFile.needs_persona_sync.is_(True),
-                ),
-                UserFile.status == UserFileStatus.COMPLETED,
-            )
-        )
-        .order_by(UserFile.created_at)
-        .limit(1)
-        .with_for_update(skip_locked=True)
-    )
-    if exclude_ids:
-        stmt = stmt.where(UserFile.id.notin_(exclude_ids))
-    file_id = db_session.execute(stmt).scalar_one_or_none()
-    db_session.commit()
-    return file_id
-
-
-# ------------------------------------------------------------------
-# Drain loops — process *all* pending work of each type
-# ------------------------------------------------------------------
-
-
-def drain_processing_loop(tenant_id: str) -> None:
-    """Process all pending PROCESSING user files."""
-    from onyx.background.celery.tasks.user_file_processing.tasks import (
-        process_user_file_impl,
-    )
-    from onyx.db.engine.sql_engine import get_session_with_current_tenant
-
-    while True:
-        with get_session_with_current_tenant() as session:
-            file_id = _claim_next_processing_file(session)
-        if file_id is None:
-            break
-        try:
-            process_user_file_impl(
-                user_file_id=str(file_id),
-                tenant_id=tenant_id,
-                redis_locking=False,
-            )
-        except Exception:
-            logger.exception(f"Failed to process user file {file_id}")
-
-
-def drain_delete_loop(tenant_id: str) -> None:
-    """Delete all pending DELETING user files."""
-    from onyx.background.celery.tasks.user_file_processing.tasks import (
-        delete_user_file_impl,
-    )
-    from onyx.db.engine.sql_engine import get_session_with_current_tenant
-
-    failed: set[UUID] = set()
-    while True:
-        with get_session_with_current_tenant() as session:
-            file_id = _claim_next_deleting_file(session, exclude_ids=failed)
-        if file_id is None:
-            break
-        try:
-            delete_user_file_impl(
-                user_file_id=str(file_id),
-                tenant_id=tenant_id,
-                redis_locking=False,
-            )
-        except Exception:
-            logger.exception(f"Failed to delete user file {file_id}")
-            failed.add(file_id)
-
-
-def drain_project_sync_loop(tenant_id: str) -> None:
-    """Sync all pending project/persona metadata for user files."""
-    from onyx.background.celery.tasks.user_file_processing.tasks import (
-        project_sync_user_file_impl,
-    )
-    from onyx.db.engine.sql_engine import get_session_with_current_tenant
-
-    failed: set[UUID] = set()
-    while True:
-        with get_session_with_current_tenant() as session:
-            file_id = _claim_next_sync_file(session, exclude_ids=failed)
-        if file_id is None:
-            break
-        try:
-            project_sync_user_file_impl(
-                user_file_id=str(file_id),
-                tenant_id=tenant_id,
-                redis_locking=False,
-            )
-        except Exception:
-            logger.exception(f"Failed to sync user file {file_id}")
-            failed.add(file_id)

backend/onyx/cache/factory.py

@@ -1,51 +0,0 @@
-from collections.abc import Callable
-
-from onyx.cache.interface import CacheBackend
-from onyx.cache.interface import CacheBackendType
-from onyx.configs.app_configs import CACHE_BACKEND
-
-
-def _build_redis_backend(tenant_id: str) -> CacheBackend:
-    from onyx.cache.redis_backend import RedisCacheBackend
-    from onyx.redis.redis_pool import redis_pool
-
-    return RedisCacheBackend(redis_pool.get_client(tenant_id))
-
-
-def _build_postgres_backend(tenant_id: str) -> CacheBackend:
-    from onyx.cache.postgres_backend import PostgresCacheBackend
-
-    return PostgresCacheBackend(tenant_id)
-
-
-_BACKEND_BUILDERS: dict[CacheBackendType, Callable[[str], CacheBackend]] = {
-    CacheBackendType.REDIS: _build_redis_backend,
-    CacheBackendType.POSTGRES: _build_postgres_backend,
-}
-
-
-def get_cache_backend(*, tenant_id: str | None = None) -> CacheBackend:
-    """Return a tenant-aware ``CacheBackend``.
-
-    If *tenant_id* is ``None``, the current tenant is read from the
-    thread-local context variable (same behaviour as ``get_redis_client``).
-    """
-    if tenant_id is None:
-        from shared_configs.contextvars import get_current_tenant_id
-
-        tenant_id = get_current_tenant_id()
-
-    builder = _BACKEND_BUILDERS.get(CACHE_BACKEND)
-    if builder is None:
-        raise ValueError(
-            f"Unsupported CACHE_BACKEND={CACHE_BACKEND!r}. "
-            f"Supported values: {[t.value for t in CacheBackendType]}"
-        )
-    return builder(tenant_id)
-
-
-def get_shared_cache_backend() -> CacheBackend:
-    """Return a ``CacheBackend`` in the shared (cross-tenant) namespace."""
-    from shared_configs.configs import DEFAULT_REDIS_PREFIX
-
-    return get_cache_backend(tenant_id=DEFAULT_REDIS_PREFIX)

backend/onyx/cache/interface.py

@@ -1,115 +0,0 @@
-import abc
-from enum import Enum
-
-from redis.exceptions import RedisError
-from sqlalchemy.exc import SQLAlchemyError
-
-TTL_KEY_NOT_FOUND = -2
-TTL_NO_EXPIRY = -1
-
-CACHE_TRANSIENT_ERRORS: tuple[type[Exception], ...] = (RedisError, SQLAlchemyError)
-"""Exception types that represent transient cache connectivity / operational
-failures.  Callers that want to fail-open (or fail-closed) on cache errors
-should catch this tuple instead of bare ``Exception``.
-
-When adding a new ``CacheBackend`` implementation, add its transient error
-base class(es) here so all call-sites pick it up automatically."""
-
-
-class CacheBackendType(str, Enum):
-    REDIS = "redis"
-    POSTGRES = "postgres"
-
-
-class CacheLock(abc.ABC):
-    """Abstract distributed lock returned by CacheBackend.lock()."""
-
-    @abc.abstractmethod
-    def acquire(
-        self,
-        blocking: bool = True,
-        blocking_timeout: float | None = None,
-    ) -> bool:
-        raise NotImplementedError
-
-    @abc.abstractmethod
-    def release(self) -> None:
-        raise NotImplementedError
-
-    @abc.abstractmethod
-    def owned(self) -> bool:
-        raise NotImplementedError
-
-    def __enter__(self) -> "CacheLock":
-        if not self.acquire():
-            raise RuntimeError("Failed to acquire lock")
-        return self
-
-    def __exit__(self, *args: object) -> None:
-        self.release()
-
-
-class CacheBackend(abc.ABC):
-    """Thin abstraction over a key-value cache with TTL, locks, and blocking lists.
-
-    Covers the subset of Redis operations used outside of Celery. When
-    CACHE_BACKEND=postgres, a PostgreSQL-backed implementation is used instead.
-    """
-
-    # -- basic key/value ---------------------------------------------------
-
-    @abc.abstractmethod
-    def get(self, key: str) -> bytes | None:
-        raise NotImplementedError
-
-    @abc.abstractmethod
-    def set(
-        self,
-        key: str,
-        value: str | bytes | int | float,
-        ex: int | None = None,
-    ) -> None:
-        raise NotImplementedError
-
-    @abc.abstractmethod
-    def delete(self, key: str) -> None:
-        raise NotImplementedError
-
-    @abc.abstractmethod
-    def exists(self, key: str) -> bool:
-        raise NotImplementedError
-
-    # -- TTL ---------------------------------------------------------------
-
-    @abc.abstractmethod
-    def expire(self, key: str, seconds: int) -> None:
-        raise NotImplementedError
-
-    @abc.abstractmethod
-    def ttl(self, key: str) -> int:
-        """Return remaining TTL in seconds.
-
-        Returns ``TTL_NO_EXPIRY`` (-1) if key exists without expiry,
-        ``TTL_KEY_NOT_FOUND`` (-2) if key is missing or expired.
-        """
-        raise NotImplementedError
-
-    # -- distributed lock --------------------------------------------------
-
-    @abc.abstractmethod
-    def lock(self, name: str, timeout: float | None = None) -> CacheLock:
-        raise NotImplementedError
-
-    # -- blocking list (used by MCP OAuth BLPOP pattern) -------------------
-
-    @abc.abstractmethod
-    def rpush(self, key: str, value: str | bytes) -> None:
-        raise NotImplementedError
-
-    @abc.abstractmethod
-    def blpop(self, keys: list[str], timeout: int = 0) -> tuple[bytes, bytes] | None:
-        """Block until a value is available on one of *keys*, or *timeout* expires.
-
-        Returns ``(key, value)`` or ``None`` on timeout.
-        """
-        raise NotImplementedError

backend/onyx/cache/postgres_backend.py

@@ -1,323 +0,0 @@
-"""PostgreSQL-backed ``CacheBackend`` for NO_VECTOR_DB deployments.
-
-Uses the ``cache_store`` table for key-value storage, PostgreSQL advisory locks
-for distributed locking, and a polling loop for the BLPOP pattern.
-"""
-
-import hashlib
-import struct
-import time
-import uuid
-from contextlib import AbstractContextManager
-from datetime import datetime
-from datetime import timedelta
-from datetime import timezone
-
-from sqlalchemy import delete
-from sqlalchemy import func
-from sqlalchemy import or_
-from sqlalchemy import select
-from sqlalchemy import update
-from sqlalchemy.dialects.postgresql import insert as pg_insert
-from sqlalchemy.orm import Session
-
-from onyx.cache.interface import CacheBackend
-from onyx.cache.interface import CacheLock
-from onyx.cache.interface import TTL_KEY_NOT_FOUND
-from onyx.cache.interface import TTL_NO_EXPIRY
-from onyx.db.models import CacheStore
-
-_LIST_KEY_PREFIX = "_q:"
-# ASCII: ':' (0x3A) < ';' (0x3B). Upper bound for range queries so [prefix+, prefix;)
-# captures all list-item keys (e.g. _q:mylist:123:uuid) without including other
-# lists whose names share a prefix (e.g. _q:mylist2:...).
-_LIST_KEY_RANGE_TERMINATOR = ";"
-_LIST_ITEM_TTL_SECONDS = 3600
-_LOCK_POLL_INTERVAL = 0.1
-_BLPOP_POLL_INTERVAL = 0.25
-
-
-def _list_item_key(key: str) -> str:
-    """Unique key for a list item. Timestamp for FIFO ordering; UUID prevents
-    collision when concurrent rpush calls occur within the same nanosecond.
-    """
-    return f"{_LIST_KEY_PREFIX}{key}:{time.time_ns()}:{uuid.uuid4().hex}"
-
-
-def _to_bytes(value: str | bytes | int | float) -> bytes:
-    if isinstance(value, bytes):
-        return value
-    return str(value).encode()
-
-
-# ------------------------------------------------------------------
-# Lock
-# ------------------------------------------------------------------
-
-
-class PostgresCacheLock(CacheLock):
-    """Advisory-lock-based distributed lock.
-
-    Uses ``get_session_with_tenant`` for connection lifecycle.  The lock is tied
-    to the session's connection; releasing or closing the session frees it.
-
-    NOTE: Unlike Redis locks, advisory locks do not auto-expire after
-    ``timeout`` seconds.  They are released when ``release()`` is
-    called or when the session is closed.
-    """
-
-    def __init__(self, lock_id: int, timeout: float | None, tenant_id: str) -> None:
-        self._lock_id = lock_id
-        self._timeout = timeout
-        self._tenant_id = tenant_id
-        self._session_cm: AbstractContextManager[Session] | None = None
-        self._session: Session | None = None
-        self._acquired = False
-
-    def acquire(
-        self,
-        blocking: bool = True,
-        blocking_timeout: float | None = None,
-    ) -> bool:
-        from onyx.db.engine.sql_engine import get_session_with_tenant
-
-        self._session_cm = get_session_with_tenant(tenant_id=self._tenant_id)
-        self._session = self._session_cm.__enter__()
-        try:
-            if not blocking:
-                return self._try_lock()
-
-            effective_timeout = blocking_timeout or self._timeout
-            deadline = (
-                (time.monotonic() + effective_timeout) if effective_timeout else None
-            )
-            while True:
-                if self._try_lock():
-                    return True
-                if deadline is not None and time.monotonic() >= deadline:
-                    return False
-                time.sleep(_LOCK_POLL_INTERVAL)
-        finally:
-            if not self._acquired:
-                self._close_session()
-
-    def release(self) -> None:
-        if not self._acquired or self._session is None:
-            return
-        try:
-            self._session.execute(select(func.pg_advisory_unlock(self._lock_id)))
-        finally:
-            self._acquired = False
-            self._close_session()
-
-    def owned(self) -> bool:
-        return self._acquired
-
-    def _close_session(self) -> None:
-        if self._session_cm is not None:
-            try:
-                self._session_cm.__exit__(None, None, None)
-            finally:
-                self._session_cm = None
-                self._session = None
-
-    def _try_lock(self) -> bool:
-        assert self._session is not None
-        result = self._session.execute(
-            select(func.pg_try_advisory_lock(self._lock_id))
-        ).scalar()
-        if result:
-            self._acquired = True
-            return True
-        return False
-
-
-# ------------------------------------------------------------------
-# Backend
-# ------------------------------------------------------------------
-
-
-class PostgresCacheBackend(CacheBackend):
-    """``CacheBackend`` backed by the ``cache_store`` table in PostgreSQL.
-
-    Each operation opens and closes its own database session so the backend
-    is safe to share across threads.  Tenant isolation is handled by
-    SQLAlchemy's ``schema_translate_map`` (set by ``get_session_with_tenant``).
-    """
-
-    def __init__(self, tenant_id: str) -> None:
-        self._tenant_id = tenant_id
-
-    # -- basic key/value ---------------------------------------------------
-
-    def get(self, key: str) -> bytes | None:
-        from onyx.db.engine.sql_engine import get_session_with_tenant
-
-        stmt = select(CacheStore.value).where(
-            CacheStore.key == key,
-            or_(CacheStore.expires_at.is_(None), CacheStore.expires_at > func.now()),
-        )
-        with get_session_with_tenant(tenant_id=self._tenant_id) as session:
-            value = session.execute(stmt).scalar_one_or_none()
-        if value is None:
-            return None
-        return bytes(value)
-
-    def set(
-        self,
-        key: str,
-        value: str | bytes | int | float,
-        ex: int | None = None,
-    ) -> None:
-        from onyx.db.engine.sql_engine import get_session_with_tenant
-
-        value_bytes = _to_bytes(value)
-        expires_at = (
-            datetime.now(timezone.utc) + timedelta(seconds=ex)
-            if ex is not None
-            else None
-        )
-        stmt = (
-            pg_insert(CacheStore)
-            .values(key=key, value=value_bytes, expires_at=expires_at)
-            .on_conflict_do_update(
-                index_elements=[CacheStore.key],
-                set_={"value": value_bytes, "expires_at": expires_at},
-            )
-        )
-        with get_session_with_tenant(tenant_id=self._tenant_id) as session:
-            session.execute(stmt)
-            session.commit()
-
-    def delete(self, key: str) -> None:
-        from onyx.db.engine.sql_engine import get_session_with_tenant
-
-        with get_session_with_tenant(tenant_id=self._tenant_id) as session:
-            session.execute(delete(CacheStore).where(CacheStore.key == key))
-            session.commit()
-
-    def exists(self, key: str) -> bool:
-        from onyx.db.engine.sql_engine import get_session_with_tenant
-
-        stmt = (
-            select(CacheStore.key)
-            .where(
-                CacheStore.key == key,
-                or_(
-                    CacheStore.expires_at.is_(None),
-                    CacheStore.expires_at > func.now(),
-                ),
-            )
-            .limit(1)
-        )
-        with get_session_with_tenant(tenant_id=self._tenant_id) as session:
-            return session.execute(stmt).first() is not None
-
-    # -- TTL ---------------------------------------------------------------
-
-    def expire(self, key: str, seconds: int) -> None:
-        from onyx.db.engine.sql_engine import get_session_with_tenant
-
-        new_exp = datetime.now(timezone.utc) + timedelta(seconds=seconds)
-        stmt = (
-            update(CacheStore).where(CacheStore.key == key).values(expires_at=new_exp)
-        )
-        with get_session_with_tenant(tenant_id=self._tenant_id) as session:
-            session.execute(stmt)
-            session.commit()
-
-    def ttl(self, key: str) -> int:
-        from onyx.db.engine.sql_engine import get_session_with_tenant
-
-        stmt = select(CacheStore.expires_at).where(CacheStore.key == key)
-        with get_session_with_tenant(tenant_id=self._tenant_id) as session:
-            result = session.execute(stmt).first()
-        if result is None:
-            return TTL_KEY_NOT_FOUND
-        expires_at: datetime | None = result[0]
-        if expires_at is None:
-            return TTL_NO_EXPIRY
-        remaining = (expires_at - datetime.now(timezone.utc)).total_seconds()
-        if remaining <= 0:
-            return TTL_KEY_NOT_FOUND
-        return int(remaining)
-
-    # -- distributed lock --------------------------------------------------
-
-    def lock(self, name: str, timeout: float | None = None) -> CacheLock:
-        return PostgresCacheLock(
-            self._lock_id_for(name), timeout, tenant_id=self._tenant_id
-        )
-
-    # -- blocking list (MCP OAuth BLPOP pattern) ---------------------------
-
-    def rpush(self, key: str, value: str | bytes) -> None:
-        self.set(_list_item_key(key), value, ex=_LIST_ITEM_TTL_SECONDS)
-
-    def blpop(self, keys: list[str], timeout: int = 0) -> tuple[bytes, bytes] | None:
-        if timeout <= 0:
-            raise ValueError(
-                "PostgresCacheBackend.blpop requires timeout > 0. "
-                "timeout=0 would block the calling thread indefinitely "
-                "with no way to interrupt short of process termination."
-            )
-        from onyx.db.engine.sql_engine import get_session_with_tenant
-
-        deadline = time.monotonic() + timeout
-        while True:
-            for key in keys:
-                lower = f"{_LIST_KEY_PREFIX}{key}:"
-                upper = f"{_LIST_KEY_PREFIX}{key}{_LIST_KEY_RANGE_TERMINATOR}"
-                stmt = (
-                    select(CacheStore)
-                    .where(
-                        CacheStore.key >= lower,
-                        CacheStore.key < upper,
-                        or_(
-                            CacheStore.expires_at.is_(None),
-                            CacheStore.expires_at > func.now(),
-                        ),
-                    )
-                    .order_by(CacheStore.key)
-                    .limit(1)
-                    .with_for_update(skip_locked=True)
-                )
-                with get_session_with_tenant(tenant_id=self._tenant_id) as session:
-                    row = session.execute(stmt).scalars().first()
-                    if row is not None:
-                        value = bytes(row.value) if row.value else b""
-                        session.delete(row)
-                        session.commit()
-                        return (key.encode(), value)
-            if time.monotonic() >= deadline:
-                return None
-            time.sleep(_BLPOP_POLL_INTERVAL)
-
-    # -- helpers -----------------------------------------------------------
-
-    def _lock_id_for(self, name: str) -> int:
-        """Map *name* to a 64-bit signed int for ``pg_advisory_lock``."""
-        h = hashlib.md5(f"{self._tenant_id}:{name}".encode()).digest()
-        return struct.unpack("q", h[:8])[0]
-
-
-# ------------------------------------------------------------------
-# Periodic cleanup
-# ------------------------------------------------------------------
-
-
-def cleanup_expired_cache_entries() -> None:
-    """Delete rows whose ``expires_at`` is in the past.
-
-    Called by the periodic poller every 5 minutes.
-    """
-    from onyx.db.engine.sql_engine import get_session_with_current_tenant
-
-    with get_session_with_current_tenant() as session:
-        session.execute(
-            delete(CacheStore).where(
-                CacheStore.expires_at.is_not(None),
-                CacheStore.expires_at < func.now(),
-            )
-        )
-        session.commit()

backend/onyx/cache/redis_backend.py

@@ -1,92 +0,0 @@
-from typing import cast
-
-from redis.client import Redis
-from redis.lock import Lock as RedisLock
-
-from onyx.cache.interface import CacheBackend
-from onyx.cache.interface import CacheLock
-
-
-class RedisCacheLock(CacheLock):
-    """Wraps ``redis.lock.Lock`` behind the ``CacheLock`` interface."""
-
-    def __init__(self, lock: RedisLock) -> None:
-        self._lock = lock
-
-    def acquire(
-        self,
-        blocking: bool = True,
-        blocking_timeout: float | None = None,
-    ) -> bool:
-        return bool(
-            self._lock.acquire(
-                blocking=blocking,
-                blocking_timeout=blocking_timeout,
-            )
-        )
-
-    def release(self) -> None:
-        self._lock.release()
-
-    def owned(self) -> bool:
-        return bool(self._lock.owned())
-
-
-class RedisCacheBackend(CacheBackend):
-    """``CacheBackend`` implementation that delegates to a ``redis.Redis`` client.
-
-    This is a thin pass-through — every method maps 1-to-1 to the underlying
-    Redis command.  ``TenantRedis`` key-prefixing is handled by the client
-    itself (provided by ``get_redis_client``).
-    """
-
-    def __init__(self, redis_client: Redis) -> None:
-        self._r = redis_client
-
-    # -- basic key/value ---------------------------------------------------
-
-    def get(self, key: str) -> bytes | None:
-        val = self._r.get(key)
-        if val is None:
-            return None
-        if isinstance(val, bytes):
-            return val
-        return str(val).encode()
-
-    def set(
-        self,
-        key: str,
-        value: str | bytes | int | float,
-        ex: int | None = None,
-    ) -> None:
-        self._r.set(key, value, ex=ex)
-
-    def delete(self, key: str) -> None:
-        self._r.delete(key)
-
-    def exists(self, key: str) -> bool:
-        return bool(self._r.exists(key))
-
-    # -- TTL ---------------------------------------------------------------
-
-    def expire(self, key: str, seconds: int) -> None:
-        self._r.expire(key, seconds)
-
-    def ttl(self, key: str) -> int:
-        return cast(int, self._r.ttl(key))
-
-    # -- distributed lock --------------------------------------------------
-
-    def lock(self, name: str, timeout: float | None = None) -> CacheLock:
-        return RedisCacheLock(self._r.lock(name, timeout=timeout))
-
-    # -- blocking list (MCP OAuth BLPOP pattern) ---------------------------
-
-    def rpush(self, key: str, value: str | bytes) -> None:
-        self._r.rpush(key, value)
-
-    def blpop(self, keys: list[str], timeout: int = 0) -> tuple[bytes, bytes] | None:
-        result = cast(list[bytes] | None, self._r.blpop(keys, timeout=timeout))
-        if result is None:
-            return None
-        return (result[0], result[1])

backend/onyx/chat/chat_processing_checker.py

@@ -1,52 +1,57 @@
 from uuid import UUID
 
-from onyx.cache.interface import CacheBackend
+from redis.client import Redis
 
+# Redis key prefixes for chat message processing
 PREFIX = "chatprocessing"
 FENCE_PREFIX = f"{PREFIX}_fence"
 FENCE_TTL = 30 * 60  # 30 minutes
 
 
 def _get_fence_key(chat_session_id: UUID) -> str:
-    """Generate the cache key for a chat session processing fence.
+    """
+    Generate the Redis key for a chat session processing a message.
 
     Args:
         chat_session_id: The UUID of the chat session
 
     Returns:
-        The fence key string. Tenant isolation is handled automatically
-        by the cache backend (Redis key-prefixing or Postgres schema routing).
+        The fence key string (tenant_id is automatically added by the Redis client)
     """
     return f"{FENCE_PREFIX}_{chat_session_id}"
 
 
 def set_processing_status(
-    chat_session_id: UUID, cache: CacheBackend, value: bool
+    chat_session_id: UUID, redis_client: Redis, value: bool
 ) -> None:
-    """Set or clear the fence for a chat session processing a message.
+    """
+    Set or clear the fence for a chat session processing a message.
 
-    If the key exists, a message is being processed.
+    If the key exists, we are processing a message. If the key does not exist, we are not processing a message.
 
     Args:
         chat_session_id: The UUID of the chat session
-        cache: Tenant-aware cache backend
+        redis_client: The Redis client to use
         value: True to set the fence, False to clear it
     """
     fence_key = _get_fence_key(chat_session_id)
+
     if value:
-        cache.set(fence_key, 0, ex=FENCE_TTL)
+        redis_client.set(fence_key, 0, ex=FENCE_TTL)
     else:
-        cache.delete(fence_key)
+        redis_client.delete(fence_key)
 
 
-def is_chat_session_processing(chat_session_id: UUID, cache: CacheBackend) -> bool:
-    """Check if the chat session is processing a message.
+def is_chat_session_processing(chat_session_id: UUID, redis_client: Redis) -> bool:
+    """
+    Check if the chat session is processing a message.
 
     Args:
         chat_session_id: The UUID of the chat session
-        cache: Tenant-aware cache backend
+        redis_client: The Redis client to use
 
     Returns:
         True if the chat session is processing a message, False otherwise
     """
-    return cache.exists(_get_fence_key(chat_session_id))
+    fence_key = _get_fence_key(chat_session_id)
+    return bool(redis_client.exists(fence_key))

backend/onyx/chat/llm_loop.py

@@ -1,7 +1,6 @@
 import json
 import time
 from collections.abc import Callable
-from typing import Any
 from typing import Literal
 
 from sqlalchemy.orm import Session
@@ -52,7 +51,6 @@ from onyx.tools.built_in_tools import STOPPING_TOOLS_NAMES
 from onyx.tools.interface import Tool
 from onyx.tools.models import ChatFile
 from onyx.tools.models import MemoryToolResponseSnapshot
-from onyx.tools.models import PythonToolRichResponse
 from onyx.tools.models import ToolCallInfo
 from onyx.tools.models import ToolCallKickoff
 from onyx.tools.models import ToolResponse
@@ -532,13 +530,11 @@ def _create_file_tool_metadata_message(
     """
     lines = [
         "You have access to the following files. Use the read_file tool to "
-        "read sections of any file. You MUST pass the file_id UUID (not the "
-        "filename) to read_file:"
+        "read sections of any file:"
     ]
     for meta in file_metadata:
         lines.append(
-            f'- file_id="{meta.file_id}" filename="{meta.filename}" '
-            f"(~{meta.approx_char_count:,} chars)"
+            f'- {meta.file_id}: "{meta.filename}" (~{meta.approx_char_count:,} chars)'
         )
 
     message_content = "\n".join(lines)
@@ -562,16 +558,12 @@ def _create_context_files_message(
     # Format as documents JSON as described in README
     documents_list = []
     for idx, file_text in enumerate(context_files.file_texts, start=1):
-        title = (
-            context_files.file_metadata[idx - 1].filename
-            if idx - 1 < len(context_files.file_metadata)
-            else None
+        documents_list.append(
+            {
+                "document": idx,
+                "contents": file_text,
+            }
         )
-        entry: dict[str, Any] = {"document": idx}
-        if title:
-            entry["title"] = title
-        entry["contents"] = file_text
-        documents_list.append(entry)
 
     documents_json = json.dumps({"documents": documents_list}, indent=2)
     message_content = f"Here are some documents provided for context, they may not all be relevant:\n{documents_json}"
@@ -967,13 +959,6 @@ def run_llm_loop(
                 ):
                     generated_images = tool_response.rich_response.generated_images
 
-                # Extract generated_files if this is a code interpreter response
-                generated_files = None
-                if isinstance(tool_response.rich_response, PythonToolRichResponse):
-                    generated_files = (
-                        tool_response.rich_response.generated_files or None
-                    )
-
                 # Persist memory if this is a memory tool response
                 memory_snapshot: MemoryToolResponseSnapshot | None = None
                 if isinstance(tool_response.rich_response, MemoryToolResponse):
@@ -1025,7 +1010,6 @@ def run_llm_loop(
                     tool_call_response=saved_response,
                     search_docs=displayed_docs or search_docs,
                     generated_images=generated_images,
-                    generated_files=generated_files,
                 )
                 # Add to state container for partial save support
                 state_container.add_tool_call(tool_call_info)

backend/onyx/chat/process_message.py

@@ -11,10 +11,9 @@ from contextvars import Token
 from uuid import UUID
 
 from pydantic import BaseModel
+from redis.client import Redis
 from sqlalchemy.orm import Session
 
-from onyx.cache.factory import get_cache_backend
-from onyx.cache.interface import CacheBackend
 from onyx.chat.chat_processing_checker import set_processing_status
 from onyx.chat.chat_state import ChatStateContainer
 from onyx.chat.chat_state import run_chat_loop_with_state_containers
@@ -80,6 +79,7 @@ from onyx.llm.request_context import reset_llm_mock_response
 from onyx.llm.request_context import set_llm_mock_response
 from onyx.llm.utils import litellm_exception_to_error_msg
 from onyx.onyxbot.slack.models import SlackContext
+from onyx.redis.redis_pool import get_redis_client
 from onyx.server.query_and_chat.models import AUTO_PLACE_AFTER_LATEST_MESSAGE
 from onyx.server.query_and_chat.models import MessageResponseIDInfo
 from onyx.server.query_and_chat.models import SendMessageRequest
@@ -448,7 +448,7 @@ def handle_stream_message_objects(
 
     llm: LLM | None = None
     chat_session: ChatSession | None = None
-    cache: CacheBackend | None = None
+    redis_client: Redis | None = None
 
     user_id = user.id
     if user.is_anonymous:
@@ -809,19 +809,19 @@ def handle_stream_message_objects(
             )
             simple_chat_history.insert(0, summary_simple)
 
-        cache = get_cache_backend()
+        redis_client = get_redis_client()
 
         reset_cancel_status(
             chat_session.id,
-            cache,
+            redis_client,
         )
 
         def check_is_connected() -> bool:
-            return check_stop_signal(chat_session.id, cache)
+            return check_stop_signal(chat_session.id, redis_client)
 
         set_processing_status(
             chat_session_id=chat_session.id,
-            cache=cache,
+            redis_client=redis_client,
             value=True,
         )
 
@@ -968,10 +968,10 @@ def handle_stream_message_objects(
             reset_llm_mock_response(mock_response_token)
 
         try:
-            if cache is not None and chat_session is not None:
+            if redis_client is not None and chat_session is not None:
                 set_processing_status(
                     chat_session_id=chat_session.id,
-                    cache=cache,
+                    redis_client=redis_client,
                     value=False,
                 )
         except Exception:

backend/onyx/chat/save_chat.py

@@ -1,5 +1,4 @@
 import json
-import mimetypes
 
 from sqlalchemy.orm import Session
 
@@ -13,41 +12,14 @@ from onyx.db.chat import create_db_search_doc
 from onyx.db.models import ChatMessage
 from onyx.db.models import ToolCall
 from onyx.db.tools import create_tool_call_no_commit
-from onyx.file_store.models import FileDescriptor
 from onyx.natural_language_processing.utils import BaseTokenizer
 from onyx.natural_language_processing.utils import get_tokenizer
-from onyx.server.query_and_chat.chat_utils import mime_type_to_chat_file_type
 from onyx.tools.models import ToolCallInfo
 from onyx.utils.logger import setup_logger
 
 logger = setup_logger()
 
 
-def _extract_referenced_file_descriptors(
-    tool_calls: list[ToolCallInfo],
-    message_text: str,
-) -> list[FileDescriptor]:
-    """Extract FileDescriptors for code interpreter files referenced in the message text."""
-    descriptors: list[FileDescriptor] = []
-    for tool_call_info in tool_calls:
-        if not tool_call_info.generated_files:
-            continue
-        for gen_file in tool_call_info.generated_files:
-            file_id = (
-                gen_file.file_link.rsplit("/", 1)[-1] if gen_file.file_link else ""
-            )
-            if file_id and file_id in message_text:
-                mime_type, _ = mimetypes.guess_type(gen_file.filename)
-                descriptors.append(
-                    FileDescriptor(
-                        id=file_id,
-                        type=mime_type_to_chat_file_type(mime_type),
-                        name=gen_file.filename,
-                    )
-                )
-    return descriptors
-
-
 def _create_and_link_tool_calls(
     tool_calls: list[ToolCallInfo],
     assistant_message: ChatMessage,
@@ -325,14 +297,5 @@ def save_chat_turn(
         citation_number_to_search_doc_id if citation_number_to_search_doc_id else None
     )
 
-    # 8. Attach code interpreter generated files that the assistant actually
-    # referenced in its response, so they are available via load_all_chat_files
-    # on subsequent turns. Files not mentioned are intermediate artifacts.
-    if message_text:
-        referenced = _extract_referenced_file_descriptors(tool_calls, message_text)
-        if referenced:
-            existing_files = assistant_message.files or []
-            assistant_message.files = existing_files + referenced
-
     # Finally save the messages, tool calls, and docs
     db_session.commit()

backend/onyx/chat/stop_signal_checker.py

@@ -1,58 +1,65 @@
 from uuid import UUID
 
-from onyx.cache.interface import CacheBackend
+from redis.client import Redis
 
+# Redis key prefixes for chat session stop signals
 PREFIX = "chatsessionstop"
 FENCE_PREFIX = f"{PREFIX}_fence"
-FENCE_TTL = 10 * 60  # 10 minutes
+FENCE_TTL = 10 * 60  # 10 minutes - defensive TTL to prevent memory leaks
 
 
 def _get_fence_key(chat_session_id: UUID) -> str:
-    """Generate the cache key for a chat session stop signal fence.
+    """
+    Generate the Redis key for a chat session stop signal fence.
 
     Args:
         chat_session_id: The UUID of the chat session
 
     Returns:
-        The fence key string. Tenant isolation is handled automatically
-        by the cache backend (Redis key-prefixing or Postgres schema routing).
+        The fence key string (tenant_id is automatically added by the Redis client)
     """
     return f"{FENCE_PREFIX}_{chat_session_id}"
 
 
-def set_fence(chat_session_id: UUID, cache: CacheBackend, value: bool) -> None:
-    """Set or clear the stop signal fence for a chat session.
+def set_fence(chat_session_id: UUID, redis_client: Redis, value: bool) -> None:
+    """
+    Set or clear the stop signal fence for a chat session.
 
     Args:
         chat_session_id: The UUID of the chat session
-        cache: Tenant-aware cache backend
+        redis_client: Redis client to use (tenant-aware client that auto-prefixes keys)
         value: True to set the fence (stop signal), False to clear it
     """
     fence_key = _get_fence_key(chat_session_id)
     if not value:
-        cache.delete(fence_key)
+        redis_client.delete(fence_key)
         return
-    cache.set(fence_key, 0, ex=FENCE_TTL)
+
+    redis_client.set(fence_key, 0, ex=FENCE_TTL)
 
 
-def is_connected(chat_session_id: UUID, cache: CacheBackend) -> bool:
-    """Check if the chat session should continue (not stopped).
+def is_connected(chat_session_id: UUID, redis_client: Redis) -> bool:
+    """
+    Check if the chat session should continue (not stopped).
 
     Args:
         chat_session_id: The UUID of the chat session to check
-        cache: Tenant-aware cache backend
+        redis_client: Redis client to use for checking the stop signal (tenant-aware client that auto-prefixes keys)
 
     Returns:
         True if the session should continue, False if it should stop
     """
-    return not cache.exists(_get_fence_key(chat_session_id))
+    fence_key = _get_fence_key(chat_session_id)
+    return not bool(redis_client.exists(fence_key))
 
 
-def reset_cancel_status(chat_session_id: UUID, cache: CacheBackend) -> None:
-    """Clear the stop signal for a chat session.
+def reset_cancel_status(chat_session_id: UUID, redis_client: Redis) -> None:
+    """
+    Clear the stop signal for a chat session.
 
     Args:
         chat_session_id: The UUID of the chat session
-        cache: Tenant-aware cache backend
+        redis_client: Redis client to use (tenant-aware client that auto-prefixes keys)
     """
-    cache.delete(_get_fence_key(chat_session_id))
+    fence_key = _get_fence_key(chat_session_id)
+    redis_client.delete(fence_key)

backend/onyx/configs/app_configs.py

@@ -6,7 +6,6 @@ from datetime import timezone
 from typing import cast
 
 from onyx.auth.schemas import AuthBackend
-from onyx.cache.interface import CacheBackendType
 from onyx.configs.constants import AuthType
 from onyx.configs.constants import QueryHistoryType
 from onyx.file_processing.enums import HtmlBasedConnectorTransformLinksStrategy
@@ -55,12 +54,6 @@ DISABLE_USER_KNOWLEDGE = os.environ.get("DISABLE_USER_KNOWLEDGE", "").lower() ==
 # are disabled but core chat, tools, user file uploads, and Projects still work.
 DISABLE_VECTOR_DB = os.environ.get("DISABLE_VECTOR_DB", "").lower() == "true"
 
-# Which backend to use for caching, locks, and ephemeral state.
-# "redis" (default) or "postgres" (only valid when DISABLE_VECTOR_DB=true).
-CACHE_BACKEND = CacheBackendType(
-    os.environ.get("CACHE_BACKEND", CacheBackendType.REDIS)
-)
-
 # Maximum token count for a single uploaded file. Files exceeding this are rejected.
 # Defaults to 100k tokens (or 10M when vector DB is disabled).
 _DEFAULT_FILE_TOKEN_LIMIT = 10_000_000 if DISABLE_VECTOR_DB else 100_000
@@ -495,7 +488,14 @@ CELERY_WORKER_PRIMARY_POOL_OVERFLOW = int(
     os.environ.get("CELERY_WORKER_PRIMARY_POOL_OVERFLOW") or 4
 )
 
-# Individual worker concurrency settings
+# Consolidated background worker (light, docprocessing, docfetching, heavy, monitoring, user_file_processing)
+# separate workers' defaults: light=24, docprocessing=6, docfetching=1, heavy=4, kg=2, monitoring=1, user_file=2
+# Total would be 40, but we use a more conservative default of 20 for the consolidated worker
+CELERY_WORKER_BACKGROUND_CONCURRENCY = int(
+    os.environ.get("CELERY_WORKER_BACKGROUND_CONCURRENCY") or 20
+)
+
+# Individual worker concurrency settings (used when USE_LIGHTWEIGHT_BACKGROUND_WORKER is False or on Kuberenetes deployments)
 CELERY_WORKER_HEAVY_CONCURRENCY = int(
     os.environ.get("CELERY_WORKER_HEAVY_CONCURRENCY") or 4
 )
@@ -812,9 +812,7 @@ RERANK_COUNT = int(os.environ.get("RERANK_COUNT") or 1000)
 # Tool Configs
 #####
 # Code Interpreter Service Configuration
-CODE_INTERPRETER_BASE_URL = os.environ.get(
-    "CODE_INTERPRETER_BASE_URL", "http://localhost:8000"
-)
+CODE_INTERPRETER_BASE_URL = os.environ.get("CODE_INTERPRETER_BASE_URL")
 
 CODE_INTERPRETER_DEFAULT_TIMEOUT_MS = int(
     os.environ.get("CODE_INTERPRETER_DEFAULT_TIMEOUT_MS") or 60_000
@@ -895,9 +893,6 @@ CUSTOM_ANSWER_VALIDITY_CONDITIONS = json.loads(
 )
 
 VESPA_REQUEST_TIMEOUT = int(os.environ.get("VESPA_REQUEST_TIMEOUT") or "15")
-VESPA_MIGRATION_REQUEST_TIMEOUT_S = int(
-    os.environ.get("VESPA_MIGRATION_REQUEST_TIMEOUT_S") or "120"
-)
 
 SYSTEM_RECURSION_LIMIT = int(os.environ.get("SYSTEM_RECURSION_LIMIT") or "1000")
 

backend/onyx/configs/constants.py

@@ -84,6 +84,7 @@ POSTGRES_CELERY_WORKER_LIGHT_APP_NAME = "celery_worker_light"
 POSTGRES_CELERY_WORKER_DOCPROCESSING_APP_NAME = "celery_worker_docprocessing"
 POSTGRES_CELERY_WORKER_DOCFETCHING_APP_NAME = "celery_worker_docfetching"
 POSTGRES_CELERY_WORKER_INDEXING_CHILD_APP_NAME = "celery_worker_indexing_child"
+POSTGRES_CELERY_WORKER_BACKGROUND_APP_NAME = "celery_worker_background"
 POSTGRES_CELERY_WORKER_HEAVY_APP_NAME = "celery_worker_heavy"
 POSTGRES_CELERY_WORKER_MONITORING_APP_NAME = "celery_worker_monitoring"
 POSTGRES_CELERY_WORKER_USER_FILE_PROCESSING_APP_NAME = (

backend/onyx/db/chat.py

@@ -98,7 +98,6 @@ def get_chat_sessions_by_user(
     db_session: Session,
     include_onyxbot_flows: bool = False,
     limit: int = 50,
-    before: datetime | None = None,
     project_id: int | None = None,
     only_non_project_chats: bool = False,
     include_failed_chats: bool = False,
@@ -113,9 +112,6 @@ def get_chat_sessions_by_user(
     if deleted is not None:
         stmt = stmt.where(ChatSession.deleted == deleted)
 
-    if before is not None:
-        stmt = stmt.where(ChatSession.time_updated < before)
-
     if limit:
         stmt = stmt.limit(limit)
 

backend/onyx/db/enums.py

@@ -186,7 +186,6 @@ class EmbeddingPrecision(str, PyEnum):
 
 class UserFileStatus(str, PyEnum):
     PROCESSING = "PROCESSING"
-    INDEXING = "INDEXING"
     COMPLETED = "COMPLETED"
     FAILED = "FAILED"
     CANCELED = "CANCELED"

backend/onyx/db/image_generation.py

@@ -202,6 +202,7 @@ def create_default_image_gen_config_from_api_key(
             api_key=api_key,
             api_base=None,
             api_version=None,
+            default_model_name=model_name,
             deployment_name=None,
             is_public=True,
         )

backend/onyx/db/llm.py

@@ -109,38 +109,45 @@ def can_user_access_llm_provider(
         is_admin: If True, bypass user group restrictions but still respect persona restrictions
 
     Access logic:
-    - is_public controls USER access (group bypass): when True, all users can access
-      regardless of group membership. When False, user must be in a whitelisted group
-      (or be admin).
-    - Persona restrictions are ALWAYS enforced when set, regardless of is_public.
-      This allows admins to make a provider available to all users while still
-      restricting which personas (assistants) can use it.
-
-    Decision matrix:
-    1. is_public=True, no personas set → everyone has access
-    2. is_public=True, personas set → all users, but only whitelisted personas
-    3. is_public=False, groups+personas set → must satisfy BOTH (admins bypass groups)
-    4. is_public=False, only groups set → must be in group (admins bypass)
-    5. is_public=False, only personas set → must use whitelisted persona
-    6. is_public=False, neither set → admin-only (locked)
+    1. If is_public=True → everyone has access (public override)
+    2. If is_public=False:
+       - Both groups AND personas set → must satisfy BOTH (AND logic, admins bypass group check)
+       - Only groups set → must be in one of the groups (OR across groups, admins bypass)
+       - Only personas set → must use one of the personas (OR across personas, applies to admins)
+       - Neither set → NOBODY has access unless admin (locked, admin-only)
     """
-    provider_group_ids = {g.id for g in (provider.groups or [])}
-    provider_persona_ids = {p.id for p in (provider.personas or [])}
-    has_groups = bool(provider_group_ids)
-    has_personas = bool(provider_persona_ids)
-
-    # Persona restrictions are always enforced when set, regardless of is_public
-    if has_personas and not (persona and persona.id in provider_persona_ids):
-        return False
-
+    # Public override - everyone has access
     if provider.is_public:
         return True
 
+    # Extract IDs once to avoid multiple iterations
+    provider_group_ids = (
+        {group.id for group in provider.groups} if provider.groups else set()
+    )
+    provider_persona_ids = (
+        {p.id for p in provider.personas} if provider.personas else set()
+    )
+
+    has_groups = bool(provider_group_ids)
+    has_personas = bool(provider_persona_ids)
+
+    # Both groups AND personas set → AND logic (must satisfy both)
+    if has_groups and has_personas:
+        # Admins bypass group check but still must satisfy persona restrictions
+        user_in_group = is_admin or bool(user_group_ids & provider_group_ids)
+        persona_allowed = persona.id in provider_persona_ids if persona else False
+        return user_in_group and persona_allowed
+
+    # Only groups set → user must be in one of the groups (admins bypass)
     if has_groups:
         return is_admin or bool(user_group_ids & provider_group_ids)
 
-    # No groups: either persona-whitelisted (already passed) or admin-only if locked
-    return has_personas or is_admin
+    # Only personas set → persona must be in allowed list (applies to admins too)
+    if has_personas:
+        return persona.id in provider_persona_ids if persona else False
+
+    # Neither groups nor personas set, and not public → admins can access
+    return is_admin
 
 
 def validate_persona_ids_exist(
@@ -206,29 +213,11 @@ def upsert_llm_provider(
     llm_provider_upsert_request: LLMProviderUpsertRequest,
     db_session: Session,
 ) -> LLMProviderView:
-    existing_llm_provider: LLMProviderModel | None = None
-    if llm_provider_upsert_request.id:
-        existing_llm_provider = fetch_existing_llm_provider_by_id(
-            id=llm_provider_upsert_request.id, db_session=db_session
-        )
-        if not existing_llm_provider:
-            raise ValueError(
-                f"LLM provider with id {llm_provider_upsert_request.id} not found"
-            )
+    existing_llm_provider = fetch_existing_llm_provider(
+        name=llm_provider_upsert_request.name, db_session=db_session
+    )
 
-        if existing_llm_provider.name != llm_provider_upsert_request.name:
-            raise ValueError(
-                f"LLM provider with id {llm_provider_upsert_request.id} name change not allowed"
-            )
-    else:
-        existing_llm_provider = fetch_existing_llm_provider(
-            name=llm_provider_upsert_request.name, db_session=db_session
-        )
-        if existing_llm_provider:
-            raise ValueError(
-                f"LLM provider with name '{llm_provider_upsert_request.name}'"
-                " already exists"
-            )
+    if not existing_llm_provider:
         existing_llm_provider = LLMProviderModel(name=llm_provider_upsert_request.name)
         db_session.add(existing_llm_provider)
 
@@ -249,7 +238,11 @@ def upsert_llm_provider(
     existing_llm_provider.api_base = api_base
     existing_llm_provider.api_version = llm_provider_upsert_request.api_version
     existing_llm_provider.custom_config = custom_config
-
+    # TODO: Remove default model name on api change
+    # Needed due to /provider/{id}/default endpoint not disclosing the default model name
+    existing_llm_provider.default_model_name = (
+        llm_provider_upsert_request.default_model_name
+    )
     existing_llm_provider.is_public = llm_provider_upsert_request.is_public
     existing_llm_provider.is_auto_mode = llm_provider_upsert_request.is_auto_mode
     existing_llm_provider.deployment_name = llm_provider_upsert_request.deployment_name
@@ -313,6 +306,15 @@ def upsert_llm_provider(
                 display_name=model_config.display_name,
             )
 
+    default_model = fetch_default_model(db_session, LLMModelFlowType.CHAT)
+    if default_model and default_model.llm_provider_id == existing_llm_provider.id:
+        _update_default_model(
+            db_session=db_session,
+            provider_id=existing_llm_provider.id,
+            model=existing_llm_provider.default_model_name,
+            flow_type=LLMModelFlowType.CHAT,
+        )
+
     # Make sure the relationship table stays up to date
     update_group_llm_provider_relationships__no_commit(
         llm_provider_id=existing_llm_provider.id,
@@ -486,22 +488,6 @@ def fetch_existing_llm_provider(
     return provider_model
 
 
-def fetch_existing_llm_provider_by_id(
-    id: int, db_session: Session
-) -> LLMProviderModel | None:
-    provider_model = db_session.scalar(
-        select(LLMProviderModel)
-        .where(LLMProviderModel.id == id)
-        .options(
-            selectinload(LLMProviderModel.model_configurations),
-            selectinload(LLMProviderModel.groups),
-            selectinload(LLMProviderModel.personas),
-        )
-    )
-
-    return provider_model
-
-
 def fetch_embedding_provider(
     db_session: Session, provider_type: EmbeddingProvider
 ) -> CloudEmbeddingProviderModel | None:
@@ -532,7 +518,6 @@ def fetch_default_model(
 ) -> ModelConfiguration | None:
     model_config = db_session.scalar(
         select(ModelConfiguration)
-        .options(selectinload(ModelConfiguration.llm_provider))
         .join(LLMModelFlow)
         .where(
             ModelConfiguration.is_visible == True,  # noqa: E712
@@ -619,13 +604,22 @@ def remove_llm_provider__no_commit(db_session: Session, provider_id: int) -> Non
     db_session.flush()
 
 
-def update_default_provider(
-    provider_id: int, model_name: str, db_session: Session
-) -> None:
+def update_default_provider(provider_id: int, db_session: Session) -> None:
+    # Attempt to get the default_model_name from the provider first
+    # TODO: Remove default_model_name check
+    provider = db_session.scalar(
+        select(LLMProviderModel).where(
+            LLMProviderModel.id == provider_id,
+        )
+    )
+
+    if provider is None:
+        raise ValueError(f"LLM Provider with id={provider_id} does not exist")
+
     _update_default_model(
         db_session,
         provider_id,
-        model_name,
+        provider.default_model_name,  # type: ignore[arg-type]
         LLMModelFlowType.CHAT,
     )
 
@@ -811,6 +805,12 @@ def sync_auto_mode_models(
             )
             changes += 1
 
+    # In Auto mode, default model is always set from GitHub config
+    default_model = llm_recommendations.get_default_model(provider.provider)
+    if default_model and provider.default_model_name != default_model.name:
+        provider.default_model_name = default_model.name
+        changes += 1
+
     db_session.commit()
     return changes
 

backend/onyx/db/models.py

@@ -103,6 +103,7 @@ from onyx.utils.encryption import encrypt_string_to_bytes
 from onyx.utils.sensitive import SensitiveValue
 from onyx.utils.headers import HeaderItemDict
 from shared_configs.enums import EmbeddingProvider
+from onyx.context.search.enums import RecencyBiasSetting
 
 # TODO: After anonymous user migration has been deployed, make user_id columns NOT NULL
 # and update Mapped[User | None] relationships to Mapped[User] where needed.
@@ -3264,6 +3265,19 @@ class Persona(Base):
     )
     name: Mapped[str] = mapped_column(String)
     description: Mapped[str] = mapped_column(String)
+    # Number of chunks to pass to the LLM for generation.
+    num_chunks: Mapped[float | None] = mapped_column(Float, nullable=True)
+    chunks_above: Mapped[int] = mapped_column(Integer)
+    chunks_below: Mapped[int] = mapped_column(Integer)
+    # Pass every chunk through LLM for evaluation, fairly expensive
+    # Can be turned off globally by admin, in which case, this setting is ignored
+    llm_relevance_filter: Mapped[bool] = mapped_column(Boolean)
+    # Enables using LLM to extract time and source type filters
+    # Can also be admin disabled globally
+    llm_filter_extraction: Mapped[bool] = mapped_column(Boolean)
+    recency_bias: Mapped[RecencyBiasSetting] = mapped_column(
+        Enum(RecencyBiasSetting, native_enum=False)
+    )
 
     # Allows the persona to specify a specific default LLM model
     # NOTE: only is applied on the actual response generation - is not used for things like
@@ -3290,8 +3304,11 @@ class Persona(Base):
     # Treated specially (cannot be user edited etc.)
     builtin_persona: Mapped[bool] = mapped_column(Boolean, default=False)
 
-    # Featured personas are highlighted in the UI
-    featured: Mapped[bool] = mapped_column(Boolean, default=False)
+    # Default personas are personas created by admins and are automatically added
+    # to all users' assistants list.
+    is_default_persona: Mapped[bool] = mapped_column(
+        Boolean, default=False, nullable=False
+    )
     # controls whether the persona is available to be selected by users
     is_visible: Mapped[bool] = mapped_column(Boolean, default=True)
     # controls the ordering of personas in the UI
@@ -4926,9 +4943,7 @@ class ScimUserMapping(Base):
     __tablename__ = "scim_user_mapping"
 
     id: Mapped[int] = mapped_column(Integer, primary_key=True)
-    external_id: Mapped[str | None] = mapped_column(
-        String, unique=True, index=True, nullable=True
-    )
+    external_id: Mapped[str] = mapped_column(String, unique=True, index=True)
     user_id: Mapped[UUID] = mapped_column(
         ForeignKey("user.id", ondelete="CASCADE"), unique=True, nullable=False
     )
@@ -4985,25 +5000,3 @@ class CodeInterpreterServer(Base):
 
     id: Mapped[int] = mapped_column(Integer, primary_key=True)
     server_enabled: Mapped[bool] = mapped_column(Boolean, nullable=False, default=True)
-
-
-class CacheStore(Base):
-    """Key-value cache table used by ``PostgresCacheBackend``.
-
-    Replaces Redis for simple KV caching, locks, and list operations
-    when ``CACHE_BACKEND=postgres`` (NO_VECTOR_DB deployments).
-
-    Intentionally separate from ``KVStore``:
-    - Stores raw bytes (LargeBinary) vs JSONB, matching Redis semantics.
-    - Has ``expires_at`` for TTL; rows are periodically garbage-collected.
-    - Holds ephemeral data (tokens, stop signals, lock state) not
-      persistent application config, so cleanup can be aggressive.
-    """
-
-    __tablename__ = "cache_store"
-
-    key: Mapped[str] = mapped_column(String, primary_key=True)
-    value: Mapped[bytes | None] = mapped_column(LargeBinary, nullable=True)
-    expires_at: Mapped[datetime.datetime | None] = mapped_column(
-        DateTime(timezone=True), nullable=True
-    )

backend/onyx/db/persona.py

@@ -18,8 +18,11 @@ from sqlalchemy.orm import Session
 from onyx.access.hierarchy_access import get_user_external_group_ids
 from onyx.auth.schemas import UserRole
 from onyx.configs.app_configs import CURATORS_CANNOT_VIEW_OR_EDIT_NON_OWNED_ASSISTANTS
+from onyx.configs.chat_configs import CONTEXT_CHUNKS_ABOVE
+from onyx.configs.chat_configs import CONTEXT_CHUNKS_BELOW
 from onyx.configs.constants import DEFAULT_PERSONA_ID
 from onyx.configs.constants import NotificationType
+from onyx.context.search.enums import RecencyBiasSetting
 from onyx.db.constants import SLACK_BOT_PERSONA_PREFIX
 from onyx.db.document_access import get_accessible_documents_by_ids
 from onyx.db.models import ConnectorCredentialPair
@@ -251,15 +254,13 @@ def create_update_persona(
     # Permission to actually use these is checked later
 
     try:
-        # Featured persona validation
-        if create_persona_request.featured:
-
-            # Curators can edit featured personas, but not make them
-            # TODO this will be reworked soon with RBAC permissions feature
+        # Default persona validation
+        if create_persona_request.is_default_persona:
+            # Curators can edit default personas, but not make them
             if user.role == UserRole.CURATOR or user.role == UserRole.GLOBAL_CURATOR:
                 pass
             elif user.role != UserRole.ADMIN:
-                raise ValueError("Only admins can make a featured persona")
+                raise ValueError("Only admins can make a default persona")
 
         # Convert incoming string UUIDs to UUID objects for DB operations
         converted_user_file_ids = None
@@ -280,6 +281,7 @@ def create_update_persona(
             document_set_ids=create_persona_request.document_set_ids,
             tool_ids=create_persona_request.tool_ids,
             is_public=create_persona_request.is_public,
+            recency_bias=create_persona_request.recency_bias,
             llm_model_provider_override=create_persona_request.llm_model_provider_override,
             llm_model_version_override=create_persona_request.llm_model_version_override,
             starter_messages=create_persona_request.starter_messages,
@@ -293,7 +295,10 @@ def create_update_persona(
             remove_image=create_persona_request.remove_image,
             search_start_date=create_persona_request.search_start_date,
             label_ids=create_persona_request.label_ids,
-            featured=create_persona_request.featured,
+            num_chunks=create_persona_request.num_chunks,
+            llm_relevance_filter=create_persona_request.llm_relevance_filter,
+            llm_filter_extraction=create_persona_request.llm_filter_extraction,
+            is_default_persona=create_persona_request.is_default_persona,
             user_file_ids=converted_user_file_ids,
             commit=False,
             hierarchy_node_ids=create_persona_request.hierarchy_node_ids,
@@ -869,6 +874,10 @@ def upsert_persona(
     user: User | None,
     name: str,
     description: str,
+    num_chunks: float,
+    llm_relevance_filter: bool,
+    llm_filter_extraction: bool,
+    recency_bias: RecencyBiasSetting,
     llm_model_provider_override: str | None,
     llm_model_version_override: str | None,
     starter_messages: list[StarterMessage] | None,
@@ -889,11 +898,13 @@ def upsert_persona(
     remove_image: bool | None = None,
     search_start_date: datetime | None = None,
     builtin_persona: bool = False,
-    featured: bool | None = None,
+    is_default_persona: bool | None = None,
     label_ids: list[int] | None = None,
     user_file_ids: list[UUID] | None = None,
     hierarchy_node_ids: list[int] | None = None,
     document_ids: list[str] | None = None,
+    chunks_above: int = CONTEXT_CHUNKS_ABOVE,
+    chunks_below: int = CONTEXT_CHUNKS_BELOW,
     replace_base_system_prompt: bool = False,
 ) -> Persona:
     """
@@ -1004,6 +1015,12 @@ def upsert_persona(
         # `default` and `built-in` properties can only be set when creating a persona.
         existing_persona.name = name
         existing_persona.description = description
+        existing_persona.num_chunks = num_chunks
+        existing_persona.chunks_above = chunks_above
+        existing_persona.chunks_below = chunks_below
+        existing_persona.llm_relevance_filter = llm_relevance_filter
+        existing_persona.llm_filter_extraction = llm_filter_extraction
+        existing_persona.recency_bias = recency_bias
         existing_persona.llm_model_provider_override = llm_model_provider_override
         existing_persona.llm_model_version_override = llm_model_version_override
         existing_persona.starter_messages = starter_messages
@@ -1017,8 +1034,10 @@ def upsert_persona(
         if label_ids is not None:
             existing_persona.labels.clear()
             existing_persona.labels = labels or []
-        existing_persona.featured = (
-            featured if featured is not None else existing_persona.featured
+        existing_persona.is_default_persona = (
+            is_default_persona
+            if is_default_persona is not None
+            else existing_persona.is_default_persona
         )
         # Update embedded prompt fields if provided
         if system_prompt is not None:
@@ -1071,6 +1090,12 @@ def upsert_persona(
             is_public=is_public,
             name=name,
             description=description,
+            num_chunks=num_chunks,
+            chunks_above=chunks_above,
+            chunks_below=chunks_below,
+            llm_relevance_filter=llm_relevance_filter,
+            llm_filter_extraction=llm_filter_extraction,
+            recency_bias=recency_bias,
             builtin_persona=builtin_persona,
             system_prompt=system_prompt or "",
             task_prompt=task_prompt or "",
@@ -1086,7 +1111,9 @@ def upsert_persona(
             display_priority=display_priority,
             is_visible=is_visible,
             search_start_date=search_start_date,
-            featured=(featured if featured is not None else False),
+            is_default_persona=(
+                is_default_persona if is_default_persona is not None else False
+            ),
             user_files=user_files or [],
             labels=labels or [],
             hierarchy_nodes=hierarchy_nodes or [],
@@ -1131,9 +1158,9 @@ def delete_old_default_personas(
     db_session.commit()
 
 
-def update_persona_featured(
+def update_persona_is_default(
     persona_id: int,
-    featured: bool,
+    is_default: bool,
     db_session: Session,
     user: User,
 ) -> None:
@@ -1141,7 +1168,7 @@ def update_persona_featured(
         db_session=db_session, persona_id=persona_id, user=user, get_editable=True
     )
 
-    persona.featured = featured
+    persona.is_default_persona = is_default
     db_session.commit()
 
 

backend/onyx/db/projects.py

@@ -9,9 +9,8 @@ from pydantic import BaseModel
 from pydantic import ConfigDict
 from sqlalchemy import func
 from sqlalchemy.orm import Session
-from starlette.background import BackgroundTasks
 
-from onyx.configs.app_configs import DISABLE_VECTOR_DB
+from onyx.background.celery.versioned_apps.client import app as client_app
 from onyx.configs.constants import FileOrigin
 from onyx.configs.constants import OnyxCeleryPriority
 from onyx.configs.constants import OnyxCeleryQueues
@@ -52,7 +51,7 @@ def create_user_files(
 ) -> CategorizedFilesResult:
 
     # Categorize the files
-    categorized_files = categorize_uploaded_files(files, db_session)
+    categorized_files = categorize_uploaded_files(files)
     # NOTE: At the moment, zip metadata is not used for user files.
     # Should revisit to decide whether this should be a feature.
     upload_response = upload_files(categorized_files.acceptable, FileOrigin.USER_FILE)
@@ -106,8 +105,8 @@ def upload_files_to_user_files_with_indexing(
     user: User,
     temp_id_map: dict[str, str] | None,
     db_session: Session,
-    background_tasks: BackgroundTasks | None = None,
 ) -> CategorizedFilesResult:
+    # Validate project ownership if a project_id is provided
     if project_id is not None and user is not None:
         if not check_project_ownership(project_id, user.id, db_session):
             raise HTTPException(status_code=404, detail="Project not found")
@@ -128,27 +127,16 @@ def upload_files_to_user_files_with_indexing(
         logger.warning(
             f"File {rejected_file.filename} rejected for {rejected_file.reason}"
         )
-
-    if DISABLE_VECTOR_DB and background_tasks is not None:
-        from onyx.background.task_utils import drain_processing_loop
-
-        background_tasks.add_task(drain_processing_loop, tenant_id)
-        for user_file in user_files:
-            logger.info(f"Queued in-process processing for user_file_id={user_file.id}")
-    else:
-        from onyx.background.celery.versioned_apps.client import app as client_app
-
-        for user_file in user_files:
-            task = client_app.send_task(
-                OnyxCeleryTask.PROCESS_SINGLE_USER_FILE,
-                kwargs={"user_file_id": user_file.id, "tenant_id": tenant_id},
-                queue=OnyxCeleryQueues.USER_FILE_PROCESSING,
-                priority=OnyxCeleryPriority.HIGH,
-            )
-            logger.info(
-                f"Triggered indexing for user_file_id={user_file.id} "
-                f"with task_id={task.id}"
-            )
+    for user_file in user_files:
+        task = client_app.send_task(
+            OnyxCeleryTask.PROCESS_SINGLE_USER_FILE,
+            kwargs={"user_file_id": user_file.id, "tenant_id": tenant_id},
+            queue=OnyxCeleryQueues.USER_FILE_PROCESSING,
+            priority=OnyxCeleryPriority.HIGH,
+        )
+        logger.info(
+            f"Triggered indexing for user_file_id={user_file.id} with task_id={task.id}"
+        )
 
     return CategorizedFilesResult(
         user_files=user_files,

backend/onyx/db/slack_channel_config.py

@@ -5,6 +5,8 @@ from sqlalchemy import select
 from sqlalchemy.orm import joinedload
 from sqlalchemy.orm import Session
 
+from onyx.configs.chat_configs import MAX_CHUNKS_FED_TO_CHAT
+from onyx.context.search.enums import RecencyBiasSetting
 from onyx.db.constants import DEFAULT_PERSONA_SLACK_CHANNEL_NAME
 from onyx.db.constants import SLACK_BOT_PERSONA_PREFIX
 from onyx.db.models import ChannelConfig
@@ -43,6 +45,8 @@ def create_slack_channel_persona(
     channel_name: str | None,
     document_set_ids: list[int],
     existing_persona_id: int | None = None,
+    num_chunks: float = MAX_CHUNKS_FED_TO_CHAT,
+    enable_auto_filters: bool = False,
 ) -> Persona:
     """NOTE: does not commit changes"""
 
@@ -69,13 +73,17 @@ def create_slack_channel_persona(
         system_prompt="",
         task_prompt="",
         datetime_aware=True,
+        num_chunks=num_chunks,
+        llm_relevance_filter=True,
+        llm_filter_extraction=enable_auto_filters,
+        recency_bias=RecencyBiasSetting.AUTO,
         tool_ids=[search_tool.id],
         document_set_ids=document_set_ids,
         llm_model_provider_override=None,
         llm_model_version_override=None,
         starter_messages=None,
         is_public=True,
-        featured=False,
+        is_default_persona=False,
         db_session=db_session,
         commit=False,
     )

backend/onyx/document_index/opensearch/opensearch_document_index.py

@@ -6,6 +6,7 @@ import httpx
 from opensearchpy import NotFoundError
 
 from onyx.access.models import DocumentAccess
+from onyx.configs.app_configs import USING_AWS_MANAGED_OPENSEARCH
 from onyx.configs.app_configs import VERIFY_CREATE_OPENSEARCH_INDEX_ON_INIT_MT
 from onyx.configs.chat_configs import NUM_RETURNED_HITS
 from onyx.configs.chat_configs import TITLE_CONTENT_RATIO
@@ -562,7 +563,12 @@ class OpenSearchDocumentIndex(DocumentIndex):
         )
 
         if not self._client.index_exists():
-            index_settings = DocumentSchema.get_index_settings_based_on_environment()
+            if USING_AWS_MANAGED_OPENSEARCH:
+                index_settings = (
+                    DocumentSchema.get_index_settings_for_aws_managed_opensearch()
+                )
+            else:
+                index_settings = DocumentSchema.get_index_settings()
             self._client.create_index(
                 mappings=expected_mappings,
                 settings=index_settings,

backend/onyx/document_index/opensearch/schema.py

@@ -12,7 +12,6 @@ from pydantic import model_validator
 from pydantic import SerializerFunctionWrapHandler
 
 from onyx.configs.app_configs import OPENSEARCH_TEXT_ANALYZER
-from onyx.configs.app_configs import USING_AWS_MANAGED_OPENSEARCH
 from onyx.document_index.interfaces_new import TenantState
 from onyx.document_index.opensearch.constants import DEFAULT_MAX_CHUNK_SIZE
 from onyx.document_index.opensearch.constants import EF_CONSTRUCTION
@@ -526,7 +525,7 @@ class DocumentSchema:
         }
 
     @staticmethod
-    def get_index_settings_for_aws_managed_opensearch_st_dev() -> dict[str, Any]:
+    def get_index_settings_for_aws_managed_opensearch() -> dict[str, Any]:
         """
         Settings for AWS-managed OpenSearch.
 
@@ -547,41 +546,3 @@ class DocumentSchema:
                 "knn.algo_param.ef_search": EF_SEARCH,
             }
         }
-
-    @staticmethod
-    def get_index_settings_for_aws_managed_opensearch_mt_cloud() -> dict[str, Any]:
-        """
-        Settings for AWS-managed OpenSearch in multi-tenant cloud.
-
-        324 shards very roughly targets a storage load of ~30Gb per shard, which
-        according to AWS OpenSearch documentation is within a good target range.
-
-        As documented above we need 2 replicas for a total of 3 copies of the
-        data because the cluster is configured with 3-AZ awareness.
-        """
-        return {
-            "index": {
-                "number_of_shards": 324,
-                "number_of_replicas": 2,
-                # Required for vector search.
-                "knn": True,
-                "knn.algo_param.ef_search": EF_SEARCH,
-            }
-        }
-
-    @staticmethod
-    def get_index_settings_based_on_environment() -> dict[str, Any]:
-        """
-        Returns the index settings based on the environment.
-        """
-        if USING_AWS_MANAGED_OPENSEARCH:
-            if MULTI_TENANT:
-                return (
-                    DocumentSchema.get_index_settings_for_aws_managed_opensearch_mt_cloud()
-                )
-            else:
-                return (
-                    DocumentSchema.get_index_settings_for_aws_managed_opensearch_st_dev()
-                )
-        else:
-            return DocumentSchema.get_index_settings()

backend/onyx/document_index/vespa/chunk_retrieval.py

@@ -1,6 +1,5 @@
 import json
 import string
-import time
 from collections.abc import Callable
 from collections.abc import Mapping
 from datetime import datetime
@@ -19,7 +18,6 @@ from onyx.background.celery.tasks.opensearch_migration.transformer import (
 )
 from onyx.configs.app_configs import LOG_VESPA_TIMING_INFORMATION
 from onyx.configs.app_configs import VESPA_LANGUAGE_OVERRIDE
-from onyx.configs.app_configs import VESPA_MIGRATION_REQUEST_TIMEOUT_S
 from onyx.context.search.models import IndexFilters
 from onyx.context.search.models import InferenceChunkUncleaned
 from onyx.document_index.interfaces import VespaChunkRequest
@@ -340,18 +338,12 @@ def get_all_chunks_paginated(
             params["continuation"] = continuation_token
 
         response: httpx.Response | None = None
-        start_time = time.monotonic()
         try:
-            with get_vespa_http_client(
-                timeout=VESPA_MIGRATION_REQUEST_TIMEOUT_S
-            ) as http_client:
+            with get_vespa_http_client() as http_client:
                 response = http_client.get(url, params=params)
                 response.raise_for_status()
         except httpx.HTTPError as e:
-            error_base = (
-                f"Failed to get chunks from Vespa slice {slice_id} with continuation token "
-                f"{continuation_token} in {time.monotonic() - start_time:.3f} seconds."
-            )
+            error_base = f"Failed to get chunks from Vespa slice {slice_id} with continuation token {continuation_token}."
             logger.exception(
                 f"Request URL: {e.request.url}\n"
                 f"Request Headers: {e.request.headers}\n"

backend/onyx/document_index/vespa/shared_utils/utils.py

@@ -52,9 +52,7 @@ def replace_invalid_doc_id_characters(text: str) -> str:
     return text.replace("'", "_")
 
 
-def get_vespa_http_client(
-    no_timeout: bool = False, http2: bool = True, timeout: int | None = None
-) -> httpx.Client:
+def get_vespa_http_client(no_timeout: bool = False, http2: bool = True) -> httpx.Client:
     """
     Configures and returns an HTTP client for communicating with Vespa,
     including authentication if needed.
@@ -66,7 +64,7 @@ def get_vespa_http_client(
             else None
         ),
         verify=False if not MANAGED_VESPA else True,
-        timeout=None if no_timeout else (timeout or VESPA_REQUEST_TIMEOUT),
+        timeout=None if no_timeout else VESPA_REQUEST_TIMEOUT,
         http2=http2,
     )
 

backend/onyx/error_handling/error_codes.py

@@ -1,101 +0,0 @@
-"""
-Standardized error codes for the Onyx backend.
-
-Usage:
-    from onyx.error_handling.error_codes import OnyxErrorCode
-    from onyx.error_handling.exceptions import OnyxError
-
-    raise OnyxError(OnyxErrorCode.UNAUTHENTICATED, "Token expired")
-"""
-
-from enum import Enum
-
-
-class OnyxErrorCode(Enum):
-    """
-    Each member is a tuple of (error_code_string, http_status_code).
-
-    The error_code_string is a stable, machine-readable identifier that
-    API consumers can match on. The http_status_code is the default HTTP
-    status to return.
-    """
-
-    # ------------------------------------------------------------------
-    # Authentication (401)
-    # ------------------------------------------------------------------
-    UNAUTHENTICATED = ("UNAUTHENTICATED", 401)
-    INVALID_TOKEN = ("INVALID_TOKEN", 401)
-    TOKEN_EXPIRED = ("TOKEN_EXPIRED", 401)
-    CSRF_FAILURE = ("CSRF_FAILURE", 403)
-
-    # ------------------------------------------------------------------
-    # Authorization (403)
-    # ------------------------------------------------------------------
-    UNAUTHORIZED = ("UNAUTHORIZED", 403)
-    INSUFFICIENT_PERMISSIONS = ("INSUFFICIENT_PERMISSIONS", 403)
-    ADMIN_ONLY = ("ADMIN_ONLY", 403)
-    EE_REQUIRED = ("EE_REQUIRED", 403)
-
-    # ------------------------------------------------------------------
-    # Validation / Bad Request (400)
-    # ------------------------------------------------------------------
-    VALIDATION_ERROR = ("VALIDATION_ERROR", 400)
-    INVALID_INPUT = ("INVALID_INPUT", 400)
-    MISSING_REQUIRED_FIELD = ("MISSING_REQUIRED_FIELD", 400)
-
-    # ------------------------------------------------------------------
-    # Not Found (404)
-    # ------------------------------------------------------------------
-    NOT_FOUND = ("NOT_FOUND", 404)
-    CONNECTOR_NOT_FOUND = ("CONNECTOR_NOT_FOUND", 404)
-    CREDENTIAL_NOT_FOUND = ("CREDENTIAL_NOT_FOUND", 404)
-    PERSONA_NOT_FOUND = ("PERSONA_NOT_FOUND", 404)
-    DOCUMENT_NOT_FOUND = ("DOCUMENT_NOT_FOUND", 404)
-    SESSION_NOT_FOUND = ("SESSION_NOT_FOUND", 404)
-    USER_NOT_FOUND = ("USER_NOT_FOUND", 404)
-
-    # ------------------------------------------------------------------
-    # Conflict (409)
-    # ------------------------------------------------------------------
-    CONFLICT = ("CONFLICT", 409)
-    DUPLICATE_RESOURCE = ("DUPLICATE_RESOURCE", 409)
-
-    # ------------------------------------------------------------------
-    # Rate Limiting / Quotas (429 / 402)
-    # ------------------------------------------------------------------
-    RATE_LIMITED = ("RATE_LIMITED", 429)
-    SEAT_LIMIT_EXCEEDED = ("SEAT_LIMIT_EXCEEDED", 402)
-
-    # ------------------------------------------------------------------
-    # Connector / Credential Errors (400-range)
-    # ------------------------------------------------------------------
-    CONNECTOR_VALIDATION_FAILED = ("CONNECTOR_VALIDATION_FAILED", 400)
-    CREDENTIAL_INVALID = ("CREDENTIAL_INVALID", 400)
-    CREDENTIAL_EXPIRED = ("CREDENTIAL_EXPIRED", 401)
-
-    # ------------------------------------------------------------------
-    # Server Errors (5xx)
-    # ------------------------------------------------------------------
-    INTERNAL_ERROR = ("INTERNAL_ERROR", 500)
-    NOT_IMPLEMENTED = ("NOT_IMPLEMENTED", 501)
-    SERVICE_UNAVAILABLE = ("SERVICE_UNAVAILABLE", 503)
-    BAD_GATEWAY = ("BAD_GATEWAY", 502)
-    LLM_PROVIDER_ERROR = ("LLM_PROVIDER_ERROR", 502)
-    GATEWAY_TIMEOUT = ("GATEWAY_TIMEOUT", 504)
-
-    def __init__(self, code: str, status_code: int) -> None:
-        self.code = code
-        self.status_code = status_code
-
-    def detail(self, message: str | None = None) -> dict[str, str]:
-        """Build a structured error detail dict.
-
-        Returns a dict like:
-            {"error_code": "UNAUTHENTICATED", "message": "Token expired"}
-
-        If no message is supplied, the error code itself is used as the message.
-        """
-        return {
-            "error_code": self.code,
-            "message": message or self.code,
-        }

backend/onyx/error_handling/exceptions.py

@@ -1,82 +0,0 @@
-"""OnyxError — the single exception type for all Onyx business errors.
-
-Raise ``OnyxError`` instead of ``HTTPException`` in business code.  A global
-FastAPI exception handler (registered via ``register_onyx_exception_handlers``)
-converts it into a JSON response with the standard
-``{"error_code": "...", "message": "..."}`` shape.
-
-Usage::
-
-    from onyx.error_handling.error_codes import OnyxErrorCode
-    from onyx.error_handling.exceptions import OnyxError
-
-    raise OnyxError(OnyxErrorCode.NOT_FOUND, "Session not found")
-
-For upstream errors with a dynamic HTTP status (e.g. billing service),
-use ``status_code_override``::
-
-    raise OnyxError(
-        OnyxErrorCode.BAD_GATEWAY,
-        detail,
-        status_code_override=upstream_status,
-    )
-"""
-
-from fastapi import FastAPI
-from fastapi import Request
-from fastapi.responses import JSONResponse
-
-from onyx.error_handling.error_codes import OnyxErrorCode
-from onyx.utils.logger import setup_logger
-
-logger = setup_logger()
-
-
-class OnyxError(Exception):
-    """Structured error that maps to a specific ``OnyxErrorCode``.
-
-    Attributes:
-        error_code: The ``OnyxErrorCode`` enum member.
-        message: Human-readable message (defaults to the error code string).
-        status_code: HTTP status — either overridden or from the error code.
-    """
-
-    def __init__(
-        self,
-        error_code: OnyxErrorCode,
-        message: str | None = None,
-        *,
-        status_code_override: int | None = None,
-    ) -> None:
-        self.error_code = error_code
-        self.message = message or error_code.code
-        self._status_code_override = status_code_override
-        super().__init__(self.message)
-
-    @property
-    def status_code(self) -> int:
-        return self._status_code_override or self.error_code.status_code
-
-
-def register_onyx_exception_handlers(app: FastAPI) -> None:
-    """Register a global handler that converts ``OnyxError`` to JSON responses.
-
-    Must be called *after* the app is created but *before* it starts serving.
-    The handler logs at WARNING for 4xx and ERROR for 5xx.
-    """
-
-    @app.exception_handler(OnyxError)
-    async def _handle_onyx_error(
-        request: Request,  # noqa: ARG001
-        exc: OnyxError,
-    ) -> JSONResponse:
-        status_code = exc.status_code
-        if status_code >= 500:
-            logger.error(f"OnyxError {exc.error_code.code}: {exc.message}")
-        elif status_code >= 400:
-            logger.warning(f"OnyxError {exc.error_code.code}: {exc.message}")
-
-        return JSONResponse(
-            status_code=status_code,
-            content=exc.error_code.detail(exc.message),
-        )

backend/onyx/federated_connectors/oauth_utils.py

@@ -4,33 +4,39 @@ import base64
 import json
 import uuid
 from typing import Any
+from typing import cast
+from typing import Dict
+from typing import Optional
 
-from onyx.cache.factory import get_cache_backend
 from onyx.configs.app_configs import WEB_DOMAIN
+from onyx.redis.redis_pool import get_redis_client
 from onyx.utils.logger import setup_logger
 
 logger = setup_logger()
 
+# Redis key prefix for OAuth state
 OAUTH_STATE_PREFIX = "federated_oauth"
-OAUTH_STATE_TTL = 300  # 5 minutes
+# Default TTL for OAuth state (5 minutes)
+OAUTH_STATE_TTL = 300
 
 
 class OAuthSession:
-    """Represents an OAuth session stored in the cache backend."""
+    """Represents an OAuth session stored in Redis."""
 
     def __init__(
         self,
         federated_connector_id: int,
         user_id: str,
-        redirect_uri: str | None = None,
-        additional_data: dict[str, Any] | None = None,
+        redirect_uri: Optional[str] = None,
+        additional_data: Optional[Dict[str, Any]] = None,
     ):
         self.federated_connector_id = federated_connector_id
         self.user_id = user_id
         self.redirect_uri = redirect_uri
         self.additional_data = additional_data or {}
 
-    def to_dict(self) -> dict[str, Any]:
+    def to_dict(self) -> Dict[str, Any]:
+        """Convert to dictionary for Redis storage."""
         return {
             "federated_connector_id": self.federated_connector_id,
             "user_id": self.user_id,
@@ -39,7 +45,8 @@ class OAuthSession:
         }
 
     @classmethod
-    def from_dict(cls, data: dict[str, Any]) -> "OAuthSession":
+    def from_dict(cls, data: Dict[str, Any]) -> "OAuthSession":
+        """Create from dictionary retrieved from Redis."""
         return cls(
             federated_connector_id=data["federated_connector_id"],
             user_id=data["user_id"],
@@ -51,27 +58,31 @@ class OAuthSession:
 def generate_oauth_state(
     federated_connector_id: int,
     user_id: str,
-    redirect_uri: str | None = None,
-    additional_data: dict[str, Any] | None = None,
+    redirect_uri: Optional[str] = None,
+    additional_data: Optional[Dict[str, Any]] = None,
     ttl: int = OAUTH_STATE_TTL,
 ) -> str:
     """
-    Generate a secure state parameter and store session data in the cache backend.
+    Generate a secure state parameter and store session data in Redis.
 
     Args:
         federated_connector_id: ID of the federated connector
         user_id: ID of the user initiating OAuth
         redirect_uri: Optional redirect URI after OAuth completion
         additional_data: Any additional data to store with the session
-        ttl: Time-to-live in seconds for the cache key
+        ttl: Time-to-live in seconds for the Redis key
 
     Returns:
         Base64-encoded state parameter
     """
     # Generate a random UUID for the state
     state_uuid = uuid.uuid4()
-    state_b64 = base64.urlsafe_b64encode(state_uuid.bytes).decode("utf-8").rstrip("=")
 
+    # Convert UUID to base64 for URL-safe state parameter
+    state_bytes = state_uuid.bytes
+    state_b64 = base64.urlsafe_b64encode(state_bytes).decode("utf-8").rstrip("=")
+
+    # Create session object
     session = OAuthSession(
         federated_connector_id=federated_connector_id,
         user_id=user_id,
@@ -79,9 +90,15 @@ def generate_oauth_state(
         additional_data=additional_data,
     )
 
-    cache = get_cache_backend()
-    cache_key = f"{OAUTH_STATE_PREFIX}:{state_uuid}"
-    cache.set(cache_key, json.dumps(session.to_dict()), ex=ttl)
+    # Store in Redis with TTL
+    redis_client = get_redis_client()
+    redis_key = f"{OAUTH_STATE_PREFIX}:{state_uuid}"
+
+    redis_client.set(
+        redis_key,
+        json.dumps(session.to_dict()),
+        ex=ttl,
+    )
 
     logger.info(
         f"Generated OAuth state for federated_connector_id={federated_connector_id}, "
@@ -108,15 +125,18 @@ def verify_oauth_state(state: str) -> OAuthSession:
     state_bytes = base64.urlsafe_b64decode(padded_state)
     state_uuid = uuid.UUID(bytes=state_bytes)
 
-    cache = get_cache_backend()
-    cache_key = f"{OAUTH_STATE_PREFIX}:{state_uuid}"
+    # Look up in Redis
+    redis_client = get_redis_client()
+    redis_key = f"{OAUTH_STATE_PREFIX}:{state_uuid}"
 
-    session_data = cache.get(cache_key)
+    session_data = cast(bytes, redis_client.get(redis_key))
     if not session_data:
-        raise ValueError(f"OAuth state not found: {state}")
+        raise ValueError(f"OAuth state not found in Redis: {state}")
 
-    cache.delete(cache_key)
+    # Delete the key after retrieval (one-time use)
+    redis_client.delete(redis_key)
 
+    # Parse and return session
     session_dict = json.loads(session_data)
     return OAuthSession.from_dict(session_dict)
 

backend/onyx/key_value_store/store.py

@@ -1,11 +1,13 @@
 import json
 from typing import cast
 
-from onyx.cache.interface import CacheBackend
+from redis.client import Redis
+
 from onyx.db.engine.sql_engine import get_session_with_current_tenant
 from onyx.db.models import KVStore
 from onyx.key_value_store.interface import KeyValueStore
 from onyx.key_value_store.interface import KvKeyNotFoundError
+from onyx.redis.redis_pool import get_redis_client
 from onyx.utils.logger import setup_logger
 from onyx.utils.special_types import JSON_ro
 
@@ -18,27 +20,22 @@ KV_REDIS_KEY_EXPIRATION = 60 * 60 * 24  # 1 Day
 
 
 class PgRedisKVStore(KeyValueStore):
-    def __init__(self, cache: CacheBackend | None = None) -> None:
-        self._cache = cache
-
-    def _get_cache(self) -> CacheBackend:
-        if self._cache is None:
-            from onyx.cache.factory import get_cache_backend
-
-            self._cache = get_cache_backend()
-        return self._cache
+    def __init__(self, redis_client: Redis | None = None) -> None:
+        # If no redis_client is provided, fall back to the context var
+        if redis_client is not None:
+            self.redis_client = redis_client
+        else:
+            self.redis_client = get_redis_client()
 
     def store(self, key: str, val: JSON_ro, encrypt: bool = False) -> None:
-        # Not encrypted in Cache backend (typically Redis), but encrypted in Postgres
+        # Not encrypted in Redis, but encrypted in Postgres
         try:
-            self._get_cache().set(
+            self.redis_client.set(
                 REDIS_KEY_PREFIX + key, json.dumps(val), ex=KV_REDIS_KEY_EXPIRATION
             )
         except Exception as e:
-            # Fallback gracefully to Postgres if Cache backend fails
-            logger.error(
-                f"Failed to set value in Cache backend for key '{key}': {str(e)}"
-            )
+            # Fallback gracefully to Postgres if Redis fails
+            logger.error(f"Failed to set value in Redis for key '{key}': {str(e)}")
 
         encrypted_val = val if encrypt else None
         plain_val = val if not encrypt else None
@@ -56,12 +53,16 @@ class PgRedisKVStore(KeyValueStore):
     def load(self, key: str, refresh_cache: bool = False) -> JSON_ro:
         if not refresh_cache:
             try:
-                cached = self._get_cache().get(REDIS_KEY_PREFIX + key)
-                if cached is not None:
-                    return json.loads(cached.decode("utf-8"))
+                redis_value = self.redis_client.get(REDIS_KEY_PREFIX + key)
+                if redis_value:
+                    if not isinstance(redis_value, bytes):
+                        raise ValueError(
+                            f"Redis value for key '{key}' is not a bytes object"
+                        )
+                    return json.loads(redis_value.decode("utf-8"))
             except Exception as e:
                 logger.error(
-                    f"Failed to get value from cache for key '{key}': {str(e)}"
+                    f"Failed to get value from Redis for key '{key}': {str(e)}"
                 )
 
         with get_session_with_current_tenant() as db_session:
@@ -78,21 +79,21 @@ class PgRedisKVStore(KeyValueStore):
                 value = None
 
             try:
-                self._get_cache().set(
+                self.redis_client.set(
                     REDIS_KEY_PREFIX + key,
                     json.dumps(value),
                     ex=KV_REDIS_KEY_EXPIRATION,
                 )
             except Exception as e:
-                logger.error(f"Failed to set value in cache for key '{key}': {str(e)}")
+                logger.error(f"Failed to set value in Redis for key '{key}': {str(e)}")
 
             return cast(JSON_ro, value)
 
     def delete(self, key: str) -> None:
         try:
-            self._get_cache().delete(REDIS_KEY_PREFIX + key)
+            self.redis_client.delete(REDIS_KEY_PREFIX + key)
         except Exception as e:
-            logger.error(f"Failed to delete value from cache for key '{key}': {str(e)}")
+            logger.error(f"Failed to delete value from Redis for key '{key}': {str(e)}")
 
         with get_session_with_current_tenant() as db_session:
             result = db_session.query(KVStore).filter_by(key=key).delete()

backend/onyx/llm/litellm_singleton/monkey_patches.py

@@ -67,18 +67,6 @@ Status checked against LiteLLM v1.81.6-nightly (2026-02-02):
    STATUS: STILL NEEDED - litellm_core_utils/litellm_logging.py lines 3185-3199 set
            usage as a dict with chat completion format instead of keeping it as
            ResponseAPIUsage. Our patch creates a deep copy before modification.
-
-7. Responses API metadata=None TypeError (_patch_responses_metadata_none):
-   - LiteLLM's @client decorator wrapper in utils.py uses kwargs.get("metadata", {})
-     to check for router calls, but when metadata is explicitly None (key exists with
-     value None), the default {} is not used
-   - This causes "argument of type 'NoneType' is not iterable" TypeError which swallows
-     the real exception (e.g. AuthenticationError for wrong API key)
-   - Surfaces as: APIConnectionError: OpenAIException - argument of type 'NoneType' is
-     not iterable
-   STATUS: STILL NEEDED - litellm/utils.py wrapper function (line 1721) does not guard
-           against metadata being explicitly None. Triggered when Responses API bridge
-           passes **litellm_params containing metadata=None.
 """
 
 import time
@@ -737,44 +725,6 @@ def _patch_logging_assembled_streaming_response() -> None:
     LiteLLMLoggingObj._get_assembled_streaming_response = _patched_get_assembled_streaming_response  # type: ignore[method-assign]
 
 
-def _patch_responses_metadata_none() -> None:
-    """
-    Patches litellm.responses to normalize metadata=None to metadata={} in kwargs.
-
-    LiteLLM's @client decorator wrapper in utils.py (line 1721) does:
-        _is_litellm_router_call = "model_group" in kwargs.get("metadata", {})
-    When metadata is explicitly None in kwargs, kwargs.get("metadata", {}) returns
-    None (the key exists, so the default is not used), causing:
-        TypeError: argument of type 'NoneType' is not iterable
-
-    This swallows the real exception (e.g. AuthenticationError) and surfaces as:
-        APIConnectionError: OpenAIException - argument of type 'NoneType' is not iterable
-
-    This happens when the Responses API bridge calls litellm.responses() with
-    **litellm_params which may contain metadata=None.
-
-    STATUS: STILL NEEDED - litellm/utils.py wrapper function uses kwargs.get("metadata", {})
-            which does not guard against metadata being explicitly None. Same pattern exists
-            on line 1407 for async path.
-    """
-    import litellm as _litellm
-    from functools import wraps
-
-    original_responses = _litellm.responses
-
-    if getattr(original_responses, "_metadata_patched", False):
-        return
-
-    @wraps(original_responses)
-    def _patched_responses(*args: Any, **kwargs: Any) -> Any:
-        if kwargs.get("metadata") is None:
-            kwargs["metadata"] = {}
-        return original_responses(*args, **kwargs)
-
-    _patched_responses._metadata_patched = True  # type: ignore[attr-defined]
-    _litellm.responses = _patched_responses
-
-
 def apply_monkey_patches() -> None:
     """
     Apply all necessary monkey patches to LiteLLM for compatibility.
@@ -786,7 +736,6 @@ def apply_monkey_patches() -> None:
     - Patching AzureOpenAIResponsesAPIConfig.should_fake_stream to enable native streaming
     - Patching ResponsesAPIResponse.model_construct to fix usage format in all code paths
     - Patching LiteLLMLoggingObj._get_assembled_streaming_response to avoid mutating original response
-    - Patching litellm.responses to fix metadata=None causing TypeError in error handling
     """
     _patch_ollama_chunk_parser()
     _patch_openai_responses_parallel_tool_calls()
@@ -794,4 +743,3 @@ def apply_monkey_patches() -> None:
     _patch_azure_responses_should_fake_stream()
     _patch_responses_api_usage_format()
     _patch_logging_assembled_streaming_response()
-    _patch_responses_metadata_none()

backend/onyx/llm/well_known_providers/auto_update_service.py

@@ -13,38 +13,44 @@ from datetime import datetime
 import httpx
 from sqlalchemy.orm import Session
 
-from onyx.cache.factory import get_cache_backend
 from onyx.configs.app_configs import AUTO_LLM_CONFIG_URL
 from onyx.db.llm import fetch_auto_mode_providers
 from onyx.db.llm import sync_auto_mode_models
 from onyx.llm.well_known_providers.auto_update_models import LLMRecommendations
+from onyx.redis.redis_pool import get_redis_client
 from onyx.utils.logger import setup_logger
 
 logger = setup_logger()
 
-_CACHE_KEY_LAST_UPDATED_AT = "auto_llm_update:last_updated_at"
-_CACHE_TTL_SECONDS = 60 * 60 * 24  # 24 hours
+# Redis key for caching the last updated timestamp (per-tenant)
+_REDIS_KEY_LAST_UPDATED_AT = "auto_llm_update:last_updated_at"
 
 
 def _get_cached_last_updated_at() -> datetime | None:
+    """Get the cached last_updated_at timestamp from Redis."""
     try:
-        value = get_cache_backend().get(_CACHE_KEY_LAST_UPDATED_AT)
-        if value is not None:
+        redis_client = get_redis_client()
+        value = redis_client.get(_REDIS_KEY_LAST_UPDATED_AT)
+        if value and isinstance(value, bytes):
+            # Value is bytes, decode to string then parse as ISO format
             return datetime.fromisoformat(value.decode("utf-8"))
     except Exception as e:
-        logger.warning(f"Failed to get cached last_updated_at: {e}")
+        logger.warning(f"Failed to get cached last_updated_at from Redis: {e}")
     return None
 
 
 def _set_cached_last_updated_at(updated_at: datetime) -> None:
+    """Set the cached last_updated_at timestamp in Redis."""
     try:
-        get_cache_backend().set(
-            _CACHE_KEY_LAST_UPDATED_AT,
+        redis_client = get_redis_client()
+        # Store as ISO format string, with 24 hour expiration
+        redis_client.set(
+            _REDIS_KEY_LAST_UPDATED_AT,
             updated_at.isoformat(),
-            ex=_CACHE_TTL_SECONDS,
+            ex=60 * 60 * 24,  # 24 hours
         )
     except Exception as e:
-        logger.warning(f"Failed to set cached last_updated_at: {e}")
+        logger.warning(f"Failed to set cached last_updated_at in Redis: {e}")
 
 
 def fetch_llm_recommendations_from_github(
@@ -142,8 +148,9 @@ def sync_llm_models_from_github(
 
 
 def reset_cache() -> None:
-    """Reset the cache timestamp. Useful for testing."""
+    """Reset the cache timestamp in Redis. Useful for testing."""
     try:
-        get_cache_backend().delete(_CACHE_KEY_LAST_UPDATED_AT)
+        redis_client = get_redis_client()
+        redis_client.delete(_REDIS_KEY_LAST_UPDATED_AT)
     except Exception as e:
-        logger.warning(f"Failed to reset cache: {e}")
+        logger.warning(f"Failed to reset cache in Redis: {e}")

backend/onyx/main.py

@@ -32,14 +32,11 @@ from onyx.auth.schemas import UserUpdate
 from onyx.auth.users import auth_backend
 from onyx.auth.users import create_onyx_oauth_router
 from onyx.auth.users import fastapi_users
-from onyx.cache.interface import CacheBackendType
 from onyx.configs.app_configs import APP_API_PREFIX
 from onyx.configs.app_configs import APP_HOST
 from onyx.configs.app_configs import APP_PORT
 from onyx.configs.app_configs import AUTH_RATE_LIMITING_ENABLED
 from onyx.configs.app_configs import AUTH_TYPE
-from onyx.configs.app_configs import CACHE_BACKEND
-from onyx.configs.app_configs import DISABLE_VECTOR_DB
 from onyx.configs.app_configs import LOG_ENDPOINT_LATENCY
 from onyx.configs.app_configs import OAUTH_CLIENT_ID
 from onyx.configs.app_configs import OAUTH_CLIENT_SECRET
@@ -59,7 +56,6 @@ from onyx.db.engine.async_sql_engine import get_sqlalchemy_async_engine
 from onyx.db.engine.connection_warmup import warm_up_connections
 from onyx.db.engine.sql_engine import get_session_with_current_tenant
 from onyx.db.engine.sql_engine import SqlEngine
-from onyx.error_handling.exceptions import register_onyx_exception_handlers
 from onyx.file_store.file_store import get_default_file_store
 from onyx.server.api_key.api import router as api_key_router
 from onyx.server.auth_check import check_router_auth
@@ -258,53 +254,8 @@ def include_auth_router_with_prefix(
     )
 
 
-def validate_cache_backend_settings() -> None:
-    """Validate that CACHE_BACKEND=postgres is only used with DISABLE_VECTOR_DB.
-
-    The Postgres cache backend eliminates the Redis dependency, but only works
-    when Celery is not running (which requires DISABLE_VECTOR_DB=true).
-    """
-    if CACHE_BACKEND == CacheBackendType.POSTGRES and not DISABLE_VECTOR_DB:
-        raise RuntimeError(
-            "CACHE_BACKEND=postgres requires DISABLE_VECTOR_DB=true. "
-            "The Postgres cache backend is only supported in no-vector-DB "
-            "deployments where Celery is replaced by the in-process task runner."
-        )
-
-
-def validate_no_vector_db_settings() -> None:
-    """Validate that DISABLE_VECTOR_DB is not combined with incompatible settings.
-
-    Raises RuntimeError if DISABLE_VECTOR_DB is set alongside MULTI_TENANT or ENABLE_CRAFT,
-    since these modes require infrastructure that is removed in no-vector-DB deployments.
-    """
-    if not DISABLE_VECTOR_DB:
-        return
-
-    if MULTI_TENANT:
-        raise RuntimeError(
-            "DISABLE_VECTOR_DB cannot be used with MULTI_TENANT. "
-            "Multi-tenant deployments require the vector database for "
-            "per-tenant document indexing and search. Run in single-tenant "
-            "mode when disabling the vector database."
-        )
-
-    from onyx.server.features.build.configs import ENABLE_CRAFT
-
-    if ENABLE_CRAFT:
-        raise RuntimeError(
-            "DISABLE_VECTOR_DB cannot be used with ENABLE_CRAFT. "
-            "Onyx Craft requires background workers for sandbox lifecycle "
-            "management, which are removed in no-vector-DB deployments. "
-            "Disable Craft (ENABLE_CRAFT=false) when disabling the vector database."
-        )
-
-
 @asynccontextmanager
 async def lifespan(app: FastAPI) -> AsyncGenerator[None, None]:  # noqa: ARG001
-    validate_no_vector_db_settings()
-    validate_cache_backend_settings()
-
     # Set recursion limit
     if SYSTEM_RECURSION_LIMIT is not None:
         sys.setrecursionlimit(SYSTEM_RECURSION_LIMIT)
@@ -373,20 +324,8 @@ async def lifespan(app: FastAPI) -> AsyncGenerator[None, None]:  # noqa: ARG001
     if AUTH_RATE_LIMITING_ENABLED:
         await setup_auth_limiter()
 
-    if DISABLE_VECTOR_DB:
-        from onyx.background.periodic_poller import recover_stuck_user_files
-        from onyx.background.periodic_poller import start_periodic_poller
-
-        recover_stuck_user_files(POSTGRES_DEFAULT_SCHEMA)
-        start_periodic_poller(POSTGRES_DEFAULT_SCHEMA)
-
     yield
 
-    if DISABLE_VECTOR_DB:
-        from onyx.background.periodic_poller import stop_periodic_poller
-
-        stop_periodic_poller()
-
     SqlEngine.reset_engine()
 
     if AUTH_RATE_LIMITING_ENABLED:
@@ -445,8 +384,6 @@ def get_application(lifespan_override: Lifespan | None = None) -> FastAPI:
         status.HTTP_500_INTERNAL_SERVER_ERROR, log_http_error
     )
 
-    register_onyx_exception_handlers(application)
-
     include_router_with_global_prefix_prepended(application, password_router)
     include_router_with_global_prefix_prepended(application, chat_router)
     include_router_with_global_prefix_prepended(application, query_router)

backend/onyx/onyxbot/slack/handlers/handle_message.py

@@ -3,12 +3,10 @@ import datetime
 from slack_sdk import WebClient
 from slack_sdk.errors import SlackApiError
 
-from onyx.auth.schemas import UserRole
 from onyx.configs.onyxbot_configs import ONYX_BOT_FEEDBACK_REMINDER
 from onyx.configs.onyxbot_configs import ONYX_BOT_REACT_EMOJI
 from onyx.db.engine.sql_engine import get_session_with_current_tenant
 from onyx.db.models import SlackChannelConfig
-from onyx.db.user_preferences import activate_user
 from onyx.db.users import add_slack_user_if_not_exists
 from onyx.db.users import get_user_by_email
 from onyx.onyxbot.slack.blocks import get_feedback_reminder_blocks
@@ -245,44 +243,6 @@ def handle_message(
                     )
                     return False
 
-            elif (
-                not existing_user.is_active
-                and existing_user.role == UserRole.SLACK_USER
-            ):
-                check_seat_fn = fetch_ee_implementation_or_noop(
-                    "onyx.db.license",
-                    "check_seat_availability",
-                    None,
-                )
-                seat_result = check_seat_fn(db_session=db_session)
-                if seat_result is not None and not seat_result.available:
-                    logger.info(
-                        f"Blocked inactive Slack user {message_info.email}: "
-                        f"{seat_result.error_message}"
-                    )
-                    respond_in_thread_or_channel(
-                        client=client,
-                        channel=channel,
-                        thread_ts=message_info.msg_to_respond,
-                        text=(
-                            "We weren't able to respond because your organization "
-                            "has reached its user seat limit. Your account is "
-                            "currently deactivated and cannot be reactivated "
-                            "until more seats are available. Please contact "
-                            "your Onyx administrator."
-                        ),
-                    )
-                    return False
-
-                activate_user(existing_user, db_session)
-                invalidate_license_cache_fn = fetch_ee_implementation_or_noop(
-                    "onyx.db.license",
-                    "invalidate_license_cache",
-                    None,
-                )
-                invalidate_license_cache_fn()
-                logger.info(f"Reactivated inactive Slack user {message_info.email}")
-
             add_slack_user_if_not_exists(db_session, message_info.email)
 
         # first check if we need to respond with a standard answer

backend/onyx/server/documents/connector.py

@@ -92,7 +92,6 @@ from onyx.db.connector_credential_pair import get_connector_credential_pairs_for
 from onyx.db.connector_credential_pair import (
     get_connector_credential_pairs_for_user_parallel,
 )
-from onyx.db.connector_credential_pair import verify_user_has_access_to_cc_pair
 from onyx.db.credentials import cleanup_gmail_credentials
 from onyx.db.credentials import cleanup_google_drive_credentials
 from onyx.db.credentials import create_credential
@@ -573,43 +572,6 @@ def _normalize_file_names_for_backwards_compatibility(
     return file_names + file_locations[len(file_names) :]
 
 
-def _fetch_and_check_file_connector_cc_pair_permissions(
-    connector_id: int,
-    user: User,
-    db_session: Session,
-    require_editable: bool,
-) -> ConnectorCredentialPair:
-    cc_pair = fetch_connector_credential_pair_for_connector(db_session, connector_id)
-    if cc_pair is None:
-        raise HTTPException(
-            status_code=404,
-            detail="No Connector-Credential Pair found for this connector",
-        )
-
-    has_requested_access = verify_user_has_access_to_cc_pair(
-        cc_pair_id=cc_pair.id,
-        db_session=db_session,
-        user=user,
-        get_editable=require_editable,
-    )
-    if has_requested_access:
-        return cc_pair
-
-    # Special case: global curators should be able to manage files
-    # for public file connectors even when they are not the creator.
-    if (
-        require_editable
-        and user.role == UserRole.GLOBAL_CURATOR
-        and cc_pair.access_type == AccessType.PUBLIC
-    ):
-        return cc_pair
-
-    raise HTTPException(
-        status_code=403,
-        detail="Access denied. User cannot manage files for this connector.",
-    )
-
-
 @router.post("/admin/connector/file/upload", tags=PUBLIC_API_TAGS)
 def upload_files_api(
     files: list[UploadFile],
@@ -621,7 +583,7 @@ def upload_files_api(
 @router.get("/admin/connector/{connector_id}/files", tags=PUBLIC_API_TAGS)
 def list_connector_files(
     connector_id: int,
-    user: User = Depends(current_curator_or_admin_user),
+    user: User = Depends(current_curator_or_admin_user),  # noqa: ARG001
     db_session: Session = Depends(get_session),
 ) -> ConnectorFilesResponse:
     """List all files in a file connector."""
@@ -634,13 +596,6 @@ def list_connector_files(
             status_code=400, detail="This endpoint only works with file connectors"
         )
 
-    _ = _fetch_and_check_file_connector_cc_pair_permissions(
-        connector_id=connector_id,
-        user=user,
-        db_session=db_session,
-        require_editable=False,
-    )
-
     file_locations = connector.connector_specific_config.get("file_locations", [])
     file_names = connector.connector_specific_config.get("file_names", [])
 
@@ -690,7 +645,7 @@ def update_connector_files(
     connector_id: int,
     files: list[UploadFile] | None = File(None),
     file_ids_to_remove: str = Form("[]"),
-    user: User = Depends(current_curator_or_admin_user),
+    user: User = Depends(current_curator_or_admin_user),  # noqa: ARG001
     db_session: Session = Depends(get_session),
 ) -> FileUploadResponse:
     """
@@ -708,13 +663,12 @@ def update_connector_files(
         )
 
     # Get the connector-credential pair for indexing/pruning triggers
-    # and validate user permissions for file management.
-    cc_pair = _fetch_and_check_file_connector_cc_pair_permissions(
-        connector_id=connector_id,
-        user=user,
-        db_session=db_session,
-        require_editable=True,
-    )
+    cc_pair = fetch_connector_credential_pair_for_connector(db_session, connector_id)
+    if cc_pair is None:
+        raise HTTPException(
+            status_code=404,
+            detail="No Connector-Credential Pair found for this connector",
+        )
 
     # Parse file IDs to remove
     try:

backend/onyx/server/features/build/sandbox/kubernetes/kubernetes_sandbox_manager.py

@@ -1133,8 +1133,7 @@ done
                 # Already deleted
                 service_deleted = True
             else:
-                logger.error(f"Error deleting Service {service_name}: {e}")
-                raise
+                logger.warning(f"Error deleting Service {service_name}: {e}")
 
         pod_deleted = False
         try:
@@ -1149,8 +1148,7 @@ done
                 # Already deleted
                 pod_deleted = True
             else:
-                logger.error(f"Error deleting Pod {pod_name}: {e}")
-                raise
+                logger.warning(f"Error deleting Pod {pod_name}: {e}")
 
         # Wait for resources to be fully deleted to prevent 409 conflicts
         # on immediate re-provisioning

backend/onyx/server/features/build/sandbox/tasks/tasks.py

@@ -80,7 +80,7 @@ def cleanup_idle_sandboxes_task(self: Task, *, tenant_id: str) -> None:  # noqa:
 
     # Prevent overlapping runs of this task
     if not lock.acquire(blocking=False):
-        task_logger.info("cleanup_idle_sandboxes_task - lock not acquired, skipping")
+        task_logger.debug("cleanup_idle_sandboxes_task - lock not acquired, skipping")
         return
 
     try:

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

@@ -32,7 +32,7 @@ from onyx.db.persona import get_persona_snapshots_for_user
 from onyx.db.persona import get_persona_snapshots_paginated
 from onyx.db.persona import mark_persona_as_deleted
 from onyx.db.persona import mark_persona_as_not_deleted
-from onyx.db.persona import update_persona_featured
+from onyx.db.persona import update_persona_is_default
 from onyx.db.persona import update_persona_label
 from onyx.db.persona import update_persona_public_status
 from onyx.db.persona import update_persona_shared
@@ -130,8 +130,8 @@ class IsPublicRequest(BaseModel):
     is_public: bool
 
 
-class IsFeaturedRequest(BaseModel):
-    featured: bool
+class IsDefaultRequest(BaseModel):
+    is_default_persona: bool
 
 
 @admin_router.patch("/{persona_id}/visible")
@@ -168,22 +168,22 @@ def patch_user_persona_public_status(
         raise HTTPException(status_code=403, detail=str(e))
 
 
-@admin_router.patch("/{persona_id}/featured")
-def patch_persona_featured_status(
+@admin_router.patch("/{persona_id}/default")
+def patch_persona_default_status(
     persona_id: int,
-    is_featured_request: IsFeaturedRequest,
+    is_default_request: IsDefaultRequest,
     user: User = Depends(current_curator_or_admin_user),
     db_session: Session = Depends(get_session),
 ) -> None:
     try:
-        update_persona_featured(
+        update_persona_is_default(
             persona_id=persona_id,
-            featured=is_featured_request.featured,
+            is_default=is_default_request.is_default_persona,
             db_session=db_session,
             user=user,
         )
     except ValueError as e:
-        logger.exception("Failed to update persona featured status")
+        logger.exception("Failed to update persona default status")
         raise HTTPException(status_code=403, detail=str(e))
 
 

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

@@ -5,6 +5,7 @@ from pydantic import BaseModel
 from pydantic import Field
 
 from onyx.configs.constants import DocumentSource
+from onyx.context.search.enums import RecencyBiasSetting
 from onyx.db.enums import HierarchyNodeType
 from onyx.db.models import Document
 from onyx.db.models import HierarchyNode
@@ -107,7 +108,11 @@ class PersonaUpsertRequest(BaseModel):
     name: str
     description: str
     document_set_ids: list[int]
+    num_chunks: float
     is_public: bool
+    recency_bias: RecencyBiasSetting
+    llm_filter_extraction: bool
+    llm_relevance_filter: bool
     llm_model_provider_override: str | None = None
     llm_model_version_override: str | None = None
     starter_messages: list[StarterMessage] | None = None
@@ -123,7 +128,7 @@ class PersonaUpsertRequest(BaseModel):
     )
     search_start_date: datetime | None = None
     label_ids: list[int] | None = None
-    featured: bool = False
+    is_default_persona: bool = False
     display_priority: int | None = None
     # Accept string UUIDs from frontend
     user_file_ids: list[str] | None = None
@@ -150,6 +155,9 @@ class MinimalPersonaSnapshot(BaseModel):
     tools: list[ToolSnapshot]
     starter_messages: list[StarterMessage] | None
 
+    llm_relevance_filter: bool
+    llm_filter_extraction: bool
+
     # only show document sets in the UI that the assistant has access to
     document_sets: list[DocumentSetSummary]
     # Counts for knowledge sources (used to determine if search tool should be enabled)
@@ -167,7 +175,7 @@ class MinimalPersonaSnapshot(BaseModel):
     is_public: bool
     is_visible: bool
     display_priority: int | None
-    featured: bool
+    is_default_persona: bool
     builtin_persona: bool
 
     # Used for filtering
@@ -206,6 +214,8 @@ class MinimalPersonaSnapshot(BaseModel):
                 if should_expose_tool_to_fe(tool)
             ],
             starter_messages=persona.starter_messages,
+            llm_relevance_filter=persona.llm_relevance_filter,
+            llm_filter_extraction=persona.llm_filter_extraction,
             document_sets=[
                 DocumentSetSummary.from_model(document_set)
                 for document_set in persona.document_sets
@@ -220,7 +230,7 @@ class MinimalPersonaSnapshot(BaseModel):
             is_public=persona.is_public,
             is_visible=persona.is_visible,
             display_priority=persona.display_priority,
-            featured=persona.featured,
+            is_default_persona=persona.is_default_persona,
             builtin_persona=persona.builtin_persona,
             labels=[PersonaLabelSnapshot.from_model(label) for label in persona.labels],
             owner=(
@@ -242,9 +252,11 @@ class PersonaSnapshot(BaseModel):
     # Return string UUIDs to frontend for consistency
     user_file_ids: list[str]
     display_priority: int | None
-    featured: bool
+    is_default_persona: bool
     builtin_persona: bool
     starter_messages: list[StarterMessage] | None
+    llm_relevance_filter: bool
+    llm_filter_extraction: bool
     tools: list[ToolSnapshot]
     labels: list["PersonaLabelSnapshot"]
     owner: MinimalUserSnapshot | None
@@ -253,6 +265,7 @@ class PersonaSnapshot(BaseModel):
     document_sets: list[DocumentSetSummary]
     llm_model_provider_override: str | None
     llm_model_version_override: str | None
+    num_chunks: float | None
     # Hierarchy nodes attached for scoped search
     hierarchy_nodes: list[HierarchyNodeSnapshot] = Field(default_factory=list)
     # Individual documents attached for scoped search
@@ -276,9 +289,11 @@ class PersonaSnapshot(BaseModel):
             icon_name=persona.icon_name,
             user_file_ids=[str(file.id) for file in persona.user_files],
             display_priority=persona.display_priority,
-            featured=persona.featured,
+            is_default_persona=persona.is_default_persona,
             builtin_persona=persona.builtin_persona,
             starter_messages=persona.starter_messages,
+            llm_relevance_filter=persona.llm_relevance_filter,
+            llm_filter_extraction=persona.llm_filter_extraction,
             tools=[
                 ToolSnapshot.from_model(tool)
                 for tool in persona.tools
@@ -309,6 +324,7 @@ class PersonaSnapshot(BaseModel):
             ],
             llm_model_provider_override=persona.llm_model_provider_override,
             llm_model_version_override=persona.llm_model_version_override,
+            num_chunks=persona.num_chunks,
             system_prompt=persona.system_prompt,
             replace_base_system_prompt=persona.replace_base_system_prompt,
             task_prompt=persona.task_prompt,
@@ -316,10 +332,12 @@ class PersonaSnapshot(BaseModel):
         )
 
 
-# Model with full context on persona's internal settings
+# Model with full context on perona's internal settings
 # This is used for flows which need to know all settings
 class FullPersonaSnapshot(PersonaSnapshot):
     search_start_date: datetime | None = None
+    llm_relevance_filter: bool = False
+    llm_filter_extraction: bool = False
 
     @classmethod
     def from_model(
@@ -342,7 +360,7 @@ class FullPersonaSnapshot(PersonaSnapshot):
             icon_name=persona.icon_name,
             user_file_ids=[str(file.id) for file in persona.user_files],
             display_priority=persona.display_priority,
-            featured=persona.featured,
+            is_default_persona=persona.is_default_persona,
             builtin_persona=persona.builtin_persona,
             starter_messages=persona.starter_messages,
             users=[
@@ -373,7 +391,10 @@ class FullPersonaSnapshot(PersonaSnapshot):
                 DocumentSetSummary.from_model(document_set_model)
                 for document_set_model in persona.document_sets
             ],
+            num_chunks=persona.num_chunks,
             search_start_date=persona.search_start_date,
+            llm_relevance_filter=persona.llm_relevance_filter,
+            llm_filter_extraction=persona.llm_filter_extraction,
             llm_model_provider_override=persona.llm_model_provider_override,
             llm_model_version_override=persona.llm_model_version_override,
             system_prompt=persona.system_prompt,

backend/onyx/server/features/projects/api.py

@@ -2,7 +2,6 @@ import json
 from uuid import UUID
 
 from fastapi import APIRouter
-from fastapi import BackgroundTasks
 from fastapi import Depends
 from fastapi import File
 from fastapi import Form
@@ -13,7 +12,13 @@ from pydantic import BaseModel
 from sqlalchemy.orm import Session
 
 from onyx.auth.users import current_user
-from onyx.configs.app_configs import DISABLE_VECTOR_DB
+from onyx.background.celery.tasks.user_file_processing.tasks import (
+    enqueue_user_file_project_sync_task,
+)
+from onyx.background.celery.tasks.user_file_processing.tasks import (
+    get_user_file_project_sync_queue_depth,
+)
+from onyx.background.celery.versioned_apps.client import app as client_app
 from onyx.configs.constants import OnyxCeleryPriority
 from onyx.configs.constants import OnyxCeleryQueues
 from onyx.configs.constants import OnyxCeleryTask
@@ -29,6 +34,7 @@ from onyx.db.models import UserProject
 from onyx.db.persona import get_personas_by_ids
 from onyx.db.projects import get_project_token_count
 from onyx.db.projects import upload_files_to_user_files_with_indexing
+from onyx.redis.redis_pool import get_redis_client
 from onyx.server.features.projects.models import CategorizedFilesSnapshot
 from onyx.server.features.projects.models import ChatSessionRequest
 from onyx.server.features.projects.models import TokenCountResponse
@@ -49,27 +55,7 @@ class UserFileDeleteResult(BaseModel):
     assistant_names: list[str] = []
 
 
-def _trigger_user_file_project_sync(
-    user_file_id: UUID,
-    tenant_id: str,
-    background_tasks: BackgroundTasks | None = None,
-) -> None:
-    if DISABLE_VECTOR_DB and background_tasks is not None:
-        from onyx.background.task_utils import drain_project_sync_loop
-
-        background_tasks.add_task(drain_project_sync_loop, tenant_id)
-        logger.info(f"Queued in-process project sync for user_file_id={user_file_id}")
-        return
-
-    from onyx.background.celery.tasks.user_file_processing.tasks import (
-        enqueue_user_file_project_sync_task,
-    )
-    from onyx.background.celery.tasks.user_file_processing.tasks import (
-        get_user_file_project_sync_queue_depth,
-    )
-    from onyx.background.celery.versioned_apps.client import app as client_app
-    from onyx.redis.redis_pool import get_redis_client
-
+def _trigger_user_file_project_sync(user_file_id: UUID, tenant_id: str) -> None:
     queue_depth = get_user_file_project_sync_queue_depth(client_app)
     if queue_depth > USER_FILE_PROJECT_SYNC_MAX_QUEUE_DEPTH:
         logger.warning(
@@ -125,7 +111,6 @@ def create_project(
 
 @router.post("/file/upload", tags=PUBLIC_API_TAGS)
 def upload_user_files(
-    bg_tasks: BackgroundTasks,
     files: list[UploadFile] = File(...),
     project_id: int | None = Form(None),
     temp_id_map: str | None = Form(None),  # JSON string mapping hashed key -> temp_id
@@ -152,12 +137,12 @@ def upload_user_files(
             user=user,
             temp_id_map=parsed_temp_id_map,
             db_session=db_session,
-            background_tasks=bg_tasks if DISABLE_VECTOR_DB else None,
         )
 
         return CategorizedFilesSnapshot.from_result(categorized_files_result)
 
     except Exception as e:
+        # Log error with type, message, and stack for easier debugging
         logger.exception(f"Error uploading files - {type(e).__name__}: {str(e)}")
         raise HTTPException(
             status_code=500,
@@ -207,7 +192,6 @@ def get_files_in_project(
 def unlink_user_file_from_project(
     project_id: int,
     file_id: UUID,
-    bg_tasks: BackgroundTasks,
     user: User = Depends(current_user),
     db_session: Session = Depends(get_session),
 ) -> Response:
@@ -224,6 +208,7 @@ def unlink_user_file_from_project(
     if project is None:
         raise HTTPException(status_code=404, detail="Project not found")
 
+    user_id = user.id
     user_file = (
         db_session.query(UserFile)
         .filter(UserFile.id == file_id, UserFile.user_id == user_id)
@@ -239,7 +224,7 @@ def unlink_user_file_from_project(
         db_session.commit()
 
     tenant_id = get_current_tenant_id()
-    _trigger_user_file_project_sync(user_file.id, tenant_id, bg_tasks)
+    _trigger_user_file_project_sync(user_file.id, tenant_id)
 
     return Response(status_code=204)
 
@@ -252,7 +237,6 @@ def unlink_user_file_from_project(
 def link_user_file_to_project(
     project_id: int,
     file_id: UUID,
-    bg_tasks: BackgroundTasks,
     user: User = Depends(current_user),
     db_session: Session = Depends(get_session),
 ) -> UserFileSnapshot:
@@ -284,7 +268,7 @@ def link_user_file_to_project(
         db_session.commit()
 
     tenant_id = get_current_tenant_id()
-    _trigger_user_file_project_sync(user_file.id, tenant_id, bg_tasks)
+    _trigger_user_file_project_sync(user_file.id, tenant_id)
 
     return UserFileSnapshot.from_model(user_file)
 
@@ -351,7 +335,7 @@ def upsert_project_instructions(
 class ProjectPayload(BaseModel):
     project: UserProjectSnapshot
     files: list[UserFileSnapshot] | None = None
-    persona_id_to_featured: dict[int, bool] | None = None
+    persona_id_to_is_default: dict[int, bool] | None = None
 
 
 @router.get(
@@ -370,11 +354,13 @@ def get_project_details(
         if session.persona_id is not None
     ]
     personas = get_personas_by_ids(persona_ids, db_session)
-    persona_id_to_featured = {persona.id: persona.featured for persona in personas}
+    persona_id_to_is_default = {
+        persona.id: persona.is_default_persona for persona in personas
+    }
     return ProjectPayload(
         project=project,
         files=files,
-        persona_id_to_featured=persona_id_to_featured,
+        persona_id_to_is_default=persona_id_to_is_default,
     )
 
 
@@ -440,7 +426,6 @@ def delete_project(
 @router.delete("/file/{file_id}", tags=PUBLIC_API_TAGS)
 def delete_user_file(
     file_id: UUID,
-    bg_tasks: BackgroundTasks,
     user: User = Depends(current_user),
     db_session: Session = Depends(get_session),
 ) -> UserFileDeleteResult:
@@ -473,25 +458,15 @@ def delete_user_file(
     db_session.commit()
 
     tenant_id = get_current_tenant_id()
-    if DISABLE_VECTOR_DB:
-        from onyx.background.task_utils import drain_delete_loop
-
-        bg_tasks.add_task(drain_delete_loop, tenant_id)
-        logger.info(f"Queued in-process delete for user_file_id={user_file.id}")
-    else:
-        from onyx.background.celery.versioned_apps.client import app as client_app
-
-        task = client_app.send_task(
-            OnyxCeleryTask.DELETE_SINGLE_USER_FILE,
-            kwargs={"user_file_id": str(user_file.id), "tenant_id": tenant_id},
-            queue=OnyxCeleryQueues.USER_FILE_DELETE,
-            priority=OnyxCeleryPriority.HIGH,
-        )
-        logger.info(
-            f"Triggered delete for user_file_id={user_file.id} "
-            f"with task_id={task.id}"
-        )
-
+    task = client_app.send_task(
+        OnyxCeleryTask.DELETE_SINGLE_USER_FILE,
+        kwargs={"user_file_id": str(user_file.id), "tenant_id": tenant_id},
+        queue=OnyxCeleryQueues.USER_FILE_DELETE,
+        priority=OnyxCeleryPriority.HIGH,
+    )
+    logger.info(
+        f"Triggered delete for user_file_id={user_file.id} with task_id={task.id}"
+    )
     return UserFileDeleteResult(
         has_associations=False, project_names=[], assistant_names=[]
     )

backend/onyx/server/features/projects/projects_file_utils.py

@@ -7,14 +7,13 @@ from PIL import UnidentifiedImageError
 from pydantic import BaseModel
 from pydantic import ConfigDict
 from pydantic import Field
-from sqlalchemy.orm import Session
 
 from onyx.configs.app_configs import FILE_TOKEN_COUNT_THRESHOLD
-from onyx.db.llm import fetch_default_llm_model
 from onyx.file_processing.extract_file_text import extract_file_text
 from onyx.file_processing.extract_file_text import get_file_ext
 from onyx.file_processing.file_types import OnyxFileExtensions
 from onyx.file_processing.password_validation import is_file_password_protected
+from onyx.llm.factory import get_default_llm
 from onyx.natural_language_processing.utils import get_tokenizer
 from onyx.utils.logger import setup_logger
 from shared_configs.configs import MULTI_TENANT
@@ -117,9 +116,7 @@ def estimate_image_tokens_for_upload(
             pass
 
 
-def categorize_uploaded_files(
-    files: list[UploadFile], db_session: Session
-) -> CategorizedFiles:
+def categorize_uploaded_files(files: list[UploadFile]) -> CategorizedFiles:
     """
     Categorize uploaded files based on text extractability and tokenized length.
 
@@ -131,11 +128,11 @@ def categorize_uploaded_files(
     """
 
     results = CategorizedFiles()
-    default_model = fetch_default_llm_model(db_session)
+    llm = get_default_llm()
 
-    model_name = default_model.name if default_model else None
-    provider_type = default_model.llm_provider.provider if default_model else None
-    tokenizer = get_tokenizer(model_name=model_name, provider_type=provider_type)
+    tokenizer = get_tokenizer(
+        model_name=llm.config.model_name, provider_type=llm.config.model_provider
+    )
 
     # Check if threshold checks should be skipped
     skip_threshold = False

backend/onyx/server/features/release_notes/utils.py

@@ -8,10 +8,10 @@ import httpx
 from sqlalchemy.orm import Session
 
 from onyx import __version__
-from onyx.cache.factory import get_shared_cache_backend
 from onyx.configs.app_configs import INSTANCE_TYPE
 from onyx.configs.constants import OnyxRedisLocks
 from onyx.db.release_notes import create_release_notifications_for_versions
+from onyx.redis.redis_pool import get_shared_redis_client
 from onyx.server.features.release_notes.constants import AUTO_REFRESH_THRESHOLD_SECONDS
 from onyx.server.features.release_notes.constants import FETCH_TIMEOUT
 from onyx.server.features.release_notes.constants import GITHUB_CHANGELOG_RAW_URL
@@ -113,46 +113,60 @@ def parse_mdx_to_release_note_entries(mdx_content: str) -> list[ReleaseNoteEntry
 
 
 def get_cached_etag() -> str | None:
-    cache = get_shared_cache_backend()
+    """Get the cached GitHub ETag from Redis."""
+    redis_client = get_shared_redis_client()
     try:
-        etag = cache.get(REDIS_KEY_ETAG)
+        etag = redis_client.get(REDIS_KEY_ETAG)
         if etag:
-            return etag.decode("utf-8")
+            return etag.decode("utf-8") if isinstance(etag, bytes) else str(etag)
         return None
     except Exception as e:
-        logger.error(f"Failed to get cached etag: {e}")
+        logger.error(f"Failed to get cached etag from Redis: {e}")
         return None
 
 
 def get_last_fetch_time() -> datetime | None:
-    cache = get_shared_cache_backend()
+    """Get the last fetch timestamp from Redis."""
+    redis_client = get_shared_redis_client()
     try:
-        raw = cache.get(REDIS_KEY_FETCHED_AT)
-        if not raw:
+        fetched_at_str = redis_client.get(REDIS_KEY_FETCHED_AT)
+        if not fetched_at_str:
             return None
 
-        last_fetch = datetime.fromisoformat(raw.decode("utf-8"))
+        decoded = (
+            fetched_at_str.decode("utf-8")
+            if isinstance(fetched_at_str, bytes)
+            else str(fetched_at_str)
+        )
+
+        last_fetch = datetime.fromisoformat(decoded)
+
+        # Defensively ensure timezone awareness
+        # fromisoformat() returns naive datetime if input lacks timezone
         if last_fetch.tzinfo is None:
+            # Assume UTC for naive datetimes
             last_fetch = last_fetch.replace(tzinfo=timezone.utc)
         else:
+            # Convert to UTC if timezone-aware
             last_fetch = last_fetch.astimezone(timezone.utc)
 
         return last_fetch
     except Exception as e:
-        logger.error(f"Failed to get last fetch time from cache: {e}")
+        logger.error(f"Failed to get last fetch time from Redis: {e}")
         return None
 
 
 def save_fetch_metadata(etag: str | None) -> None:
-    cache = get_shared_cache_backend()
+    """Save ETag and fetch timestamp to Redis."""
+    redis_client = get_shared_redis_client()
     now = datetime.now(timezone.utc)
 
     try:
-        cache.set(REDIS_KEY_FETCHED_AT, now.isoformat(), ex=REDIS_CACHE_TTL)
+        redis_client.set(REDIS_KEY_FETCHED_AT, now.isoformat(), ex=REDIS_CACHE_TTL)
         if etag:
-            cache.set(REDIS_KEY_ETAG, etag, ex=REDIS_CACHE_TTL)
+            redis_client.set(REDIS_KEY_ETAG, etag, ex=REDIS_CACHE_TTL)
     except Exception as e:
-        logger.error(f"Failed to save fetch metadata to cache: {e}")
+        logger.error(f"Failed to save fetch metadata to Redis: {e}")
 
 
 def is_cache_stale() -> bool:
@@ -182,10 +196,11 @@ def ensure_release_notes_fresh_and_notify(db_session: Session) -> None:
     if not is_cache_stale():
         return
 
-    cache = get_shared_cache_backend()
-    lock = cache.lock(
+    # Acquire lock to prevent concurrent fetches
+    redis_client = get_shared_redis_client()
+    lock = redis_client.lock(
         OnyxRedisLocks.RELEASE_NOTES_FETCH_LOCK,
-        timeout=90,
+        timeout=90,  # 90 second timeout for the lock
     )
 
     # Non-blocking acquire - if we can't get the lock, another request is handling it

backend/onyx/server/kg/api.py

@@ -5,6 +5,7 @@ from sqlalchemy.orm import Session
 from onyx.auth.users import current_admin_user
 from onyx.configs.constants import TMP_DRALPHA_PERSONA_NAME
 from onyx.configs.kg_configs import KG_BETA_ASSISTANT_DESCRIPTION
+from onyx.context.search.enums import RecencyBiasSetting
 from onyx.db.engine.sql_engine import get_session
 from onyx.db.entities import get_entity_stats_by_grounded_source_name
 from onyx.db.entity_type import get_configured_entity_types
@@ -133,7 +134,11 @@ def enable_or_disable_kg(
         system_prompt=KG_BETA_ASSISTANT_SYSTEM_PROMPT,
         task_prompt=KG_BETA_ASSISTANT_TASK_PROMPT,
         datetime_aware=False,
+        num_chunks=25,
+        llm_relevance_filter=False,
         is_public=False,
+        llm_filter_extraction=False,
+        recency_bias=RecencyBiasSetting.NO_DECAY,
         document_set_ids=[],
         tool_ids=[search_tool.id, kg_tool.id],
         llm_model_provider_override=None,
@@ -142,7 +147,7 @@ def enable_or_disable_kg(
         users=[user.id],
         groups=[],
         label_ids=[],
-        featured=False,
+        is_default_persona=False,
         display_priority=0,
         user_file_ids=[],
     )

backend/onyx/server/manage/image_generation/api.py

@@ -97,6 +97,7 @@ def _build_llm_provider_request(
             ),  # Only this from source
             api_base=api_base,  # From request
             api_version=api_version,  # From request
+            default_model_name=model_name,
             deployment_name=deployment_name,  # From request
             is_public=True,
             groups=[],
@@ -135,6 +136,7 @@ def _build_llm_provider_request(
         api_key=api_key,
         api_base=api_base,
         api_version=api_version,
+        default_model_name=model_name,
         deployment_name=deployment_name,
         is_public=True,
         groups=[],
@@ -166,6 +168,7 @@ def _create_image_gen_llm_provider__no_commit(
         api_key=provider_request.api_key,
         api_base=provider_request.api_base,
         api_version=provider_request.api_version,
+        default_model_name=provider_request.default_model_name,
         deployment_name=provider_request.deployment_name,
         is_public=provider_request.is_public,
         custom_config=provider_request.custom_config,

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

@@ -22,10 +22,7 @@ from onyx.auth.users import current_chat_accessible_user
 from onyx.db.engine.sql_engine import get_session
 from onyx.db.enums import LLMModelFlowType
 from onyx.db.llm import can_user_access_llm_provider
-from onyx.db.llm import fetch_default_llm_model
-from onyx.db.llm import fetch_default_vision_model
 from onyx.db.llm import fetch_existing_llm_provider
-from onyx.db.llm import fetch_existing_llm_provider_by_id
 from onyx.db.llm import fetch_existing_llm_providers
 from onyx.db.llm import fetch_existing_models
 from onyx.db.llm import fetch_persona_with_groups
@@ -55,12 +52,11 @@ from onyx.llm.well_known_providers.llm_provider_options import (
 )
 from onyx.server.manage.llm.models import BedrockFinalModelResponse
 from onyx.server.manage.llm.models import BedrockModelsRequest
-from onyx.server.manage.llm.models import DefaultModel
 from onyx.server.manage.llm.models import LLMCost
 from onyx.server.manage.llm.models import LLMProviderDescriptor
-from onyx.server.manage.llm.models import LLMProviderResponse
 from onyx.server.manage.llm.models import LLMProviderUpsertRequest
 from onyx.server.manage.llm.models import LLMProviderView
+from onyx.server.manage.llm.models import ModelConfigurationUpsertRequest
 from onyx.server.manage.llm.models import OllamaFinalModelResponse
 from onyx.server.manage.llm.models import OllamaModelDetails
 from onyx.server.manage.llm.models import OllamaModelsRequest
@@ -237,9 +233,12 @@ def test_llm_configuration(
 
     test_api_key = test_llm_request.api_key
     test_custom_config = test_llm_request.custom_config
-    if test_llm_request.id:
-        existing_provider = fetch_existing_llm_provider_by_id(
-            id=test_llm_request.id, db_session=db_session
+    if test_llm_request.name:
+        # NOTE: we are querying by name. we probably should be querying by an invariant id, but
+        # as it turns out the name is not editable in the UI and other code also keys off name,
+        # so we won't rock the boat just yet.
+        existing_provider = fetch_existing_llm_provider(
+            name=test_llm_request.name, db_session=db_session
         )
         if existing_provider:
             test_custom_config = _restore_masked_custom_config_values(
@@ -269,7 +268,7 @@ def test_llm_configuration(
 
     llm = get_llm(
         provider=test_llm_request.provider,
-        model=test_llm_request.model,
+        model=test_llm_request.default_model_name,
         api_key=test_api_key,
         api_base=test_llm_request.api_base,
         api_version=test_llm_request.api_version,
@@ -304,7 +303,7 @@ def list_llm_providers(
     include_image_gen: bool = Query(False),
     _: User = Depends(current_admin_user),
     db_session: Session = Depends(get_session),
-) -> LLMProviderResponse[LLMProviderView]:
+) -> list[LLMProviderView]:
     start_time = datetime.now(timezone.utc)
     logger.debug("Starting to fetch LLM providers")
 
@@ -329,15 +328,7 @@ def list_llm_providers(
     duration = (end_time - start_time).total_seconds()
     logger.debug(f"Completed fetching LLM providers in {duration:.2f} seconds")
 
-    return LLMProviderResponse[LLMProviderView].from_models(
-        providers=llm_provider_list,
-        default_text=DefaultModel.from_model_config(
-            fetch_default_llm_model(db_session)
-        ),
-        default_vision=DefaultModel.from_model_config(
-            fetch_default_vision_model(db_session)
-        ),
-    )
+    return llm_provider_list
 
 
 @admin_router.put("/provider")
@@ -353,44 +344,18 @@ def put_llm_provider(
     # validate request (e.g. if we're intending to create but the name already exists we should throw an error)
     # NOTE: may involve duplicate fetching to Postgres, but we're assuming SQLAlchemy is smart enough to cache
     # the result
-    existing_provider = None
-    if llm_provider_upsert_request.id:
-        existing_provider = fetch_existing_llm_provider_by_id(
-            id=llm_provider_upsert_request.id, db_session=db_session
-        )
-
-    # Check name constraints
-    # TODO: Once port from name to id is complete, unique name will no longer be required
-    if existing_provider and llm_provider_upsert_request.name != existing_provider.name:
-        raise HTTPException(
-            status_code=400,
-            detail="Renaming providers is not currently supported",
-        )
-
-    found_provider = fetch_existing_llm_provider(
+    existing_provider = fetch_existing_llm_provider(
         name=llm_provider_upsert_request.name, db_session=db_session
     )
-    if found_provider is not None and found_provider is not existing_provider:
-        raise HTTPException(
-            status_code=400,
-            detail=f"Provider with name={llm_provider_upsert_request.name} already exists",
-        )
-
     if existing_provider and is_creation:
         raise HTTPException(
             status_code=400,
-            detail=(
-                f"LLM Provider with name {llm_provider_upsert_request.name} and "
-                f"id={llm_provider_upsert_request.id} already exists"
-            ),
+            detail=f"LLM Provider with name {llm_provider_upsert_request.name} already exists",
         )
     elif not existing_provider and not is_creation:
         raise HTTPException(
             status_code=400,
-            detail=(
-                f"LLM Provider with name {llm_provider_upsert_request.name} and "
-                f"id={llm_provider_upsert_request.id} does not exist"
-            ),
+            detail=f"LLM Provider with name {llm_provider_upsert_request.name} does not exist",
         )
 
     # SSRF Protection: Validate api_base and custom_config match stored values
@@ -428,6 +393,22 @@ def put_llm_provider(
                 deduplicated_personas.append(persona_id)
         llm_provider_upsert_request.personas = deduplicated_personas
 
+    default_model_found = False
+
+    for model_configuration in llm_provider_upsert_request.model_configurations:
+        if model_configuration.name == llm_provider_upsert_request.default_model_name:
+            model_configuration.is_visible = True
+            default_model_found = True
+
+    # TODO: Remove this logic on api change
+    # Believed to be a dead pathway but we want to be safe for now
+    if not default_model_found:
+        llm_provider_upsert_request.model_configurations.append(
+            ModelConfigurationUpsertRequest(
+                name=llm_provider_upsert_request.default_model_name, is_visible=True
+            )
+        )
+
     # the llm api key is sanitized when returned to clients, so the only time we
     # should get a real key is when it is explicitly changed
     if existing_provider and not llm_provider_upsert_request.api_key_changed:
@@ -457,8 +438,8 @@ def put_llm_provider(
             config = fetch_llm_recommendations_from_github()
             if config and llm_provider_upsert_request.provider in config.providers:
                 # Refetch the provider to get the updated model
-                updated_provider = fetch_existing_llm_provider_by_id(
-                    id=result.id, db_session=db_session
+                updated_provider = fetch_existing_llm_provider(
+                    name=llm_provider_upsert_request.name, db_session=db_session
                 )
                 if updated_provider:
                     sync_auto_mode_models(
@@ -479,48 +460,37 @@ def put_llm_provider(
 @admin_router.delete("/provider/{provider_id}")
 def delete_llm_provider(
     provider_id: int,
-    force: bool = Query(False),
     _: User = Depends(current_admin_user),
     db_session: Session = Depends(get_session),
 ) -> None:
     try:
-        if not force:
-            model = fetch_default_llm_model(db_session)
-
-            if model and model.llm_provider_id == provider_id:
-                raise HTTPException(
-                    status_code=400,
-                    detail="Cannot delete the default LLM provider",
-                )
-
         remove_llm_provider(db_session, provider_id)
     except ValueError as e:
         raise HTTPException(status_code=404, detail=str(e))
 
 
-@admin_router.post("/default")
+@admin_router.post("/provider/{provider_id}/default")
 def set_provider_as_default(
-    default_model_request: DefaultModel,
+    provider_id: int,
     _: User = Depends(current_admin_user),
     db_session: Session = Depends(get_session),
 ) -> None:
-    update_default_provider(
-        provider_id=default_model_request.provider_id,
-        model_name=default_model_request.model_name,
-        db_session=db_session,
-    )
+    update_default_provider(provider_id=provider_id, db_session=db_session)
 
 
-@admin_router.post("/default-vision")
+@admin_router.post("/provider/{provider_id}/default-vision")
 def set_provider_as_default_vision(
-    default_model: DefaultModel,
+    provider_id: int,
+    vision_model: str | None = Query(
+        None, description="The default vision model to use"
+    ),
     _: User = Depends(current_admin_user),
     db_session: Session = Depends(get_session),
 ) -> None:
+    if vision_model is None:
+        raise HTTPException(status_code=404, detail="Vision model not provided")
     update_default_vision_provider(
-        provider_id=default_model.provider_id,
-        vision_model=default_model.model_name,
-        db_session=db_session,
+        provider_id=provider_id, vision_model=vision_model, db_session=db_session
     )
 
 
@@ -546,7 +516,7 @@ def get_auto_config(
 def get_vision_capable_providers(
     _: User = Depends(current_admin_user),
     db_session: Session = Depends(get_session),
-) -> LLMProviderResponse[VisionProviderResponse]:
+) -> list[VisionProviderResponse]:
     """Return a list of LLM providers and their models that support image input"""
     vision_models = fetch_existing_models(
         db_session=db_session, flow_types=[LLMModelFlowType.VISION]
@@ -575,13 +545,7 @@ def get_vision_capable_providers(
     ]
 
     logger.debug(f"Found {len(vision_provider_response)} vision-capable providers")
-
-    return LLMProviderResponse[VisionProviderResponse].from_models(
-        providers=vision_provider_response,
-        default_vision=DefaultModel.from_model_config(
-            fetch_default_vision_model(db_session)
-        ),
-    )
+    return vision_provider_response
 
 
 """Endpoints for all"""
@@ -591,7 +555,7 @@ def get_vision_capable_providers(
 def list_llm_provider_basics(
     user: User = Depends(current_chat_accessible_user),
     db_session: Session = Depends(get_session),
-) -> LLMProviderResponse[LLMProviderDescriptor]:
+) -> list[LLMProviderDescriptor]:
     """Get LLM providers accessible to the current user.
 
     Returns:
@@ -613,9 +577,9 @@ def list_llm_provider_basics(
     for provider in all_providers:
         # Use centralized access control logic with persona=None since we're
         # listing providers without a specific persona context. This correctly:
-        # - Includes public providers WITHOUT persona restrictions
+        # - Includes all public providers
         # - Includes providers user can access via group membership
-        # - Excludes providers with persona restrictions (requires specific persona)
+        # - Excludes persona-only restricted providers (requires specific persona)
         # - Excludes non-public providers with no restrictions (admin-only)
         if can_user_access_llm_provider(
             provider, user_group_ids, persona=None, is_admin=is_admin
@@ -628,15 +592,7 @@ def list_llm_provider_basics(
         f"Completed fetching {len(accessible_providers)} user-accessible providers in {duration:.2f} seconds"
     )
 
-    return LLMProviderResponse[LLMProviderDescriptor].from_models(
-        providers=accessible_providers,
-        default_text=DefaultModel.from_model_config(
-            fetch_default_llm_model(db_session)
-        ),
-        default_vision=DefaultModel.from_model_config(
-            fetch_default_vision_model(db_session)
-        ),
-    )
+    return accessible_providers
 
 
 def get_valid_model_names_for_persona(
@@ -648,7 +604,7 @@ def get_valid_model_names_for_persona(
 
     Returns a list of model names (e.g., ["gpt-4o", "claude-3-5-sonnet"]) that are
     available to the user when using this persona, respecting all RBAC restrictions.
-    Public providers are included unless they have persona restrictions that exclude this persona.
+    Public providers are always included.
     """
     persona = fetch_persona_with_groups(db_session, persona_id)
     if not persona:
@@ -662,7 +618,7 @@ def get_valid_model_names_for_persona(
 
     valid_models = []
     for llm_provider_model in all_providers:
-        # Check access with persona context — respects all RBAC restrictions
+        # Public providers always included, restricted checked via RBAC
         if can_user_access_llm_provider(
             llm_provider_model, user_group_ids, persona, is_admin=is_admin
         ):
@@ -679,11 +635,11 @@ def list_llm_providers_for_persona(
     persona_id: int,
     user: User = Depends(current_chat_accessible_user),
     db_session: Session = Depends(get_session),
-) -> LLMProviderResponse[LLMProviderDescriptor]:
+) -> list[LLMProviderDescriptor]:
     """Get LLM providers for a specific persona.
 
     Returns providers that the user can access when using this persona:
-    - Public providers (respecting persona restrictions if set)
+    - All public providers (is_public=True) - ALWAYS included
     - Restricted providers user can access via group/persona restrictions
 
     This endpoint is used for background fetching of restricted providers
@@ -712,7 +668,7 @@ def list_llm_providers_for_persona(
     llm_provider_list: list[LLMProviderDescriptor] = []
 
     for llm_provider_model in all_providers:
-        # Check access with persona context — respects persona restrictions
+        # Use simplified access check - public providers always included
         if can_user_access_llm_provider(
             llm_provider_model, user_group_ids, persona, is_admin=is_admin
         ):
@@ -726,51 +682,7 @@ def list_llm_providers_for_persona(
         f"Completed fetching {len(llm_provider_list)} LLM providers for persona {persona_id} in {duration:.2f} seconds"
     )
 
-    # Get the default model and vision model for the persona
-    # TODO: Port persona's over to use ID
-    persona_default_provider = persona.llm_model_provider_override
-    persona_default_model = persona.llm_model_version_override
-
-    default_text_model = fetch_default_llm_model(db_session)
-    default_vision_model = fetch_default_vision_model(db_session)
-
-    # Build default_text and default_vision using persona overrides when available,
-    # falling back to the global defaults.
-    default_text = DefaultModel.from_model_config(default_text_model)
-    default_vision = DefaultModel.from_model_config(default_vision_model)
-
-    if persona_default_provider:
-        provider = fetch_existing_llm_provider(persona_default_provider, db_session)
-        if provider and can_user_access_llm_provider(
-            provider, user_group_ids, persona, is_admin=is_admin
-        ):
-            if persona_default_model:
-                # Persona specifies both provider and model — use them directly
-                default_text = DefaultModel(
-                    provider_id=provider.id,
-                    model_name=persona_default_model,
-                )
-            else:
-                # Persona specifies only the provider — pick a visible (public) model,
-                # falling back to any model on this provider
-                visible_model = next(
-                    (mc for mc in provider.model_configurations if mc.is_visible),
-                    None,
-                )
-                fallback_model = visible_model or next(
-                    iter(provider.model_configurations), None
-                )
-                if fallback_model:
-                    default_text = DefaultModel(
-                        provider_id=provider.id,
-                        model_name=fallback_model.name,
-                    )
-
-    return LLMProviderResponse[LLMProviderDescriptor].from_models(
-        providers=llm_provider_list,
-        default_text=default_text,
-        default_vision=default_vision,
-    )
+    return llm_provider_list
 
 
 @admin_router.get("/provider-contextual-cost")

backend/onyx/server/manage/llm/models.py

@@ -1,9 +1,5 @@
-from __future__ import annotations
-
 from typing import Any
-from typing import Generic
 from typing import TYPE_CHECKING
-from typing import TypeVar
 
 from pydantic import BaseModel
 from pydantic import Field
@@ -25,22 +21,50 @@ if TYPE_CHECKING:
         ModelConfiguration as ModelConfigurationModel,
     )
 
-T = TypeVar("T", "LLMProviderDescriptor", "LLMProviderView", "VisionProviderResponse")
+
+# TODO: Clear this up on api refactor
+# There is still logic that requires sending each providers default model name
+# There is no logic that requires sending the providers default vision model name
+# We only send for the one that is actually the default
+def get_default_llm_model_name(llm_provider_model: "LLMProviderModel") -> str:
+    """Find the default conversation model name for a provider.
+
+    Returns the model name if found, otherwise returns empty string.
+    """
+    for model_config in llm_provider_model.model_configurations:
+        for flow in model_config.llm_model_flows:
+            if flow.is_default and flow.llm_model_flow_type == LLMModelFlowType.CHAT:
+                return model_config.name
+    return ""
+
+
+def get_default_vision_model_name(llm_provider_model: "LLMProviderModel") -> str | None:
+    """Find the default vision model name for a provider.
+
+    Returns the model name if found, otherwise returns None.
+    """
+    for model_config in llm_provider_model.model_configurations:
+        for flow in model_config.llm_model_flows:
+            if flow.is_default and flow.llm_model_flow_type == LLMModelFlowType.VISION:
+                return model_config.name
+    return None
 
 
 class TestLLMRequest(BaseModel):
     # provider level
-    id: int | None = None
+    name: str | None = None
     provider: str
-    model: str
     api_key: str | None = None
     api_base: str | None = None
     api_version: str | None = None
     custom_config: dict[str, str] | None = None
 
     # model level
+    default_model_name: str
     deployment_name: str | None = None
 
+    model_configurations: list["ModelConfigurationUpsertRequest"]
+
     # if try and use the existing API/custom config key
     api_key_changed: bool
     custom_config_changed: bool
@@ -56,10 +80,13 @@ class LLMProviderDescriptor(BaseModel):
     """A descriptor for an LLM provider that can be safely viewed by
     non-admin users. Used when giving a list of available LLMs."""
 
-    id: int
     name: str
     provider: str
     provider_display_name: str  # Human-friendly name like "Claude (Anthropic)"
+    default_model_name: str
+    is_default_provider: bool | None
+    is_default_vision_provider: bool | None
+    default_vision_model: str | None
     model_configurations: list["ModelConfigurationView"]
 
     @classmethod
@@ -72,12 +99,24 @@ class LLMProviderDescriptor(BaseModel):
         )
 
         provider = llm_provider_model.provider
+        default_model_name = get_default_llm_model_name(llm_provider_model)
+        default_vision_model = get_default_vision_model_name(llm_provider_model)
+
+        is_default_provider = bool(default_model_name)
+        is_default_vision_provider = default_vision_model is not None
+
+        default_model_name = (
+            default_model_name or llm_provider_model.default_model_name or ""
+        )
 
         return cls(
-            id=llm_provider_model.id,
             name=llm_provider_model.name,
             provider=provider,
             provider_display_name=get_provider_display_name(provider),
+            default_model_name=default_model_name,
+            is_default_provider=is_default_provider,
+            is_default_vision_provider=is_default_vision_provider,
+            default_vision_model=default_vision_model,
             model_configurations=filter_model_configurations(
                 llm_provider_model.model_configurations, provider
             ),
@@ -91,17 +130,18 @@ class LLMProvider(BaseModel):
     api_base: str | None = None
     api_version: str | None = None
     custom_config: dict[str, str] | None = None
+    default_model_name: str
     is_public: bool = True
     is_auto_mode: bool = False
     groups: list[int] = Field(default_factory=list)
     personas: list[int] = Field(default_factory=list)
     deployment_name: str | None = None
+    default_vision_model: str | None = None
 
 
 class LLMProviderUpsertRequest(LLMProvider):
     # should only be used for a "custom" provider
     # for default providers, the built-in model names are used
-    id: int | None = None
     api_key_changed: bool = False
     custom_config_changed: bool = False
     model_configurations: list["ModelConfigurationUpsertRequest"] = []
@@ -117,6 +157,8 @@ class LLMProviderView(LLMProvider):
     """Stripped down representation of LLMProvider for display / limited access info only"""
 
     id: int
+    is_default_provider: bool | None = None
+    is_default_vision_provider: bool | None = None
     model_configurations: list["ModelConfigurationView"]
 
     @classmethod
@@ -138,6 +180,16 @@ class LLMProviderView(LLMProvider):
 
         provider = llm_provider_model.provider
 
+        default_model_name = get_default_llm_model_name(llm_provider_model)
+        default_vision_model = get_default_vision_model_name(llm_provider_model)
+
+        is_default_provider = bool(default_model_name)
+        is_default_vision_provider = default_vision_model is not None
+
+        default_model_name = (
+            default_model_name or llm_provider_model.default_model_name or ""
+        )
+
         return cls(
             id=llm_provider_model.id,
             name=llm_provider_model.name,
@@ -150,6 +202,10 @@ class LLMProviderView(LLMProvider):
             api_base=llm_provider_model.api_base,
             api_version=llm_provider_model.api_version,
             custom_config=llm_provider_model.custom_config,
+            default_model_name=default_model_name,
+            is_default_provider=is_default_provider,
+            is_default_vision_provider=is_default_vision_provider,
+            default_vision_model=default_vision_model,
             is_public=llm_provider_model.is_public,
             is_auto_mode=llm_provider_model.is_auto_mode,
             groups=groups,
@@ -369,38 +425,3 @@ class OpenRouterFinalModelResponse(BaseModel):
         int | None
     )  # From OpenRouter API context_length (may be missing for some models)
     supports_image_input: bool
-
-
-class DefaultModel(BaseModel):
-    provider_id: int
-    model_name: str
-
-    @classmethod
-    def from_model_config(
-        cls, model_config: ModelConfigurationModel | None
-    ) -> DefaultModel | None:
-        if not model_config:
-            return None
-        return cls(
-            provider_id=model_config.llm_provider_id,
-            model_name=model_config.name,
-        )
-
-
-class LLMProviderResponse(BaseModel, Generic[T]):
-    providers: list[T]
-    default_text: DefaultModel | None = None
-    default_vision: DefaultModel | None = None
-
-    @classmethod
-    def from_models(
-        cls,
-        providers: list[T],
-        default_text: DefaultModel | None = None,
-        default_vision: DefaultModel | None = None,
-    ) -> LLMProviderResponse[T]:
-        return cls(
-            providers=providers,
-            default_text=default_text,
-            default_vision=default_vision,
-        )

backend/onyx/server/manage/slack_bot.py

@@ -198,6 +198,7 @@ def patch_slack_channel_config(
             channel_name=channel_config["channel_name"],
             document_set_ids=slack_channel_config_creation_request.document_sets,
             existing_persona_id=existing_persona_id,
+            enable_auto_filters=slack_channel_config_creation_request.enable_auto_filters,
         ).id
 
     slack_channel_config_model = update_slack_channel_config(

backend/onyx/server/query_and_chat/chat_backend.py

@@ -13,13 +13,13 @@ from fastapi import Request
 from fastapi import Response
 from fastapi.responses import StreamingResponse
 from pydantic import BaseModel
+from redis.client import Redis
 from sqlalchemy.orm import Session
 
 from onyx.auth.api_key import get_hashed_api_key_from_request
 from onyx.auth.pat import get_hashed_pat_from_request
 from onyx.auth.users import current_chat_accessible_user
 from onyx.auth.users import current_user
-from onyx.cache.factory import get_cache_backend
 from onyx.chat.chat_processing_checker import is_chat_session_processing
 from onyx.chat.chat_state import ChatStateContainer
 from onyx.chat.chat_utils import convert_chat_history_basic
@@ -67,6 +67,7 @@ from onyx.llm.constants import LlmProviderNames
 from onyx.llm.factory import get_default_llm
 from onyx.llm.factory import get_llm_for_persona
 from onyx.llm.factory import get_llm_token_counter
+from onyx.redis.redis_pool import get_redis_client
 from onyx.secondary_llm_flows.chat_session_naming import generate_chat_session_name
 from onyx.server.api_key_usage import check_api_key_usage
 from onyx.server.query_and_chat.models import ChatFeedbackRequest
@@ -151,20 +152,10 @@ def get_user_chat_sessions(
     project_id: int | None = None,
     only_non_project_chats: bool = True,
     include_failed_chats: bool = False,
-    page_size: int = Query(default=50, ge=1, le=100),
-    before: str | None = Query(default=None),
 ) -> ChatSessionsResponse:
     user_id = user.id
 
     try:
-        before_dt = (
-            datetime.datetime.fromisoformat(before) if before is not None else None
-        )
-    except ValueError:
-        raise HTTPException(status_code=422, detail="Invalid 'before' timestamp format")
-
-    try:
-        # Fetch one extra to determine if there are more results
         chat_sessions = get_chat_sessions_by_user(
             user_id=user_id,
             deleted=False,
@@ -172,16 +163,11 @@ def get_user_chat_sessions(
             project_id=project_id,
             only_non_project_chats=only_non_project_chats,
             include_failed_chats=include_failed_chats,
-            limit=page_size + 1,
-            before=before_dt,
         )
 
     except ValueError:
         raise ValueError("Chat session does not exist or has been deleted")
 
-    has_more = len(chat_sessions) > page_size
-    chat_sessions = chat_sessions[:page_size]
-
     return ChatSessionsResponse(
         sessions=[
             ChatSessionDetails(
@@ -195,8 +181,7 @@ def get_user_chat_sessions(
                 current_temperature_override=chat.temperature_override,
             )
             for chat in chat_sessions
-        ],
-        has_more=has_more,
+        ]
     )
 
 
@@ -329,7 +314,7 @@ def get_chat_session(
     ]
 
     try:
-        is_processing = is_chat_session_processing(session_id, get_cache_backend())
+        is_processing = is_chat_session_processing(session_id, get_redis_client())
         # Edit the last message to indicate loading (Overriding default message value)
         if is_processing and chat_message_details:
             last_msg = chat_message_details[-1]
@@ -926,10 +911,11 @@ async def search_chats(
 def stop_chat_session(
     chat_session_id: UUID,
     user: User = Depends(current_user),  # noqa: ARG001
+    redis_client: Redis = Depends(get_redis_client),
 ) -> dict[str, str]:
     """
-    Stop a chat session by setting a stop signal.
+    Stop a chat session by setting a stop signal in Redis.
     This endpoint is called by the frontend when the user clicks the stop button.
     """
-    set_fence(chat_session_id, get_cache_backend(), True)
+    set_fence(chat_session_id, redis_client, True)
     return {"message": "Chat session stopped"}

backend/onyx/server/query_and_chat/models.py

@@ -192,7 +192,6 @@ class ChatSessionDetails(BaseModel):
 
 class ChatSessionsResponse(BaseModel):
     sessions: list[ChatSessionDetails]
-    has_more: bool = False
 
 
 class ChatMessageDetail(BaseModel):

backend/onyx/server/settings/models.py

@@ -19,7 +19,6 @@ class ApplicationStatus(str, Enum):
     PAYMENT_REMINDER = "payment_reminder"
     GRACE_PERIOD = "grace_period"
     GATED_ACCESS = "gated_access"
-    SEAT_LIMIT_EXCEEDED = "seat_limit_exceeded"
 
 
 class Notification(BaseModel):
@@ -83,10 +82,6 @@ class Settings(BaseModel):
     # Default Assistant settings
     disable_default_assistant: bool | None = False
 
-    # Seat usage - populated by license enforcement when seat limit is exceeded
-    seat_count: int | None = None
-    used_seats: int | None = None
-
     # OpenSearch migration
     opensearch_indexing_enabled: bool = False
 

backend/onyx/server/settings/store.py

@@ -1,4 +1,3 @@
-from onyx.cache.factory import get_cache_backend
 from onyx.configs.app_configs import DISABLE_USER_KNOWLEDGE
 from onyx.configs.app_configs import ENABLE_OPENSEARCH_INDEXING_FOR_ONYX
 from onyx.configs.app_configs import ONYX_QUERY_HISTORY_TYPE
@@ -7,8 +6,11 @@ from onyx.configs.constants import KV_SETTINGS_KEY
 from onyx.configs.constants import OnyxRedisLocks
 from onyx.key_value_store.factory import get_kv_store
 from onyx.key_value_store.interface import KvKeyNotFoundError
+from onyx.redis.redis_pool import get_redis_client
 from onyx.server.settings.models import Settings
 from onyx.utils.logger import setup_logger
+from shared_configs.configs import MULTI_TENANT
+from shared_configs.contextvars import get_current_tenant_id
 
 logger = setup_logger()
 
@@ -31,22 +33,30 @@ def load_settings() -> Settings:
         logger.error(f"Error loading settings from KV store: {str(e)}")
         settings = Settings()
 
-    cache = get_cache_backend()
+    tenant_id = get_current_tenant_id() if MULTI_TENANT else None
+    redis_client = get_redis_client(tenant_id=tenant_id)
 
     try:
-        value = cache.get(OnyxRedisLocks.ANONYMOUS_USER_ENABLED)
+        value = redis_client.get(OnyxRedisLocks.ANONYMOUS_USER_ENABLED)
         if value is not None:
+            assert isinstance(value, bytes)
             anonymous_user_enabled = int(value.decode("utf-8")) == 1
         else:
+            # Default to False
             anonymous_user_enabled = False
-            cache.set(OnyxRedisLocks.ANONYMOUS_USER_ENABLED, "0", ex=SETTINGS_TTL)
+            # Optionally store the default back to Redis
+            redis_client.set(
+                OnyxRedisLocks.ANONYMOUS_USER_ENABLED, "0", ex=SETTINGS_TTL
+            )
     except Exception as e:
-        logger.error(f"Error loading anonymous user setting from cache: {str(e)}")
+        # Log the error and reset to default
+        logger.error(f"Error loading anonymous user setting from Redis: {str(e)}")
         anonymous_user_enabled = False
 
     settings.anonymous_user_enabled = anonymous_user_enabled
     settings.query_history_type = ONYX_QUERY_HISTORY_TYPE
 
+    # Override user knowledge setting if disabled via environment variable
     if DISABLE_USER_KNOWLEDGE:
         settings.user_knowledge_enabled = False
 
@@ -56,10 +66,11 @@ def load_settings() -> Settings:
 
 
 def store_settings(settings: Settings) -> None:
-    cache = get_cache_backend()
+    tenant_id = get_current_tenant_id() if MULTI_TENANT else None
+    redis_client = get_redis_client(tenant_id=tenant_id)
 
     if settings.anonymous_user_enabled is not None:
-        cache.set(
+        redis_client.set(
             OnyxRedisLocks.ANONYMOUS_USER_ENABLED,
             "1" if settings.anonymous_user_enabled else "0",
             ex=SETTINGS_TTL,

backend/onyx/setup.py

@@ -25,7 +25,6 @@ from onyx.db.enums import EmbeddingPrecision
 from onyx.db.index_attempt import cancel_indexing_attempts_past_model
 from onyx.db.index_attempt import expire_index_attempts
 from onyx.db.llm import fetch_default_llm_model
-from onyx.db.llm import fetch_existing_llm_provider
 from onyx.db.llm import update_default_provider
 from onyx.db.llm import upsert_llm_provider
 from onyx.db.search_settings import get_active_search_settings
@@ -255,18 +254,14 @@ def setup_postgres(db_session: Session) -> None:
         logger.notice("Setting up default OpenAI LLM for dev.")
 
         llm_model = GEN_AI_MODEL_VERSION or "gpt-4o-mini"
-        provider_name = "DevEnvPresetOpenAI"
-        existing = fetch_existing_llm_provider(
-            name=provider_name, db_session=db_session
-        )
         model_req = LLMProviderUpsertRequest(
-            id=existing.id if existing else None,
-            name=provider_name,
+            name="DevEnvPresetOpenAI",
             provider=LlmProviderNames.OPENAI,
             api_key=GEN_AI_API_KEY,
             api_base=None,
             api_version=None,
             custom_config=None,
+            default_model_name=llm_model,
             is_public=True,
             groups=[],
             model_configurations=[
@@ -278,9 +273,7 @@ def setup_postgres(db_session: Session) -> None:
         new_llm_provider = upsert_llm_provider(
             llm_provider_upsert_request=model_req, db_session=db_session
         )
-        update_default_provider(
-            provider_id=new_llm_provider.id, model_name=llm_model, db_session=db_session
-        )
+        update_default_provider(provider_id=new_llm_provider.id, db_session=db_session)
 
 
 def update_default_multipass_indexing(db_session: Session) -> None:

backend/onyx/tools/models.py

@@ -93,8 +93,6 @@ class ToolResponse(BaseModel):
         # | WebContentResponse
         # This comes from custom tools, tool result needs to be saved
         | CustomToolCallSummary
-        # This comes from code interpreter, carries generated files
-        | PythonToolRichResponse
         # If the rich response is a string, this is what's saved to the tool call in the DB
         | str
         | None  # If nothing needs to be persisted outside of the string value passed to the LLM
@@ -195,12 +193,6 @@ class ChatFile(BaseModel):
     model_config = ConfigDict(arbitrary_types_allowed=True)
 
 
-class PythonToolRichResponse(BaseModel):
-    """Rich response from the Python tool carrying generated files."""
-
-    generated_files: list[PythonExecutionFile] = []
-
-
 class PythonToolOverrideKwargs(BaseModel):
     """Override kwargs for the Python/Code Interpreter tool."""
 
@@ -253,7 +245,6 @@ class ToolCallInfo(BaseModel):
     tool_call_response: str
     search_docs: list[SearchDoc] | None = None
     generated_images: list[GeneratedImage] | None = None
-    generated_files: list[PythonExecutionFile] | None = None
 
 
 CHAT_SESSION_ID_PLACEHOLDER = "CHAT_SESSION_ID"

backend/onyx/tools/tool_implementations/python/code_interpreter_client.py

@@ -1,5 +1,4 @@
 import json
-import time
 from collections.abc import Generator
 from typing import Literal
 from typing import TypedDict
@@ -13,9 +12,6 @@ from onyx.utils.logger import setup_logger
 
 logger = setup_logger()
 
-_HEALTH_CACHE_TTL_SECONDS = 30
-_health_cache: dict[str, tuple[float, bool]] = {}
-
 
 class FileInput(TypedDict):
     """Input file to be staged in execution workspace"""
@@ -102,32 +98,16 @@ class CodeInterpreterClient:
             payload["files"] = files
         return payload
 
-    def health(self, use_cache: bool = False) -> bool:
-        """Check if the Code Interpreter service is healthy
-
-        Args:
-            use_cache: When True, return a cached result if available and
-                       within the TTL window. The cache is always populated
-                       after a live request regardless of this flag.
-        """
-        if use_cache:
-            cached = _health_cache.get(self.base_url)
-            if cached is not None:
-                cached_at, cached_result = cached
-                if time.monotonic() - cached_at < _HEALTH_CACHE_TTL_SECONDS:
-                    return cached_result
-
+    def health(self) -> bool:
+        """Check if the Code Interpreter service is healthy"""
         url = f"{self.base_url}/health"
         try:
             response = self.session.get(url, timeout=5)
             response.raise_for_status()
-            result = response.json().get("status") == "ok"
+            return response.json().get("status") == "ok"
         except Exception as e:
             logger.warning(f"Exception caught when checking health, e={e}")
-            result = False
-
-        _health_cache[self.base_url] = (time.monotonic(), result)
-        return result
+            return False
 
     def execute(
         self,

backend/onyx/tools/tool_implementations/python/python_tool.py

@@ -23,7 +23,6 @@ from onyx.tools.interface import Tool
 from onyx.tools.models import LlmPythonExecutionResult
 from onyx.tools.models import PythonExecutionFile
 from onyx.tools.models import PythonToolOverrideKwargs
-from onyx.tools.models import PythonToolRichResponse
 from onyx.tools.models import ToolCallException
 from onyx.tools.models import ToolResponse
 from onyx.tools.tool_implementations.python.code_interpreter_client import (
@@ -108,11 +107,7 @@ class PythonTool(Tool[PythonToolOverrideKwargs]):
         if not CODE_INTERPRETER_BASE_URL:
             return False
         server = fetch_code_interpreter_server(db_session)
-        if not server.server_enabled:
-            return False
-
-        client = CodeInterpreterClient()
-        return client.health(use_cache=True)
+        return server.server_enabled
 
     def tool_definition(self) -> dict:
         return {
@@ -330,9 +325,7 @@ class PythonTool(Tool[PythonToolOverrideKwargs]):
             llm_response = adapter.dump_json(result).decode()
 
             return ToolResponse(
-                rich_response=PythonToolRichResponse(
-                    generated_files=generated_files,
-                ),
+                rich_response=None,  # No rich response needed for Python tool
                 llm_facing_response=llm_response,
             )
 

backend/onyx/tools/tool_implementations/web_search/web_search_tool.py

@@ -1,5 +1,6 @@
 import json
 from typing import Any
+from typing import cast
 
 from sqlalchemy.orm import Session
 from typing_extensions import override
@@ -56,30 +57,6 @@ def _sanitize_query(query: str) -> str:
     return " ".join(sanitized.split())
 
 
-def _normalize_queries_input(raw: Any) -> list[str]:
-    """Coerce LLM output to a list of sanitized query strings.
-
-    Accepts a bare string or a list (possibly with non-string elements).
-    Sanitizes each query (strip control chars, normalize whitespace) and
-    drops empty or whitespace-only entries.
-    """
-    if isinstance(raw, str):
-        raw = raw.strip()
-        if not raw:
-            return []
-        raw = [raw]
-    elif not isinstance(raw, list):
-        return []
-    result: list[str] = []
-    for q in raw:
-        if q is None:
-            continue
-        sanitized = _sanitize_query(str(q))
-        if sanitized:
-            result.append(sanitized)
-    return result
-
-
 class WebSearchTool(Tool[WebSearchToolOverrideKwargs]):
     NAME = "web_search"
     DESCRIPTION = "Search the web for information."
@@ -212,7 +189,13 @@ class WebSearchTool(Tool[WebSearchToolOverrideKwargs]):
                     f'like: {{"queries": ["your search query here"]}}'
                 ),
             )
-        queries = _normalize_queries_input(llm_kwargs[QUERIES_FIELD])
+        raw_queries = cast(list[str], llm_kwargs[QUERIES_FIELD])
+
+        # Normalize queries:
+        # - remove control characters (null bytes, etc.) that LLMs sometimes produce
+        # - collapse whitespace and strip
+        # - drop empty/whitespace-only queries
+        queries = [sanitized for q in raw_queries if (sanitized := _sanitize_query(q))]
         if not queries:
             raise ToolCallException(
                 message=(

backend/pyproject.toml

@@ -8,3 +8,37 @@ dependencies = [
 
 [tool.uv.sources]
 onyx = { workspace = true }
+
+[tool.mypy]
+plugins = "sqlalchemy.ext.mypy.plugin"
+mypy_path = "backend"
+explicit_package_bases = true
+disallow_untyped_defs = true
+warn_unused_ignores = true
+enable_error_code = ["possibly-undefined"]
+strict_equality = true
+# Patterns match paths whether mypy is run from backend/ (CI) or repo root (e.g. VS Code extension with target ./backend)
+exclude = [
+  "(?:^|/)generated/",
+  "(?:^|/)\\.venv/",
+  "(?:^|/)onyx/server/features/build/sandbox/kubernetes/docker/skills/",
+  "(?:^|/)onyx/server/features/build/sandbox/kubernetes/docker/templates/",
+]
+
+[[tool.mypy.overrides]]
+module = "alembic.versions.*"
+disable_error_code = ["var-annotated"]
+
+[[tool.mypy.overrides]]
+module = "alembic_tenants.versions.*"
+disable_error_code = ["var-annotated"]
+
+[[tool.mypy.overrides]]
+module = "generated.*"
+follow_imports = "silent"
+ignore_errors = true
+
+[[tool.mypy.overrides]]
+module = "transformers.*"
+follow_imports = "skip"
+ignore_errors = true

backend/requirements/default.txt

@@ -528,7 +528,7 @@ lxml==5.3.0
     #   unstructured
     #   xmlsec
     #   zeep
-lxml-html-clean==0.4.4
+lxml-html-clean==0.4.3
     # via lxml
 magika==0.6.3
     # via markitdown
@@ -809,7 +809,7 @@ pypandoc-binary==1.16.2
     # via onyx
 pyparsing==3.2.5
     # via httplib2
-pypdf==6.7.5
+pypdf==6.7.3
     # via
     #   onyx
     #   unstructured-client

backend/scripts/dev_run_background_jobs.py

@@ -16,6 +16,10 @@ def monitor_process(process_name: str, process: subprocess.Popen) -> None:
 
 
 def run_jobs() -> None:
+    # Check if we should use lightweight mode, defaults to True, change to False to use separate background workers
+    use_lightweight = True
+
+    # command setup
     cmd_worker_primary = [
         "celery",
         "-A",
@@ -70,48 +74,6 @@ def run_jobs() -> None:
         "--queues=connector_doc_fetching",
     ]
 
-    cmd_worker_heavy = [
-        "celery",
-        "-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,csv_generation,sandbox",
-    ]
-
-    cmd_worker_monitoring = [
-        "celery",
-        "-A",
-        "onyx.background.celery.versioned_apps.monitoring",
-        "worker",
-        "--pool=threads",
-        "--concurrency=1",
-        "--prefetch-multiplier=1",
-        "--loglevel=INFO",
-        "--hostname=monitoring@%n",
-        "-Q",
-        "monitoring",
-    ]
-
-    cmd_worker_user_file_processing = [
-        "celery",
-        "-A",
-        "onyx.background.celery.versioned_apps.user_file_processing",
-        "worker",
-        "--pool=threads",
-        "--concurrency=2",
-        "--prefetch-multiplier=1",
-        "--loglevel=INFO",
-        "--hostname=user_file_processing@%n",
-        "-Q",
-        "user_file_processing,user_file_project_sync,user_file_delete",
-    ]
-
     cmd_beat = [
         "celery",
         "-A",
@@ -120,31 +82,144 @@ def run_jobs() -> None:
         "--loglevel=INFO",
     ]
 
-    all_workers = [
-        ("PRIMARY", cmd_worker_primary),
-        ("LIGHT", cmd_worker_light),
-        ("DOCPROCESSING", cmd_worker_docprocessing),
-        ("DOCFETCHING", cmd_worker_docfetching),
-        ("HEAVY", cmd_worker_heavy),
-        ("MONITORING", cmd_worker_monitoring),
-        ("USER_FILE_PROCESSING", cmd_worker_user_file_processing),
-        ("BEAT", cmd_beat),
-    ]
+    # Prepare background worker commands based on mode
+    if use_lightweight:
+        print("Starting workers in LIGHTWEIGHT mode (single background worker)")
+        cmd_worker_background = [
+            "celery",
+            "-A",
+            "onyx.background.celery.versioned_apps.background",
+            "worker",
+            "--pool=threads",
+            "--concurrency=6",
+            "--prefetch-multiplier=1",
+            "--loglevel=INFO",
+            "--hostname=background@%n",
+            "-Q",
+            "connector_pruning,connector_doc_permissions_sync,connector_external_group_sync,csv_generation,monitoring,user_file_processing,user_file_project_sync,user_file_delete,opensearch_migration",
+        ]
+        background_workers = [("BACKGROUND", cmd_worker_background)]
+    else:
+        print("Starting workers in STANDARD mode (separate background workers)")
+        cmd_worker_heavy = [
+            "celery",
+            "-A",
+            "onyx.background.celery.versioned_apps.heavy",
+            "worker",
+            "--pool=threads",
+            "--concurrency=4",
+            "--prefetch-multiplier=1",
+            "--loglevel=INFO",
+            "--hostname=heavy@%n",
+            "-Q",
+            "connector_pruning,sandbox",
+        ]
+        cmd_worker_monitoring = [
+            "celery",
+            "-A",
+            "onyx.background.celery.versioned_apps.monitoring",
+            "worker",
+            "--pool=threads",
+            "--concurrency=1",
+            "--prefetch-multiplier=1",
+            "--loglevel=INFO",
+            "--hostname=monitoring@%n",
+            "-Q",
+            "monitoring",
+        ]
+        cmd_worker_user_file_processing = [
+            "celery",
+            "-A",
+            "onyx.background.celery.versioned_apps.user_file_processing",
+            "worker",
+            "--pool=threads",
+            "--concurrency=2",
+            "--prefetch-multiplier=1",
+            "--loglevel=INFO",
+            "--hostname=user_file_processing@%n",
+            "-Q",
+            "user_file_processing,user_file_project_sync,connector_doc_permissions_sync,connector_external_group_sync,csv_generation,user_file_delete",
+        ]
+        background_workers = [
+            ("HEAVY", cmd_worker_heavy),
+            ("MONITORING", cmd_worker_monitoring),
+            ("USER_FILE_PROCESSING", cmd_worker_user_file_processing),
+        ]
 
-    processes = []
-    for name, cmd in all_workers:
+    # spawn processes
+    worker_primary_process = subprocess.Popen(
+        cmd_worker_primary, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, text=True
+    )
+
+    worker_light_process = subprocess.Popen(
+        cmd_worker_light, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, text=True
+    )
+
+    worker_docprocessing_process = subprocess.Popen(
+        cmd_worker_docprocessing,
+        stdout=subprocess.PIPE,
+        stderr=subprocess.STDOUT,
+        text=True,
+    )
+
+    worker_docfetching_process = subprocess.Popen(
+        cmd_worker_docfetching,
+        stdout=subprocess.PIPE,
+        stderr=subprocess.STDOUT,
+        text=True,
+    )
+
+    beat_process = subprocess.Popen(
+        cmd_beat, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, text=True
+    )
+
+    # Spawn background worker processes based on mode
+    background_processes = []
+    for name, cmd in background_workers:
         process = subprocess.Popen(
             cmd, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, text=True
         )
-        processes.append((name, process))
+        background_processes.append((name, process))
 
-    threads = []
-    for name, process in processes:
+    # monitor threads
+    worker_primary_thread = threading.Thread(
+        target=monitor_process, args=("PRIMARY", worker_primary_process)
+    )
+    worker_light_thread = threading.Thread(
+        target=monitor_process, args=("LIGHT", worker_light_process)
+    )
+    worker_docprocessing_thread = threading.Thread(
+        target=monitor_process, args=("DOCPROCESSING", worker_docprocessing_process)
+    )
+    worker_docfetching_thread = threading.Thread(
+        target=monitor_process, args=("DOCFETCHING", worker_docfetching_process)
+    )
+    beat_thread = threading.Thread(target=monitor_process, args=("BEAT", beat_process))
+
+    # Create monitor threads for background workers
+    background_threads = []
+    for name, process in background_processes:
         thread = threading.Thread(target=monitor_process, args=(name, process))
-        threads.append(thread)
+        background_threads.append(thread)
+
+    # Start all threads
+    worker_primary_thread.start()
+    worker_light_thread.start()
+    worker_docprocessing_thread.start()
+    worker_docfetching_thread.start()
+    beat_thread.start()
+
+    for thread in background_threads:
         thread.start()
 
-    for thread in threads:
+    # Wait for all threads
+    worker_primary_thread.join()
+    worker_light_thread.join()
+    worker_docprocessing_thread.join()
+    worker_docfetching_thread.join()
+    beat_thread.join()
+
+    for thread in background_threads:
         thread.join()
 
 

backend/scripts/supervisord_entrypoint.sh

@@ -1,5 +1,23 @@
 #!/bin/sh
-# Entrypoint script for supervisord
+# Entrypoint script for supervisord that sets environment variables
+# for controlling which celery workers to start
+
+# Default to lightweight mode if not set
+if [ -z "$USE_LIGHTWEIGHT_BACKGROUND_WORKER" ]; then
+    export USE_LIGHTWEIGHT_BACKGROUND_WORKER="true"
+fi
+
+# Set the complementary variable for supervisord
+# because it doesn't support %(not ENV_USE_LIGHTWEIGHT_BACKGROUND_WORKER) syntax
+if [ "$USE_LIGHTWEIGHT_BACKGROUND_WORKER" = "true" ]; then
+    export USE_SEPARATE_BACKGROUND_WORKERS="false"
+else
+    export USE_SEPARATE_BACKGROUND_WORKERS="true"
+fi
+
+echo "Worker mode configuration:"
+echo "  USE_LIGHTWEIGHT_BACKGROUND_WORKER=$USE_LIGHTWEIGHT_BACKGROUND_WORKER"
+echo "  USE_SEPARATE_BACKGROUND_WORKERS=$USE_SEPARATE_BACKGROUND_WORKERS"
 
 # Launch supervisord with environment variables available
 exec /usr/bin/supervisord -c /etc/supervisor/conf.d/supervisord.conf

backend/supervisord.conf

@@ -39,6 +39,7 @@ autorestart=true
 startsecs=10
 stopasgroup=true
 
+# Standard mode: Light worker for fast operations
 # NOTE: only allowing configuration here and not in the other celery workers,
 # since this is often the bottleneck for "sync" jobs (e.g. document set syncing,
 # user group syncing, deletion, etc.)
@@ -53,7 +54,26 @@ redirect_stderr=true
 autorestart=true
 startsecs=10
 stopasgroup=true
+autostart=%(ENV_USE_SEPARATE_BACKGROUND_WORKERS)s
 
+# Lightweight mode: single consolidated background worker
+# Used when USE_LIGHTWEIGHT_BACKGROUND_WORKER=true (default)
+# Consolidates: light, docprocessing, docfetching, heavy, monitoring, user_file_processing
+[program:celery_worker_background]
+command=celery -A onyx.background.celery.versioned_apps.background worker
+    --loglevel=INFO
+    --hostname=background@%%n
+    -Q vespa_metadata_sync,connector_deletion,doc_permissions_upsert,checkpoint_cleanup,index_attempt_cleanup,sandbox,docprocessing,connector_doc_fetching,connector_pruning,connector_doc_permissions_sync,connector_external_group_sync,csv_generation,monitoring,user_file_processing,user_file_project_sync,opensearch_migration
+stdout_logfile=/var/log/celery_worker_background.log
+stdout_logfile_maxbytes=16MB
+redirect_stderr=true
+autorestart=true
+startsecs=10
+stopasgroup=true
+autostart=%(ENV_USE_LIGHTWEIGHT_BACKGROUND_WORKER)s
+
+# Standard mode: separate workers for different background tasks
+# Used when USE_LIGHTWEIGHT_BACKGROUND_WORKER=false
 [program:celery_worker_heavy]
 command=celery -A onyx.background.celery.versioned_apps.heavy worker
     --loglevel=INFO
@@ -65,7 +85,9 @@ redirect_stderr=true
 autorestart=true
 startsecs=10
 stopasgroup=true
+autostart=%(ENV_USE_SEPARATE_BACKGROUND_WORKERS)s
 
+# Standard mode: Document processing worker
 [program:celery_worker_docprocessing]
 command=celery -A onyx.background.celery.versioned_apps.docprocessing worker
     --loglevel=INFO
@@ -77,6 +99,7 @@ redirect_stderr=true
 autorestart=true
 startsecs=10
 stopasgroup=true
+autostart=%(ENV_USE_SEPARATE_BACKGROUND_WORKERS)s
 
 [program:celery_worker_user_file_processing]
 command=celery -A onyx.background.celery.versioned_apps.user_file_processing worker
@@ -89,7 +112,9 @@ redirect_stderr=true
 autorestart=true
 startsecs=10
 stopasgroup=true
+autostart=%(ENV_USE_SEPARATE_BACKGROUND_WORKERS)s
 
+# Standard mode: Document fetching worker
 [program:celery_worker_docfetching]
 command=celery -A onyx.background.celery.versioned_apps.docfetching worker
     --loglevel=INFO
@@ -101,6 +126,7 @@ redirect_stderr=true
 autorestart=true
 startsecs=10
 stopasgroup=true
+autostart=%(ENV_USE_SEPARATE_BACKGROUND_WORKERS)s
 
 [program:celery_worker_monitoring]
 command=celery -A onyx.background.celery.versioned_apps.monitoring worker
@@ -113,6 +139,7 @@ redirect_stderr=true
 autorestart=true
 startsecs=10
 stopasgroup=true
+autostart=%(ENV_USE_SEPARATE_BACKGROUND_WORKERS)s
 
 
 # Job scheduler for periodic tasks
@@ -170,6 +197,7 @@ command=tail -qF
     /var/log/celery_beat.log
     /var/log/celery_worker_primary.log
     /var/log/celery_worker_light.log
+    /var/log/celery_worker_background.log
     /var/log/celery_worker_heavy.log
     /var/log/celery_worker_docprocessing.log
     /var/log/celery_worker_monitoring.log

backend/tests/daily/llm/test_bedrock.py

@@ -17,7 +17,7 @@ def test_bedrock_llm_configuration(client: TestClient) -> None:
     # Prepare the test request payload
     test_request: dict[str, Any] = {
         "provider": LlmProviderNames.BEDROCK,
-        "model": _DEFAULT_BEDROCK_MODEL,
+        "default_model_name": _DEFAULT_BEDROCK_MODEL,
         "api_key": None,
         "api_base": None,
         "api_version": None,
@@ -44,7 +44,7 @@ def test_bedrock_llm_configuration_invalid_key(client: TestClient) -> None:
     # Prepare the test request payload with invalid credentials
     test_request: dict[str, Any] = {
         "provider": LlmProviderNames.BEDROCK,
-        "model": _DEFAULT_BEDROCK_MODEL,
+        "default_model_name": _DEFAULT_BEDROCK_MODEL,
         "api_key": None,
         "api_base": None,
         "api_version": None,