use redis completion signal to double check exit code

Update initial seeding for latency requirements

.github/actions/custom-build-and-push/action.yml

@@ -32,16 +32,20 @@ inputs:
     description: 'Cache destinations'
     required: false
   retry-wait-time:
-    description: 'Time to wait before retry in seconds'
+    description: 'Time to wait before attempt 2 in seconds'
     required: false
-    default: '5'
+    default: '60'
+  retry-wait-time-2:
+    description: 'Time to wait before attempt 3 in seconds'
+    required: false
+    default: '120'
 
 runs:
   using: "composite"
   steps:
-    - name: Build and push Docker image (First Attempt)
+    - name: Build and push Docker image (Attempt 1 of 3)
       id: buildx1
-      uses: docker/build-push-action@v5
+      uses: docker/build-push-action@v6
       continue-on-error: true
       with:
         context: ${{ inputs.context }}
@@ -54,16 +58,17 @@ runs:
         cache-from: ${{ inputs.cache-from }}
         cache-to: ${{ inputs.cache-to }}
 
-    - name: Wait to retry
+    - name: Wait before attempt 2
       if: steps.buildx1.outcome != 'success'
       run: |
         echo "First attempt failed. Waiting ${{ inputs.retry-wait-time }} seconds before retry..."
         sleep ${{ inputs.retry-wait-time }}
       shell: bash
 
-    - name: Build and push Docker image (Retry Attempt)
+    - name: Build and push Docker image (Attempt 2 of 3)
+      id: buildx2
       if: steps.buildx1.outcome != 'success'
-      uses: docker/build-push-action@v5
+      uses: docker/build-push-action@v6
       with:
         context: ${{ inputs.context }}
         file: ${{ inputs.file }}
@@ -74,3 +79,31 @@ runs:
         tags: ${{ inputs.tags }}
         cache-from: ${{ inputs.cache-from }}
         cache-to: ${{ inputs.cache-to }}
+
+    - name: Wait before attempt 3
+      if: steps.buildx1.outcome != 'success' && steps.buildx2.outcome != 'success'
+      run: |
+        echo "Second attempt failed. Waiting ${{ inputs.retry-wait-time-2 }} seconds before retry..."
+        sleep ${{ inputs.retry-wait-time-2 }}
+      shell: bash
+
+    - name: Build and push Docker image (Attempt 3 of 3)
+      id: buildx3
+      if: steps.buildx1.outcome != 'success' && steps.buildx2.outcome != 'success'
+      uses: docker/build-push-action@v6
+      with:
+        context: ${{ inputs.context }}
+        file: ${{ inputs.file }}
+        platforms: ${{ inputs.platforms }}
+        pull: ${{ inputs.pull }}
+        push: ${{ inputs.push }}
+        load: ${{ inputs.load }}
+        tags: ${{ inputs.tags }}
+        cache-from: ${{ inputs.cache-from }}
+        cache-to: ${{ inputs.cache-to }}
+
+    - name: Report failure
+      if: steps.buildx1.outcome != 'success' && steps.buildx2.outcome != 'success' && steps.buildx3.outcome != 'success'
+      run: |
+        echo "All attempts failed. Possible transient infrastucture issues? Try again later or inspect logs for details."
+      shell: bash

.github/pull_request_template.md

@@ -6,20 +6,24 @@
 [Describe the tests you ran to verify your changes]
 
 
-## Accepted Risk
-[Any know risks or failure modes to point out to reviewers]
+## Accepted Risk (provide if relevant)
+N/A
 
 
-## Related Issue(s)
-[If applicable, link to the issue(s) this PR addresses]
+## Related Issue(s) (provide if relevant)
+N/A
 
 
-## Checklist:
-- [ ] All of the automated tests pass
-- [ ] All PR comments are addressed and marked resolved
-- [ ] If there are migrations, they have been rebased to latest main
-- [ ] If there are new dependencies, they are added to the requirements
-- [ ] If there are new environment variables, they are added to all of the deployment methods
-- [ ] If there are new APIs that don't require auth, they are added to PUBLIC_ENDPOINT_SPECS
-- [ ] Docker images build and basic functionalities work
-- [ ] Author has done a final read through of the PR right before merge
+## Mental Checklist:
+- All of the automated tests pass
+- All PR comments are addressed and marked resolved
+- If there are migrations, they have been rebased to latest main
+- If there are new dependencies, they are added to the requirements
+- If there are new environment variables, they are added to all of the deployment methods
+- If there are new APIs that don't require auth, they are added to PUBLIC_ENDPOINT_SPECS
+- Docker images build and basic functionalities work
+- Author has done a final read through of the PR right before merge
+
+## Backporting (check the box to trigger backport action)
+Note: You have to check that the action passes, otherwise resolve the conflicts manually and tag the patches.
+- [ ] This PR should be backported (make sure to check that the backport attempt succeeds)

.github/workflows/check-backend-changes.yml

@@ -1,23 +0,0 @@
-name: Check Backend Changes
-
-on:
-  workflow_call:
-    outputs:
-      run-tests:
-        description: "Whether to run tests based on backend changes"
-        value: ${{ jobs.check-run-needed.outputs.run-tests }}
-
-jobs:
-  check-run-needed:
-    runs-on: ubuntu-latest
-    outputs:
-      run-tests: ${{ steps.check.outputs.run-tests }}
-    steps:
-      - uses: actions/checkout@v4
-      - id: check
-        run: |
-          if git diff --name-only ${{ github.event.before }} ${{ github.sha }} | grep -q '^backend/'; then
-            echo "run-tests=true" >> $GITHUB_OUTPUT
-          else
-            echo "run-tests=false" >> $GITHUB_OUTPUT
-          fi

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

@@ -3,47 +3,61 @@ name: Build and Push Backend Image on Tag
 on:
   push:
     tags:
-      - '*'
+      - "*"
 
 env:
-  REGISTRY_IMAGE: danswer/danswer-backend
+  REGISTRY_IMAGE: ${{ contains(github.ref_name, 'cloud') && 'danswer/danswer-backend-cloud' || 'danswer/danswer-backend' }}
+  LATEST_TAG: ${{ contains(github.ref_name, 'latest') }}
 
 jobs:
   build-and-push:
-    # TODO: make this a matrix build like the web containers
-    runs-on: 
-      group: amd64-image-builders
+    # TODO: investigate a matrix build like the web container
+    # See https://runs-on.com/runners/linux/
+    runs-on: [runs-on, runner=8cpu-linux-x64, "run-id=${{ github.run_id }}"]
 
     steps:
-    - name: Checkout code
-      uses: actions/checkout@v2
+      - name: Checkout code
+        uses: actions/checkout@v4
 
-    - name: Set up Docker Buildx
-      uses: docker/setup-buildx-action@v3
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
 
-    - name: Login to Docker Hub
-      uses: docker/login-action@v3
-      with:
-        username: ${{ secrets.DOCKER_USERNAME }}
-        password: ${{ secrets.DOCKER_TOKEN }}
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          username: ${{ secrets.DOCKER_USERNAME }}
+          password: ${{ secrets.DOCKER_TOKEN }}
 
-    - name: Backend Image Docker Build and Push
-      uses: docker/build-push-action@v5
-      with:
-        context: ./backend
-        file: ./backend/Dockerfile
-        platforms: linux/amd64,linux/arm64
-        push: true
-        tags: |
-          ${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}
-          ${{ env.REGISTRY_IMAGE }}:latest
-        build-args: |
-          DANSWER_VERSION=${{ github.ref_name }}
+      - name: Install build-essential
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y build-essential
 
-    - name: Run Trivy vulnerability scanner
-      uses: aquasecurity/trivy-action@master
-      with:
-        # To run locally: trivy image --severity HIGH,CRITICAL danswer/danswer-backend
-        image-ref: docker.io/${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}
-        severity: 'CRITICAL,HIGH'
-        trivyignores: ./backend/.trivyignore
+      - name: Backend Image Docker Build and Push
+        uses: docker/build-push-action@v5
+        with:
+          context: ./backend
+          file: ./backend/Dockerfile
+          platforms: linux/amd64,linux/arm64
+          push: true
+          tags: |
+            ${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}
+            ${{ env.LATEST_TAG == 'true' && format('{0}:latest', env.REGISTRY_IMAGE) || '' }}
+          build-args: |
+            DANSWER_VERSION=${{ github.ref_name }}
+
+      # trivy has their own rate limiting issues causing this action to flake
+      # we worked around it by hardcoding to different db repos in env
+      # can re-enable when they figure it out
+      # https://github.com/aquasecurity/trivy/discussions/7538
+      # https://github.com/aquasecurity/trivy-action/issues/389
+      - name: Run Trivy vulnerability scanner
+        uses: aquasecurity/trivy-action@master
+        env:
+          TRIVY_DB_REPOSITORY: "public.ecr.aws/aquasecurity/trivy-db:2"
+          TRIVY_JAVA_DB_REPOSITORY: "public.ecr.aws/aquasecurity/trivy-java-db:1"
+        with:
+          # To run locally: trivy image --severity HIGH,CRITICAL danswer/danswer-backend
+          image-ref: docker.io/${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}
+          severity: "CRITICAL,HIGH"
+          trivyignores: ./backend/.trivyignore

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

@@ -0,0 +1,137 @@
+name: Build and Push Cloud Web Image on Tag
+# Identical to the web container build, but with correct image tag and build args
+
+on:
+  push:
+    tags:
+      - "*"
+
+env:
+  REGISTRY_IMAGE: danswer/danswer-web-server-cloud
+  LATEST_TAG: ${{ contains(github.ref_name, 'latest') }}
+
+jobs:
+  build:
+    runs-on:
+      - runs-on
+      - runner=${{ matrix.platform == 'linux/amd64' && '8cpu-linux-x64' || '8cpu-linux-arm64' }}
+      - run-id=${{ github.run_id }}
+      - tag=platform-${{ matrix.platform }}
+    strategy:
+      fail-fast: false
+      matrix:
+        platform:
+          - linux/amd64
+          - linux/arm64
+
+    steps:
+      - name: Prepare
+        run: |
+          platform=${{ matrix.platform }}
+          echo "PLATFORM_PAIR=${platform//\//-}" >> $GITHUB_ENV
+
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Docker meta
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          images: ${{ env.REGISTRY_IMAGE }}
+          tags: |
+            type=raw,value=${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}
+            type=raw,value=${{ env.LATEST_TAG == 'true' && format('{0}:latest', env.REGISTRY_IMAGE) || '' }}
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          username: ${{ secrets.DOCKER_USERNAME }}
+          password: ${{ secrets.DOCKER_TOKEN }}
+
+      - name: Build and push by digest
+        id: build
+        uses: docker/build-push-action@v5
+        with:
+          context: ./web
+          file: ./web/Dockerfile
+          platforms: ${{ matrix.platform }}
+          push: true
+          build-args: |
+            DANSWER_VERSION=${{ github.ref_name }}
+            NEXT_PUBLIC_CLOUD_ENABLED=true
+            NEXT_PUBLIC_POSTHOG_KEY=${{ secrets.POSTHOG_KEY }}
+            NEXT_PUBLIC_POSTHOG_HOST=${{ secrets.POSTHOG_HOST }}
+            NEXT_PUBLIC_SENTRY_DSN=${{ secrets.SENTRY_DSN }}
+            NEXT_PUBLIC_GTM_ENABLED=true
+          # needed due to weird interactions with the builds for different platforms
+          no-cache: true
+          labels: ${{ steps.meta.outputs.labels }}
+          outputs: type=image,name=${{ env.REGISTRY_IMAGE }},push-by-digest=true,name-canonical=true,push=true
+
+      - name: Export digest
+        run: |
+          mkdir -p /tmp/digests
+          digest="${{ steps.build.outputs.digest }}"
+          touch "/tmp/digests/${digest#sha256:}"
+
+      - name: Upload digest
+        uses: actions/upload-artifact@v4
+        with:
+          name: digests-${{ env.PLATFORM_PAIR }}
+          path: /tmp/digests/*
+          if-no-files-found: error
+          retention-days: 1
+
+  merge:
+    runs-on: ubuntu-latest
+    needs:
+      - build
+    steps:
+      - name: Download digests
+        uses: actions/download-artifact@v4
+        with:
+          path: /tmp/digests
+          pattern: digests-*
+          merge-multiple: true
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Docker meta
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          images: ${{ env.REGISTRY_IMAGE }}
+
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          username: ${{ secrets.DOCKER_USERNAME }}
+          password: ${{ secrets.DOCKER_TOKEN }}
+
+      - name: Create manifest list and push
+        working-directory: /tmp/digests
+        run: |
+          docker buildx imagetools create $(jq -cr '.tags | map("-t " + .) | join(" ")' <<< "$DOCKER_METADATA_OUTPUT_JSON") \
+            $(printf '${{ env.REGISTRY_IMAGE }}@sha256:%s ' *)
+
+      - name: Inspect image
+        run: |
+          docker buildx imagetools inspect ${{ env.REGISTRY_IMAGE }}:${{ steps.meta.outputs.version }}
+
+      # trivy has their own rate limiting issues causing this action to flake
+      # we worked around it by hardcoding to different db repos in env
+      # can re-enable when they figure it out
+      # https://github.com/aquasecurity/trivy/discussions/7538
+      # https://github.com/aquasecurity/trivy-action/issues/389
+      - name: Run Trivy vulnerability scanner
+        uses: aquasecurity/trivy-action@master
+        env:
+          TRIVY_DB_REPOSITORY: "public.ecr.aws/aquasecurity/trivy-db:2"
+          TRIVY_JAVA_DB_REPOSITORY: "public.ecr.aws/aquasecurity/trivy-java-db:1"
+        with:
+          image-ref: docker.io/${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}
+          severity: "CRITICAL,HIGH"

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

@@ -3,41 +3,53 @@ name: Build and Push Model Server Image on Tag
 on:
   push:
     tags:
-      - '*'
+      - "*"
+
+env:
+  REGISTRY_IMAGE: ${{ contains(github.ref_name, 'cloud') && 'danswer/danswer-model-server-cloud' || 'danswer/danswer-model-server' }}
+  LATEST_TAG: ${{ contains(github.ref_name, 'latest') }}
 
 jobs:
   build-and-push:
-    runs-on: 
-      group: amd64-image-builders
+    # See https://runs-on.com/runners/linux/
+    runs-on: [runs-on, runner=8cpu-linux-x64, "run-id=${{ github.run_id }}"]
 
     steps:
-    - name: Checkout code
-      uses: actions/checkout@v2
+      - name: Checkout code
+        uses: actions/checkout@v4
 
-    - name: Set up Docker Buildx
-      uses: docker/setup-buildx-action@v3
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
 
-    - name: Login to Docker Hub
-      uses: docker/login-action@v3
-      with:
-        username: ${{ secrets.DOCKER_USERNAME }}
-        password: ${{ secrets.DOCKER_TOKEN }}
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          username: ${{ secrets.DOCKER_USERNAME }}
+          password: ${{ secrets.DOCKER_TOKEN }}
 
-    - name: Model Server Image Docker Build and Push
-      uses: docker/build-push-action@v5
-      with:
-        context: ./backend
-        file: ./backend/Dockerfile.model_server
-        platforms: linux/amd64,linux/arm64
-        push: true
-        tags: |
-          danswer/danswer-model-server:${{ github.ref_name }}
-          danswer/danswer-model-server:latest
-        build-args: |
-          DANSWER_VERSION=${{ github.ref_name }}
+      - name: Model Server Image Docker Build and Push
+        uses: docker/build-push-action@v5
+        with:
+          context: ./backend
+          file: ./backend/Dockerfile.model_server
+          platforms: linux/amd64,linux/arm64
+          push: true
+          tags: |
+            ${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}
+            ${{ env.LATEST_TAG == 'true' && format('{0}:latest', env.REGISTRY_IMAGE) || '' }}
+          build-args: |
+            DANSWER_VERSION=${{ github.ref_name }}
 
-    - name: Run Trivy vulnerability scanner
-      uses: aquasecurity/trivy-action@master
-      with:
-        image-ref: docker.io/danswer/danswer-model-server:${{ github.ref_name }}
-        severity: 'CRITICAL,HIGH'
+      # trivy has their own rate limiting issues causing this action to flake
+      # we worked around it by hardcoding to different db repos in env
+      # can re-enable when they figure it out
+      # https://github.com/aquasecurity/trivy/discussions/7538
+      # https://github.com/aquasecurity/trivy-action/issues/389
+      - name: Run Trivy vulnerability scanner
+        uses: aquasecurity/trivy-action@master
+        env:
+          TRIVY_DB_REPOSITORY: "public.ecr.aws/aquasecurity/trivy-db:2"
+          TRIVY_JAVA_DB_REPOSITORY: "public.ecr.aws/aquasecurity/trivy-java-db:1"
+        with:
+          image-ref: docker.io/danswer/danswer-model-server:${{ github.ref_name }}
+          severity: "CRITICAL,HIGH"

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

@@ -7,11 +7,15 @@ on:
 
 env:
   REGISTRY_IMAGE: danswer/danswer-web-server
-
+  LATEST_TAG: ${{ contains(github.ref_name, 'latest') }}
+  
 jobs:
   build:
-    runs-on: 
-      group: ${{ matrix.platform == 'linux/amd64' && 'amd64-image-builders' || 'arm64-image-builders' }}
+    runs-on:
+      - runs-on
+      - runner=${{ matrix.platform == 'linux/amd64' && '8cpu-linux-x64' || '8cpu-linux-arm64' }}
+      - run-id=${{ github.run_id }}
+      - tag=platform-${{ matrix.platform }}
     strategy:
       fail-fast: false
       matrix:
@@ -35,7 +39,7 @@ jobs:
           images: ${{ env.REGISTRY_IMAGE }}
           tags: |
             type=raw,value=${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}
-            type=raw,value=${{ env.REGISTRY_IMAGE }}:latest
+            type=raw,value=${{ env.LATEST_TAG == 'true' && format('{0}:latest', env.REGISTRY_IMAGE) || '' }}
       
       - name: Set up Docker Buildx
         uses: docker/setup-buildx-action@v3
@@ -112,8 +116,16 @@ jobs:
         run: |
           docker buildx imagetools inspect ${{ env.REGISTRY_IMAGE }}:${{ steps.meta.outputs.version }}
 
+    # trivy has their own rate limiting issues causing this action to flake
+    # we worked around it by hardcoding to different db repos in env
+    # can re-enable when they figure it out
+    # https://github.com/aquasecurity/trivy/discussions/7538
+    # https://github.com/aquasecurity/trivy-action/issues/389
       - name: Run Trivy vulnerability scanner
         uses: aquasecurity/trivy-action@master
+        env:
+          TRIVY_DB_REPOSITORY: 'public.ecr.aws/aquasecurity/trivy-db:2'
+          TRIVY_JAVA_DB_REPOSITORY: 'public.ecr.aws/aquasecurity/trivy-java-db:1'
         with:
           image-ref: docker.io/${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}
           severity: 'CRITICAL,HIGH'

.github/workflows/docker-tag-latest.yml

@@ -1,3 +1,6 @@
+# This workflow is set up to be manually triggered via the GitHub Action tab.
+# Given a version, it will tag those backend and webserver images as "latest".
+
 name: Tag Latest Version
 
 on:
@@ -9,7 +12,9 @@ on:
 
 jobs:
   tag:
-    runs-on: ubuntu-latest
+    # See https://runs-on.com/runners/linux/
+    # use a lower powered instance since this just does i/o to docker hub
+    runs-on: [runs-on,runner=2cpu-linux-x64,"run-id=${{ github.run_id }}"]
     steps:
     - name: Set up Docker Buildx
       uses: docker/setup-buildx-action@v1

.github/workflows/hotfix-release-branches.yml

@@ -0,0 +1,172 @@
+# This workflow is intended to be manually triggered via the GitHub Action tab.
+# Given a hotfix branch, it will attempt to open a PR to all release branches and
+# by default auto merge them
+
+name: Hotfix release branches
+
+on:
+  workflow_dispatch:
+    inputs:
+      hotfix_commit:
+        description: 'Hotfix commit hash'
+        required: true
+      hotfix_suffix:
+        description: 'Hotfix branch suffix (e.g. hotfix/v0.8-{suffix})'
+        required: true
+      release_branch_pattern:
+        description: 'Release branch pattern (regex)'
+        required: true
+        default: 'release/.*'
+      auto_merge:
+        description: 'Automatically merge the hotfix PRs'
+        required: true
+        type: choice
+        default: 'true'
+        options:
+          - true
+          - false
+          
+jobs:
+  hotfix_release_branches:
+    permissions: write-all
+    # See https://runs-on.com/runners/linux/
+    # use a lower powered instance since this just does i/o to docker hub
+    runs-on: [runs-on,runner=2cpu-linux-x64,"run-id=${{ github.run_id }}"]
+    steps:
+    
+      # needs RKUO_DEPLOY_KEY for write access to merge PR's
+      - name: Checkout Repository
+        uses: actions/checkout@v4
+        with:
+          ssh-key: "${{ secrets.RKUO_DEPLOY_KEY }}"
+          fetch-depth: 0
+          
+      - name: Set up Git user
+        run: |
+          git config user.name "Richard Kuo [bot]"
+          git config user.email "rkuo[bot]@danswer.ai"
+
+      - name: Fetch All Branches
+        run: |
+          git fetch --all --prune
+
+      - name: Verify Hotfix Commit Exists
+        run: |
+          git rev-parse --verify "${{ github.event.inputs.hotfix_commit }}" || { echo "Commit not found: ${{ github.event.inputs.hotfix_commit }}"; exit 1; }
+
+      - name: Get Release Branches
+        id: get_release_branches
+        run: |
+          BRANCHES=$(git branch -r | grep -E "${{ github.event.inputs.release_branch_pattern }}" | sed 's|origin/||' | tr -d ' ')
+          if [ -z "$BRANCHES" ]; then
+            echo "No release branches found matching pattern '${{ github.event.inputs.release_branch_pattern }}'."
+            exit 1
+          fi
+          
+          echo "Found release branches:"
+          echo "$BRANCHES"
+          
+          # Join the branches into a single line separated by commas
+          BRANCHES_JOINED=$(echo "$BRANCHES" | tr '\n' ',' | sed 's/,$//')
+
+          # Set the branches as an output
+          echo "branches=$BRANCHES_JOINED" >> $GITHUB_OUTPUT
+
+      # notes on all the vagaries of wiring up automated PR's
+      # https://github.com/peter-evans/create-pull-request/blob/main/docs/concepts-guidelines.md#triggering-further-workflow-runs
+      # we must use a custom token for GH_TOKEN to trigger the subsequent PR checks
+      - name: Create and Merge Pull Requests to Matching Release Branches
+        env:
+          HOTFIX_COMMIT: ${{ github.event.inputs.hotfix_commit }}
+          HOTFIX_SUFFIX: ${{ github.event.inputs.hotfix_suffix }}
+          AUTO_MERGE: ${{ github.event.inputs.auto_merge }}
+          GH_TOKEN: ${{ secrets.RKUO_PERSONAL_ACCESS_TOKEN }}
+        run: |
+          # Get the branches from the previous step
+          BRANCHES="${{ steps.get_release_branches.outputs.branches }}"
+
+          # Convert BRANCHES to an array
+          IFS=$',' read -ra BRANCH_ARRAY <<< "$BRANCHES"
+
+          # Loop through each release branch and create and merge a PR
+          for RELEASE_BRANCH in "${BRANCH_ARRAY[@]}"; do
+            echo "Processing $RELEASE_BRANCH..."
+            
+            # Parse out the release version by removing "release/" from the branch name
+            RELEASE_VERSION=${RELEASE_BRANCH#release/}
+            echo "Release version parsed: $RELEASE_VERSION"
+            
+            HOTFIX_BRANCH="hotfix/${RELEASE_VERSION}-${HOTFIX_SUFFIX}"
+            echo "Creating PR from $HOTFIX_BRANCH to $RELEASE_BRANCH"
+
+            # Checkout the release branch
+            echo "Checking out $RELEASE_BRANCH"
+            git checkout "$RELEASE_BRANCH"
+
+            # Create the new hotfix branch
+            if git rev-parse --verify "$HOTFIX_BRANCH" >/dev/null 2>&1; then
+              echo "Hotfix branch $HOTFIX_BRANCH already exists. Skipping branch creation."
+            else
+              echo "Branching $RELEASE_BRANCH to $HOTFIX_BRANCH"
+              git checkout -b "$HOTFIX_BRANCH"
+            fi
+            
+            # Check if the hotfix commit is a merge commit
+            if git rev-list --merges -n 1 "$HOTFIX_COMMIT" >/dev/null 2>&1; then
+              # -m 1 uses the target branch as the base (which is what we want)
+              echo "Hotfix commit $HOTFIX_COMMIT is a merge commit, using -m 1 for cherry-pick"
+              CHERRY_PICK_CMD="git cherry-pick -m 1 $HOTFIX_COMMIT"
+            else
+              CHERRY_PICK_CMD="git cherry-pick $HOTFIX_COMMIT"
+            fi
+
+            # Perform the cherry-pick
+            echo "Executing: $CHERRY_PICK_CMD"
+            eval "$CHERRY_PICK_CMD"
+
+            if [ $? -ne 0 ]; then
+              echo "Cherry-pick failed for $HOTFIX_COMMIT on $HOTFIX_BRANCH. Aborting..."
+              git cherry-pick --abort
+              continue
+            fi
+
+            # Push the hotfix branch to the remote
+            echo "Pushing $HOTFIX_BRANCH..."
+            git push origin "$HOTFIX_BRANCH"
+            echo "Hotfix branch $HOTFIX_BRANCH created and pushed."
+            
+            # Check if PR already exists
+            EXISTING_PR=$(gh pr list --head "$HOTFIX_BRANCH" --base "$RELEASE_BRANCH" --state open --json number --jq '.[0].number')
+
+            if [ -n "$EXISTING_PR" ]; then
+              echo "An open PR already exists: #$EXISTING_PR. Skipping..."
+              continue
+            fi
+            
+            # Create a new PR and capture the output
+            PR_OUTPUT=$(gh pr create --title "Merge $HOTFIX_BRANCH into $RELEASE_BRANCH" \
+              --body "Automated PR to merge \`$HOTFIX_BRANCH\` into \`$RELEASE_BRANCH\`." \
+              --head "$HOTFIX_BRANCH" --base "$RELEASE_BRANCH")
+
+            # Extract the URL from the output
+            PR_URL=$(echo "$PR_OUTPUT" | grep -Eo 'https://github.com/[^ ]+')
+            echo "Pull request created: $PR_URL"
+
+            # Extract PR number from URL
+            PR_NUMBER=$(basename "$PR_URL")
+            echo "Pull request created: $PR_NUMBER"
+
+            if [ "$AUTO_MERGE" == "true" ]; then
+              echo "Attempting to merge pull request #$PR_NUMBER"
+
+              # Attempt to merge the PR
+              gh pr merge "$PR_NUMBER" --merge --auto --delete-branch
+
+              if [ $? -eq 0 ]; then
+                echo "Pull request #$PR_NUMBER merged successfully."
+              else
+                # Optionally, handle the error or continue
+                echo "Failed to merge pull request #$PR_NUMBER."
+              fi
+            fi
+          done

.github/workflows/nightly-close-stale-issues.yml

@@ -0,0 +1,23 @@
+name: 'Nightly - Close stale issues and PRs'
+on:
+  schedule:
+    - cron: '0 11 * * *' # Runs every day at 3 AM PST / 4 AM PDT / 11 AM UTC
+
+permissions:
+  # contents: write # only for delete-branch option
+  issues: write
+  pull-requests: write
+  
+jobs:
+  stale:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/stale@v9
+        with:
+          stale-issue-message: 'This issue is stale because it has been open 75 days with no activity. Remove stale label or comment or this will be closed in 15 days.'
+          stale-pr-message: 'This PR is stale because it has been open 75 days with no activity. Remove stale label or comment or this will be closed in 15 days.'
+          close-issue-message: 'This issue was closed because it has been stalled for 90 days with no activity.'
+          close-pr-message: 'This PR was closed because it has been stalled for 90 days with no activity.'
+          days-before-stale: 75
+#           days-before-close: 90  # uncomment after we test stale behavior
+          

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

@@ -0,0 +1,76 @@
+# Scan for problematic software licenses
+
+# trivy has their own rate limiting issues causing this action to flake
+# we worked around it by hardcoding to different db repos in env
+# can re-enable when they figure it out
+# https://github.com/aquasecurity/trivy/discussions/7538
+# https://github.com/aquasecurity/trivy-action/issues/389
+
+name: 'Nightly - Scan licenses'
+on:
+#   schedule:
+#     - cron: '0 14 * * *'  # Runs every day at 6 AM PST / 7 AM PDT / 2 PM UTC
+  workflow_dispatch:  # Allows manual triggering
+
+permissions:
+  actions: read
+  contents: read
+  security-events: write
+  
+jobs:
+  scan-licenses:
+    # See https://runs-on.com/runners/linux/
+    runs-on: [runs-on,runner=2cpu-linux-x64,"run-id=${{ github.run_id }}"]
+
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+        
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.11'
+          cache: 'pip'
+          cache-dependency-path: |
+            backend/requirements/default.txt
+            backend/requirements/dev.txt
+            backend/requirements/model_server.txt
+      
+      - name: Get explicit and transitive dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install --retries 5 --timeout 30 -r backend/requirements/default.txt
+          pip install --retries 5 --timeout 30 -r backend/requirements/dev.txt
+          pip install --retries 5 --timeout 30 -r backend/requirements/model_server.txt
+          pip freeze > requirements-all.txt
+                    
+      - name: Check python
+        id: license_check_report
+        uses: pilosus/action-pip-license-checker@v2
+        with:
+          requirements: 'requirements-all.txt'
+          fail: 'Copyleft'
+          exclude: '(?i)^(pylint|aio[-_]*).*'
+          
+      - name: Print report
+        if: ${{ always() }}
+        run: echo "${{ steps.license_check_report.outputs.report }}"
+      
+      - name: Install npm dependencies
+        working-directory: ./web
+        run: npm ci
+        
+      - name: Run Trivy vulnerability scanner in repo mode
+        uses: aquasecurity/trivy-action@0.28.0
+        with:
+          scan-type: fs
+          scanners: license
+          format: table
+#           format: sarif
+#           output: trivy-results.sarif
+          severity: HIGH,CRITICAL
+
+#       - name: Upload Trivy scan results to GitHub Security tab
+#         uses: github/codeql-action/upload-sarif@v3
+#         with:
+#           sarif_file: trivy-results.sarif

.github/workflows/pr-backport-autotrigger.yml

@@ -0,0 +1,124 @@
+name: Backport on Merge
+
+# Note this workflow does not trigger the builds, be sure to manually tag the branches to trigger the builds
+
+on:
+  pull_request:
+    types: [closed]  # Later we check for merge so only PRs that go in can get backported
+
+permissions:
+  contents: write
+  actions: write
+
+jobs:
+  backport:
+    if: github.event.pull_request.merged == true
+    runs-on: ubuntu-latest
+    env:
+      GITHUB_TOKEN: ${{ secrets.YUHONG_GH_ACTIONS }}
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+        with:
+          ssh-key: "${{ secrets.RKUO_DEPLOY_KEY }}"
+          fetch-depth: 0
+
+      - name: Set up Git user
+        run: |
+          git config user.name "Richard Kuo [bot]"
+          git config user.email "rkuo[bot]@danswer.ai"
+          git fetch --prune
+      
+      - name: Check for Backport Checkbox
+        id: checkbox-check
+        run: |
+          PR_BODY="${{ github.event.pull_request.body }}"
+          if [[ "$PR_BODY" == *"[x] This PR should be backported"* ]]; then
+            echo "backport=true" >> $GITHUB_OUTPUT
+          else
+            echo "backport=false" >> $GITHUB_OUTPUT
+          fi
+
+      - name: List and sort release branches
+        id: list-branches
+        run: |
+          git fetch --all --tags
+          BRANCHES=$(git for-each-ref --format='%(refname:short)' refs/remotes/origin/release/* | sed 's|origin/release/||' | sort -Vr)
+          BETA=$(echo "$BRANCHES" | head -n 1)
+          STABLE=$(echo "$BRANCHES" | head -n 2 | tail -n 1)
+          echo "beta=release/$BETA" >> $GITHUB_OUTPUT
+          echo "stable=release/$STABLE" >> $GITHUB_OUTPUT
+          # Fetch latest tags for beta and stable
+          LATEST_BETA_TAG=$(git tag -l "v[0-9]*.[0-9]*.[0-9]*-beta.[0-9]*" | grep -E "^v[0-9]+\.[0-9]+\.[0-9]+-beta\.[0-9]+$" | grep -v -- "-cloud" | sort -Vr | head -n 1)
+          LATEST_STABLE_TAG=$(git tag -l "v[0-9]*.[0-9]*.[0-9]*" | grep -E "^v[0-9]+\.[0-9]+\.[0-9]+$" | sort -Vr | head -n 1)
+          
+          # Handle case where no beta tags exist
+          if [[ -z "$LATEST_BETA_TAG" ]]; then
+            NEW_BETA_TAG="v1.0.0-beta.1"
+          else
+            NEW_BETA_TAG=$(echo $LATEST_BETA_TAG | awk -F '[.-]' '{print $1 "." $2 "." $3 "-beta." ($NF+1)}')
+          fi
+          
+          # Increment latest stable tag
+          NEW_STABLE_TAG=$(echo $LATEST_STABLE_TAG | awk -F '.' '{print $1 "." $2 "." ($3+1)}')
+          echo "latest_beta_tag=$LATEST_BETA_TAG" >> $GITHUB_OUTPUT
+          echo "latest_stable_tag=$LATEST_STABLE_TAG" >> $GITHUB_OUTPUT
+          echo "new_beta_tag=$NEW_BETA_TAG" >> $GITHUB_OUTPUT
+          echo "new_stable_tag=$NEW_STABLE_TAG" >> $GITHUB_OUTPUT
+
+      - name: Echo branch and tag information
+        run: |
+          echo "Beta branch: ${{ steps.list-branches.outputs.beta }}"
+          echo "Stable branch: ${{ steps.list-branches.outputs.stable }}"
+          echo "Latest beta tag: ${{ steps.list-branches.outputs.latest_beta_tag }}"
+          echo "Latest stable tag: ${{ steps.list-branches.outputs.latest_stable_tag }}"
+          echo "New beta tag: ${{ steps.list-branches.outputs.new_beta_tag }}"
+          echo "New stable tag: ${{ steps.list-branches.outputs.new_stable_tag }}"
+
+      - name: Trigger Backport
+        if: steps.checkbox-check.outputs.backport == 'true'
+        run: |
+          set -e
+          echo "Backporting to beta ${{ steps.list-branches.outputs.beta }} and stable ${{ steps.list-branches.outputs.stable }}"
+          
+          # Echo the merge commit SHA
+          echo "Merge commit SHA: ${{ github.event.pull_request.merge_commit_sha }}"
+          
+          # Fetch all history for all branches and tags
+          git fetch --prune
+
+          # Reset and prepare the beta branch
+          git checkout ${{ steps.list-branches.outputs.beta }}
+          echo "Last 5 commits on beta branch:"
+          git log -n 5 --pretty=format:"%H"
+          echo ""  # Newline for formatting
+
+          # Cherry-pick the merge commit from the merged PR
+          git cherry-pick -m 1 ${{ github.event.pull_request.merge_commit_sha }} || {
+            echo "Cherry-pick to beta failed due to conflicts."
+            exit 1
+          }
+          
+          # Create new beta branch/tag
+          git tag ${{ steps.list-branches.outputs.new_beta_tag }}
+          # Push the changes and tag to the beta branch using PAT
+          git push origin ${{ steps.list-branches.outputs.beta }}
+          git push origin ${{ steps.list-branches.outputs.new_beta_tag }}
+
+          # Reset and prepare the stable branch
+          git checkout ${{ steps.list-branches.outputs.stable }}
+          echo "Last 5 commits on stable branch:"
+          git log -n 5 --pretty=format:"%H"
+          echo ""  # Newline for formatting
+          
+          # Cherry-pick the merge commit from the merged PR
+          git cherry-pick -m 1 ${{ github.event.pull_request.merge_commit_sha }} || {
+            echo "Cherry-pick to stable failed due to conflicts."
+            exit 1
+          }
+          
+          # Create new stable branch/tag
+          git tag ${{ steps.list-branches.outputs.new_stable_tag }}
+          # Push the changes and tag to the stable branch using PAT
+          git push origin ${{ steps.list-branches.outputs.stable }}
+          git push origin ${{ steps.list-branches.outputs.new_stable_tag }}

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

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

.github/workflows/pr-helm-chart-testing.yml

@@ -0,0 +1,72 @@
+name: Helm - Lint and Test Charts
+
+on:
+  merge_group:
+  pull_request:
+    branches: [ main ]
+  workflow_dispatch:  # Allows manual triggering
+  
+jobs:
+  helm-chart-check:
+    # See https://runs-on.com/runners/linux/
+    runs-on: [runs-on,runner=8cpu-linux-x64,hdd=256,"run-id=${{ github.run_id }}"]
+
+    # fetch-depth 0 is required for helm/chart-testing-action
+    steps:
+    - name: Checkout code
+      uses: actions/checkout@v4
+      with:
+        fetch-depth: 0
+        
+    - name: Set up Helm
+      uses: azure/setup-helm@v4.2.0
+      with:
+        version: v3.14.4
+      
+    - name: Set up chart-testing
+      uses: helm/chart-testing-action@v2.6.1
+
+    # even though we specify chart-dirs in ct.yaml, it isn't used by ct for the list-changed command...
+    - name: Run chart-testing (list-changed)
+      id: list-changed
+      run: |
+        echo "default_branch: ${{ github.event.repository.default_branch }}"
+        changed=$(ct list-changed --remote origin --target-branch ${{ github.event.repository.default_branch }} --chart-dirs deployment/helm/charts)
+        echo "list-changed output: $changed"
+        if [[ -n "$changed" ]]; then
+          echo "changed=true" >> "$GITHUB_OUTPUT"
+        fi
+
+#     rkuo: I don't think we need python?
+#     - name: Set up Python
+#       uses: actions/setup-python@v5
+#       with:
+#         python-version: '3.11'
+#         cache: 'pip'
+#         cache-dependency-path: |
+#           backend/requirements/default.txt
+#           backend/requirements/dev.txt
+#           backend/requirements/model_server.txt
+#     - run: |
+#         python -m pip install --upgrade pip
+#         pip install --retries 5 --timeout 30 -r backend/requirements/default.txt
+#         pip install --retries 5 --timeout 30 -r backend/requirements/dev.txt
+#         pip install --retries 5 --timeout 30 -r backend/requirements/model_server.txt
+
+    # lint all charts if any changes were detected
+    - name: Run chart-testing (lint)
+      if: steps.list-changed.outputs.changed == 'true'
+      run: ct lint --config ct.yaml --all
+      # the following would lint only changed charts, but linting isn't expensive
+      # run: ct lint --config ct.yaml --target-branch ${{ github.event.repository.default_branch }}
+
+    - name: Create kind cluster
+      if: steps.list-changed.outputs.changed == 'true'
+      uses: helm/kind-action@v1.10.0
+
+    - name: Run chart-testing (install)
+      if: steps.list-changed.outputs.changed == 'true'
+      run: ct install --all --helm-extra-set-args="--set=nginx.enabled=false" --debug --config ct.yaml
+      # the following would install only changed charts, but we only have one chart so 
+      # don't worry about that for now
+      # run: ct install --target-branch ${{ github.event.repository.default_branch }}

.github/workflows/pr-helm-chart-testing.yml.disabled.txt

@@ -1,67 +0,0 @@
-# This workflow is intentionally disabled while we're still working on it
-# It's close to ready, but a race condition needs to be fixed with
-# API server and Vespa startup, and it needs to have a way to build/test against
-# local containers
-
-name: Helm - Lint and Test Charts
-
-on:
-  merge_group:
-  pull_request:
-    branches: [ main ]
-
-jobs:
-  lint-test:
-    runs-on: Amd64
-
-    # fetch-depth 0 is required for helm/chart-testing-action
-    steps:
-    - name: Checkout code
-      uses: actions/checkout@v3
-      with:
-        fetch-depth: 0
-        
-    - name: Set up Helm
-      uses: azure/setup-helm@v4.2.0
-      with:
-        version: v3.14.4
-      
-    - name: Set up Python
-      uses: actions/setup-python@v4
-      with:
-        python-version: '3.11'
-        cache: 'pip'
-        cache-dependency-path: |
-          backend/requirements/default.txt
-          backend/requirements/dev.txt
-          backend/requirements/model_server.txt
-    - run: |
-        python -m pip install --upgrade pip
-        pip install -r backend/requirements/default.txt
-        pip install -r backend/requirements/dev.txt
-        pip install -r backend/requirements/model_server.txt
-
-    - name: Set up chart-testing
-      uses: helm/chart-testing-action@v2.6.1
-
-    - name: Run chart-testing (list-changed)
-      id: list-changed
-      run: |
-        changed=$(ct list-changed --target-branch ${{ github.event.repository.default_branch }})
-        if [[ -n "$changed" ]]; then
-          echo "changed=true" >> "$GITHUB_OUTPUT"
-        fi
-
-    - name: Run chart-testing (lint)
-#       if: steps.list-changed.outputs.changed == 'true'
-      run: ct lint --all --config ct.yaml --target-branch ${{ github.event.repository.default_branch }}
-
-    - name: Create kind cluster
-#       if: steps.list-changed.outputs.changed == 'true'
-      uses: helm/kind-action@v1.10.0
-
-    - name: Run chart-testing (install)
-#       if: steps.list-changed.outputs.changed == 'true'
-      run: ct install --all --config ct.yaml
-#       run: ct install --target-branch ${{ github.event.repository.default_branch }}
-      

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

@@ -0,0 +1,243 @@
+name: Run Integration Tests v2
+concurrency:
+  group: Run-Integration-Tests-${{ github.workflow }}-${{ github.head_ref || github.event.workflow_run.head_branch || github.run_id }}
+  cancel-in-progress: true
+
+on:
+  merge_group:
+  pull_request:
+    branches:
+      - main
+      - 'release/**'
+
+env:
+  OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+  SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
+  CONFLUENCE_TEST_SPACE_URL: ${{ secrets.CONFLUENCE_TEST_SPACE_URL }}
+  CONFLUENCE_USER_NAME: ${{ secrets.CONFLUENCE_USER_NAME }}
+  CONFLUENCE_ACCESS_TOKEN: ${{ secrets.CONFLUENCE_ACCESS_TOKEN }}
+  
+jobs:
+  integration-tests:
+    # See https://runs-on.com/runners/linux/
+    runs-on: [runs-on,runner=8cpu-linux-x64,ram=16,"run-id=${{ github.run_id }}"]
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          username: ${{ secrets.DOCKER_USERNAME }}
+          password: ${{ secrets.DOCKER_TOKEN }}
+
+      # tag every docker image with "test" so that we can spin up the correct set
+      # of images during testing
+      
+      # We don't need to build the Web Docker image since it's not yet used
+      # in the integration tests. We have a separate action to verify that it builds 
+      # successfully.
+      - name: Pull Web Docker image
+        run: |
+          docker pull danswer/danswer-web-server:latest
+          docker tag danswer/danswer-web-server:latest danswer/danswer-web-server:test
+
+      # we use the runs-on cache for docker builds
+      # in conjunction with runs-on runners, it has better speed and unlimited caching
+      # https://runs-on.com/caching/s3-cache-for-github-actions/
+      # https://runs-on.com/caching/docker/
+      # https://github.com/moby/buildkit#s3-cache-experimental
+      
+      # images are built and run locally for testing purposes. Not pushed.
+      - name: Build Backend Docker image
+        uses: ./.github/actions/custom-build-and-push
+        with:
+          context: ./backend
+          file: ./backend/Dockerfile
+          platforms: linux/amd64
+          tags: danswer/danswer-backend:test
+          push: false
+          load: true
+          cache-from: type=s3,prefix=cache/${{ github.repository }}/integration-tests/backend/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }}
+          cache-to: type=s3,prefix=cache/${{ github.repository }}/integration-tests/backend/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }},mode=max
+
+      - name: Build Model Server Docker image
+        uses: ./.github/actions/custom-build-and-push
+        with:
+          context: ./backend
+          file: ./backend/Dockerfile.model_server
+          platforms: linux/amd64
+          tags: danswer/danswer-model-server:test
+          push: false
+          load: true
+          cache-from: type=s3,prefix=cache/${{ github.repository }}/integration-tests/model-server/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }}
+          cache-to: type=s3,prefix=cache/${{ github.repository }}/integration-tests/model-server/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }},mode=max
+      
+      - name: Build integration test Docker image
+        uses: ./.github/actions/custom-build-and-push
+        with:
+          context: ./backend
+          file: ./backend/tests/integration/Dockerfile
+          platforms: linux/amd64
+          tags: danswer/danswer-integration:test
+          push: false
+          load: true
+          cache-from: type=s3,prefix=cache/${{ github.repository }}/integration-tests/integration/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }}
+          cache-to: type=s3,prefix=cache/${{ github.repository }}/integration-tests/integration/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }},mode=max
+
+      # Start containers for multi-tenant tests
+      - name: Start Docker containers for multi-tenant tests
+        run: |
+          cd deployment/docker_compose
+          ENABLE_PAID_ENTERPRISE_EDITION_FEATURES=true \
+          MULTI_TENANT=true \
+          AUTH_TYPE=basic \
+          REQUIRE_EMAIL_VERIFICATION=false \
+          DISABLE_TELEMETRY=true \
+          IMAGE_TAG=test \
+          docker compose -f docker-compose.dev.yml -p danswer-stack up -d
+        id: start_docker_multi_tenant
+
+      # In practice, `cloud` Auth type would require OAUTH credentials to be set.
+      - name: Run Multi-Tenant Integration Tests
+        run: |
+          echo "Running integration tests..."
+          docker run --rm --network danswer-stack_default \
+            --name test-runner \
+            -e POSTGRES_HOST=relational_db \
+            -e POSTGRES_USER=postgres \
+            -e POSTGRES_PASSWORD=password \
+            -e POSTGRES_DB=postgres \
+            -e VESPA_HOST=index \
+            -e REDIS_HOST=cache \
+            -e API_SERVER_HOST=api_server \
+            -e OPENAI_API_KEY=${OPENAI_API_KEY} \
+            -e SLACK_BOT_TOKEN=${SLACK_BOT_TOKEN} \
+            -e TEST_WEB_HOSTNAME=test-runner \
+            -e AUTH_TYPE=cloud \
+            -e MULTI_TENANT=true \
+            danswer/danswer-integration:test \
+            /app/tests/integration/multitenant_tests
+        continue-on-error: true
+        id: run_multitenant_tests
+
+      - name: Check multi-tenant test results
+        run: |
+          if [ ${{ steps.run_tests.outcome }} == 'failure' ]; then
+            echo "Integration tests failed. Exiting with error."
+            exit 1
+          else
+            echo "All integration tests passed successfully."
+          fi 
+
+      - name: Stop multi-tenant Docker containers
+        run: |
+          cd deployment/docker_compose
+          docker compose -f docker-compose.dev.yml -p danswer-stack down -v
+
+
+      - name: Start Docker containers 
+        run: |
+          cd deployment/docker_compose
+          ENABLE_PAID_ENTERPRISE_EDITION_FEATURES=true \
+          AUTH_TYPE=basic \
+          REQUIRE_EMAIL_VERIFICATION=false \
+          DISABLE_TELEMETRY=true \
+          IMAGE_TAG=test \
+          docker compose -f docker-compose.dev.yml -p danswer-stack up -d
+        id: start_docker
+
+      - name: Wait for service to be ready
+        run: |
+          echo "Starting wait-for-service script..."
+          
+          docker logs -f danswer-stack-api_server-1 &
+
+          start_time=$(date +%s)
+          timeout=300  # 5 minutes in seconds
+          
+          while true; do
+            current_time=$(date +%s)
+            elapsed_time=$((current_time - start_time))
+            
+            if [ $elapsed_time -ge $timeout ]; then
+              echo "Timeout reached. Service did not become ready in 5 minutes."
+              exit 1
+            fi
+            
+            # Use curl with error handling to ignore specific exit code 56
+            response=$(curl -s -o /dev/null -w "%{http_code}" http://localhost:8080/health || echo "curl_error")
+            
+            if [ "$response" = "200" ]; then
+              echo "Service is ready!"
+              break
+            elif [ "$response" = "curl_error" ]; then
+              echo "Curl encountered an error, possibly exit code 56. Continuing to retry..."
+            else
+              echo "Service not ready yet (HTTP status $response). Retrying in 5 seconds..."
+            fi
+            
+            sleep 5
+          done
+          echo "Finished waiting for service."
+
+      - name: Run Standard Integration Tests
+        run: |
+          echo "Running integration tests..."
+          docker run --rm --network danswer-stack_default \
+            --name test-runner \
+            -e POSTGRES_HOST=relational_db \
+            -e POSTGRES_USER=postgres \
+            -e POSTGRES_PASSWORD=password \
+            -e POSTGRES_DB=postgres \
+            -e VESPA_HOST=index \
+            -e REDIS_HOST=cache \
+            -e API_SERVER_HOST=api_server \
+            -e OPENAI_API_KEY=${OPENAI_API_KEY} \
+            -e SLACK_BOT_TOKEN=${SLACK_BOT_TOKEN} \
+            -e CONFLUENCE_TEST_SPACE_URL=${CONFLUENCE_TEST_SPACE_URL} \
+            -e CONFLUENCE_USER_NAME=${CONFLUENCE_USER_NAME} \
+            -e CONFLUENCE_ACCESS_TOKEN=${CONFLUENCE_ACCESS_TOKEN} \
+            -e TEST_WEB_HOSTNAME=test-runner \
+            danswer/danswer-integration:test \
+            /app/tests/integration/tests \
+            /app/tests/integration/connector_job_tests
+        continue-on-error: true
+        id: run_tests
+
+      - name: Check test results
+        run: |
+          if [ ${{ steps.run_tests.outcome }} == 'failure' ]; then
+            echo "Integration tests failed. Exiting with error."
+            exit 1
+          else
+            echo "All integration tests passed successfully."
+          fi
+
+      # save before stopping the containers so the logs can be captured
+      - name: Save Docker logs
+        if: success() || failure()
+        run: |
+          cd deployment/docker_compose
+          docker compose -f docker-compose.dev.yml -p danswer-stack logs > docker-compose.log
+          mv docker-compose.log ${{ github.workspace }}/docker-compose.log
+
+      - name: Stop Docker containers
+        run: |
+          cd deployment/docker_compose
+          docker compose -f docker-compose.dev.yml -p danswer-stack down -v
+      
+      - name: Upload logs
+        if: success() || failure()
+        uses: actions/upload-artifact@v4
+        with:
+          name: docker-logs
+          path: ${{ github.workspace }}/docker-compose.log
+
+      - name: Stop Docker containers
+        run: |
+          cd deployment/docker_compose
+          docker compose -f docker-compose.dev.yml -p danswer-stack down -v

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

@@ -3,23 +3,21 @@ name: Python Checks
 on:
   merge_group:
   pull_request:
-    branches: [ main ]
+    branches:
+      - main
+      - 'release/**'
 
 jobs:
-  check-changes:
-    uses: ./.github/workflows/check-backend-changes.yml
-
   mypy-check:
-    needs: check-changes
-    if: needs.check-changes.outputs.run-tests == 'true'
-    runs-on: ubuntu-latest
+    # See https://runs-on.com/runners/linux/
+    runs-on: [runs-on,runner=8cpu-linux-x64,"run-id=${{ github.run_id }}"]
 
     steps:
     - name: Checkout code
-      uses: actions/checkout@v3
+      uses: actions/checkout@v4
 
     - name: Set up Python
-      uses: actions/setup-python@v4
+      uses: actions/setup-python@v5
       with:
         python-version: '3.11'
         cache: 'pip'
@@ -29,9 +27,9 @@ jobs:
           backend/requirements/model_server.txt
     - run: |
         python -m pip install --upgrade pip
-        pip install -r backend/requirements/default.txt
-        pip install -r backend/requirements/dev.txt
-        pip install -r backend/requirements/model_server.txt
+        pip install --retries 5 --timeout 30 -r backend/requirements/default.txt
+        pip install --retries 5 --timeout 30 -r backend/requirements/dev.txt
+        pip install --retries 5 --timeout 30 -r backend/requirements/model_server.txt
 
     - name: Run MyPy
       run: |
@@ -52,10 +50,3 @@ jobs:
       run: |
         cd backend
         black --check .
-
-  skip-tests:
-    needs: check-changes
-    if: needs.check-changes.outputs.run-tests == 'false'
-    runs-on: ubuntu-latest
-    steps:
-      - run: echo "No changes in backend, skipping this test."

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

@@ -15,15 +15,22 @@ env:
   CONFLUENCE_TEST_PAGE_ID: ${{ secrets.CONFLUENCE_TEST_PAGE_ID }}
   CONFLUENCE_USER_NAME: ${{ secrets.CONFLUENCE_USER_NAME }}
   CONFLUENCE_ACCESS_TOKEN: ${{ secrets.CONFLUENCE_ACCESS_TOKEN }}
+  # Jira
+  JIRA_USER_EMAIL: ${{ secrets.JIRA_USER_EMAIL }}
+  JIRA_API_TOKEN: ${{ secrets.JIRA_API_TOKEN }}
+  # Google
+  GOOGLE_DRIVE_SERVICE_ACCOUNT_JSON_STR: ${{ secrets.GOOGLE_DRIVE_SERVICE_ACCOUNT_JSON_STR }}
+  GOOGLE_DRIVE_OAUTH_CREDENTIALS_JSON_STR_TEST_USER_1: ${{ secrets.GOOGLE_DRIVE_OAUTH_CREDENTIALS_JSON_STR_TEST_USER_1 }}
+  GOOGLE_DRIVE_OAUTH_CREDENTIALS_JSON_STR: ${{ secrets.GOOGLE_DRIVE_OAUTH_CREDENTIALS_JSON_STR }}
+  GOOGLE_GMAIL_SERVICE_ACCOUNT_JSON_STR: ${{ secrets.GOOGLE_GMAIL_SERVICE_ACCOUNT_JSON_STR }}
+  GOOGLE_GMAIL_OAUTH_CREDENTIALS_JSON_STR: ${{ secrets.GOOGLE_GMAIL_OAUTH_CREDENTIALS_JSON_STR }}
+  # Slab
+  SLAB_BOT_TOKEN: ${{ secrets.SLAB_BOT_TOKEN }}
 
 jobs:
-  check-changes:
-    uses: ./.github/workflows/check-backend-changes.yml
-
   connectors-check:
-    needs: check-changes
-    if: needs.check-changes.outputs.run-tests == 'true'
-    runs-on: ubuntu-latest
+    # See https://runs-on.com/runners/linux/
+    runs-on: [runs-on,runner=8cpu-linux-x64,"run-id=${{ github.run_id }}"]
 
     env:
       PYTHONPATH: ./backend
@@ -33,7 +40,7 @@ jobs:
         uses: actions/checkout@v4
 
       - name: Set up Python
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@v5
         with:
           python-version: "3.11"
           cache: "pip"
@@ -44,8 +51,8 @@ jobs:
       - name: Install Dependencies
         run: |
           python -m pip install --upgrade pip
-          pip install -r backend/requirements/default.txt
-          pip install -r backend/requirements/dev.txt
+          pip install --retries 5 --timeout 30 -r backend/requirements/default.txt
+          pip install --retries 5 --timeout 30 -r backend/requirements/dev.txt
 
       - name: Run Tests
         shell: script -q -e -c "bash --noprofile --norc -eo pipefail {0}"
@@ -60,10 +67,3 @@ jobs:
             -H 'Content-type: application/json' \
             --data '{"text":"Scheduled Connector Tests failed! Check the run at: https://github.com/${{ github.repository }}/actions/runs/${{ github.run_id }}"}' \
             $SLACK_WEBHOOK
-
-  skip-tests:
-    needs: check-changes
-    if: needs.check-changes.outputs.run-tests == 'false'
-    runs-on: ubuntu-latest
-    steps:
-      - run: echo "No changes in backend, skipping this test."

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

@@ -0,0 +1,58 @@
+name: Connector Tests
+
+on:
+  schedule:
+    # This cron expression runs the job daily at 16:00 UTC (9am PT)
+    - cron: "0 16 * * *"
+
+env:
+  # Bedrock
+  AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }}
+  AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
+  AWS_REGION_NAME: ${{ secrets.AWS_REGION_NAME }}
+
+  # OpenAI
+  OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+
+jobs:
+  model-check:
+    # See https://runs-on.com/runners/linux/
+    runs-on: [runs-on,runner=8cpu-linux-x64,"run-id=${{ github.run_id }}"]
+
+    env:
+      PYTHONPATH: ./backend
+
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+          cache: "pip"
+          cache-dependency-path: |
+            backend/requirements/default.txt
+            backend/requirements/dev.txt
+
+      - name: Install Dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install --retries 5 --timeout 30 -r backend/requirements/default.txt
+          pip install --retries 5 --timeout 30 -r backend/requirements/dev.txt
+
+      - name: Run Tests
+        shell: script -q -e -c "bash --noprofile --norc -eo pipefail {0}"
+        run: |
+          py.test -o junit_family=xunit2 -xv --ff backend/tests/daily/llm
+          py.test -o junit_family=xunit2 -xv --ff backend/tests/daily/embedding
+
+      - name: Alert on Failure
+        if: failure() && github.event_name == 'schedule'
+        env:
+          SLACK_WEBHOOK: ${{ secrets.SLACK_WEBHOOK }}
+        run: |
+          curl -X POST \
+            -H 'Content-type: application/json' \
+            --data '{"text":"Scheduled Model Tests failed! Check the run at: https://github.com/${{ github.repository }}/actions/runs/${{ github.run_id }}"}' \
+            $SLACK_WEBHOOK

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

@@ -3,26 +3,25 @@ name: Python Unit Tests
 on:
   merge_group:
   pull_request:
-    branches: [ main ]
+    branches:
+      - main
+      - 'release/**'
 
 jobs:
-  check-changes:
-    uses: ./.github/workflows/check-backend-changes.yml
-
   backend-check:
-    needs: check-changes
-    if: needs.check-changes.outputs.run-tests == 'true'
-    runs-on: ubuntu-latest
+    # See https://runs-on.com/runners/linux/
+    runs-on: [runs-on,runner=8cpu-linux-x64,"run-id=${{ github.run_id }}"]
 
     env:
       PYTHONPATH: ./backend
-
+      REDIS_CLOUD_PYTEST_PASSWORD: ${{ secrets.REDIS_CLOUD_PYTEST_PASSWORD }}
+      
     steps:
     - name: Checkout code
       uses: actions/checkout@v4
 
     - name: Set up Python
-      uses: actions/setup-python@v4
+      uses: actions/setup-python@v5
       with:
         python-version: '3.11'
         cache: 'pip'
@@ -33,17 +32,9 @@ jobs:
     - name: Install Dependencies
       run: |
         python -m pip install --upgrade pip
-        pip install -r backend/requirements/default.txt
-        pip install -r backend/requirements/dev.txt
+        pip install --retries 5 --timeout 30 -r backend/requirements/default.txt
+        pip install --retries 5 --timeout 30 -r backend/requirements/dev.txt
 
     - name: Run Tests
       shell: script -q -e -c "bash --noprofile --norc -eo pipefail {0}"
       run: py.test -o junit_family=xunit2 -xv --ff backend/tests/unit
-
-
-  skip-tests:
-    needs: check-changes
-    if: needs.check-changes.outputs.run-tests == 'false'
-    runs-on: ubuntu-latest
-    steps:
-      - run: echo "No changes in backend, skipping this test."

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

@@ -1,6 +1,6 @@
 name: Quality Checks PR
 concurrency:
-  group: Quality-Checks-PR-${{ github.head_ref }}
+  group: Quality-Checks-PR-${{ github.workflow }}-${{ github.head_ref || github.event.workflow_run.head_branch || github.run_id }}
   cancel-in-progress: true
 
 on:
@@ -9,7 +9,8 @@ on:
 
 jobs:
   quality-checks:
-    runs-on: ubuntu-latest
+    # See https://runs-on.com/runners/linux/
+    runs-on: [runs-on,runner=8cpu-linux-x64,"run-id=${{ github.run_id }}"]
     steps:
       - uses: actions/checkout@v4
         with:
@@ -17,6 +18,6 @@ jobs:
       - uses: actions/setup-python@v5
         with:
           python-version: "3.11"
-      - uses: pre-commit/action@v3.0.0
+      - uses: pre-commit/action@v3.0.1
         with:
           extra_args: ${{ github.event_name == 'pull_request' && format('--from-ref {0} --to-ref {1}', github.event.pull_request.base.sha, github.event.pull_request.head.sha) || '' }}

.github/workflows/run-it.yml

@@ -1,172 +0,0 @@
-name: Run Integration Tests
-concurrency:
-  group: Run-Integration-Tests-${{ github.head_ref }}
-  cancel-in-progress: true
-
-on:
-  merge_group:
-  pull_request:
-    branches: [ main ]
-
-env:
-  OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
-
-jobs:
-  check-changes:
-    uses: ./.github/workflows/check-backend-changes.yml
-
-  integration-tests:
-    needs: check-changes
-    if: needs.check-changes.outputs.run-tests == 'true'
-    runs-on: 
-      group: 'arm64-image-builders'
-    steps:
-      - name: Checkout code
-        uses: actions/checkout@v4
-
-      - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v3
-
-      - name: Login to Docker Hub
-        uses: docker/login-action@v3
-        with:
-          username: ${{ secrets.DOCKER_USERNAME }}
-          password: ${{ secrets.DOCKER_TOKEN }}
-
-      # NOTE: we don't need to build the Web Docker image since it's not used
-      # during the IT for now. We have a separate action to verify it builds 
-      # succesfully
-      - name: Pull Web Docker image
-        run: |
-          docker pull danswer/danswer-web-server:latest
-          docker tag danswer/danswer-web-server:latest danswer/danswer-web-server:it
-
-      - name: Build Backend Docker image
-        uses: ./.github/actions/custom-build-and-push
-        with:
-          context: ./backend
-          file: ./backend/Dockerfile
-          platforms: linux/arm64
-          tags: danswer/danswer-backend:it
-          cache-from: type=registry,ref=danswer/danswer-backend:it
-          cache-to: |
-            type=registry,ref=danswer/danswer-backend:it,mode=max
-            type=inline
-
-      - name: Build Model Server Docker image
-        uses: ./.github/actions/custom-build-and-push
-        with:
-          context: ./backend
-          file: ./backend/Dockerfile.model_server
-          platforms: linux/arm64
-          tags: danswer/danswer-model-server:it
-          cache-from: type=registry,ref=danswer/danswer-model-server:it
-          cache-to: |
-            type=registry,ref=danswer/danswer-model-server:it,mode=max
-            type=inline
-
-      - name: Build integration test Docker image
-        uses: ./.github/actions/custom-build-and-push
-        with:
-          context: ./backend
-          file: ./backend/tests/integration/Dockerfile
-          platforms: linux/arm64
-          tags: danswer/integration-test-runner:it
-          cache-from: type=registry,ref=danswer/integration-test-runner:it
-          cache-to: |
-            type=registry,ref=danswer/integration-test-runner:it,mode=max
-            type=inline
-
-      - name: Start Docker containers
-        run: |
-          cd deployment/docker_compose
-          ENABLE_PAID_ENTERPRISE_EDITION_FEATURES=true \
-          AUTH_TYPE=basic \
-          REQUIRE_EMAIL_VERIFICATION=false \
-          DISABLE_TELEMETRY=true \
-          IMAGE_TAG=it \
-          docker compose -f docker-compose.dev.yml -p danswer-stack up -d
-        id: start_docker
-
-      - name: Wait for service to be ready
-        run: |
-          echo "Starting wait-for-service script..."
-          
-          start_time=$(date +%s)
-          timeout=300  # 5 minutes in seconds
-          
-          while true; do
-            current_time=$(date +%s)
-            elapsed_time=$((current_time - start_time))
-            
-            if [ $elapsed_time -ge $timeout ]; then
-              echo "Timeout reached. Service did not become ready in 5 minutes."
-              exit 1
-            fi
-            
-            # Use curl with error handling to ignore specific exit code 56
-            response=$(curl -s -o /dev/null -w "%{http_code}" http://localhost:8080/health || echo "curl_error")
-            
-            if [ "$response" = "200" ]; then
-              echo "Service is ready!"
-              break
-            elif [ "$response" = "curl_error" ]; then
-              echo "Curl encountered an error, possibly exit code 56. Continuing to retry..."
-            else
-              echo "Service not ready yet (HTTP status $response). Retrying in 5 seconds..."
-            fi
-            
-            sleep 5
-          done
-          echo "Finished waiting for service."
-
-      - name: Run integration tests
-        run: |
-          echo "Running integration tests..."
-          docker run --rm --network danswer-stack_default \
-            -e POSTGRES_HOST=relational_db \
-            -e POSTGRES_USER=postgres \
-            -e POSTGRES_PASSWORD=password \
-            -e POSTGRES_DB=postgres \
-            -e VESPA_HOST=index \
-            -e REDIS_HOST=cache \
-            -e API_SERVER_HOST=api_server \
-            -e OPENAI_API_KEY=${OPENAI_API_KEY} \
-            danswer/integration-test-runner:it
-        continue-on-error: true
-        id: run_tests
-
-      - name: Check test results
-        run: |
-          if [ ${{ steps.run_tests.outcome }} == 'failure' ]; then
-            echo "Integration tests failed. Exiting with error."
-            exit 1
-          else
-            echo "All integration tests passed successfully."
-          fi
-
-      - name: Save Docker logs
-        if: success() || failure()
-        run: |
-          cd deployment/docker_compose
-          docker compose -f docker-compose.dev.yml -p danswer-stack logs > docker-compose.log
-          mv docker-compose.log ${{ github.workspace }}/docker-compose.log
-      
-      - name: Upload logs
-        if: success() || failure()
-        uses: actions/upload-artifact@v3
-        with:
-          name: docker-logs
-          path: ${{ github.workspace }}/docker-compose.log
-
-      - name: Stop Docker containers
-        run: |
-          cd deployment/docker_compose
-          docker compose -f docker-compose.dev.yml -p danswer-stack down -v
-
-  skip-tests:
-    needs: check-changes
-    if: needs.check-changes.outputs.run-tests == 'false'
-    runs-on: ubuntu-latest
-    steps:
-      - run: echo "No changes in backend, skipping this test."

.github/workflows/tag-nightly.yml

@@ -0,0 +1,54 @@
+name: Nightly Tag Push
+
+on:
+  schedule:
+    - cron: '0 10 * * *' # Runs every day at 2 AM PST / 3 AM PDT / 10 AM UTC
+
+permissions:
+  contents: write  # Allows pushing tags to the repository
+
+jobs:
+  create-and-push-tag:
+    runs-on: [runs-on,runner=2cpu-linux-x64,"run-id=${{ github.run_id }}"]
+
+    steps:
+    # actions using GITHUB_TOKEN cannot trigger another workflow, but we do want this to trigger docker pushes
+    # see https://github.com/orgs/community/discussions/27028#discussioncomment-3254367 for the workaround we
+    # implement here which needs an actual user's deploy key
+    - name: Checkout code
+      uses: actions/checkout@v4
+      with:
+        ssh-key: "${{ secrets.RKUO_DEPLOY_KEY }}"
+
+    - name: Set up Git user
+      run: |
+        git config user.name "Richard Kuo [bot]"
+        git config user.email "rkuo[bot]@danswer.ai"
+
+    - name: Check for existing nightly tag
+      id: check_tag
+      run: |
+        if git tag --points-at HEAD --list "nightly-latest*" | grep -q .; then
+          echo "A tag starting with 'nightly-latest' already exists on HEAD."
+          echo "tag_exists=true" >> $GITHUB_OUTPUT
+        else
+          echo "No tag starting with 'nightly-latest' exists on HEAD."
+          echo "tag_exists=false" >> $GITHUB_OUTPUT
+        fi
+        
+    # don't tag again if HEAD already has a nightly-latest tag on it
+    - name: Create Nightly Tag
+      if: steps.check_tag.outputs.tag_exists == 'false'
+      env:
+        DATE: ${{ github.run_id }}
+      run: |
+        TAG_NAME="nightly-latest-$(date +'%Y%m%d')"
+        echo "Creating tag: $TAG_NAME"
+        git tag $TAG_NAME
+
+    - name: Push Tag
+      if: steps.check_tag.outputs.tag_exists == 'false'
+      run: |
+        TAG_NAME="nightly-latest-$(date +'%Y%m%d')"
+        git push origin $TAG_NAME
+        

.gitignore

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

.prettierignore

@@ -0,0 +1 @@
+backend/tests/integration/tests/pruning/website

.vscode/launch.template.jsonc

@@ -6,19 +6,69 @@
     // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
     "version": "0.2.0",
     "compounds": [
+		{
+			// Dummy entry used to label the group
+			"name": "--- Compound ---",
+		  	"configurations": [
+		  		"--- Individual ---"
+		  	],
+			"presentation": {
+				 "group": "1",
+			 }
+		},
         {
             "name": "Run All Danswer Services",
             "configurations": [
                 "Web Server",
                 "Model Server",
                 "API Server",
-                "Indexing",
-                "Background Jobs",
-                "Slack Bot"
-            ]
-        }
+                "Slack Bot",
+		  		"Celery primary", 
+		  		"Celery light", 
+		  		"Celery heavy", 
+		  		"Celery indexing", 
+		  		"Celery beat",
+            ],
+			"presentation": {
+				 "group": "1",
+			 }
+        },
+		{
+			"name": "Web / Model / API",
+		  	"configurations": [
+		  		"Web Server", 
+		  		"Model Server", 
+		  		"API Server",
+		  	],
+			"presentation": {
+				 "group": "1",
+			 }
+		},
+		{
+			"name": "Celery (all)",
+		  	"configurations": [
+		  		"Celery primary", 
+		  		"Celery light", 
+		  		"Celery heavy", 
+		  		"Celery indexing", 
+		  		"Celery beat"
+		  	],
+			"presentation": {
+				 "group": "1",
+			 }
+		}
     ],
     "configurations": [
+		{
+	    	// Dummy entry used to label the group
+			"name": "--- Individual ---",
+			"type": "node",
+			"request": "launch",
+			"presentation": {
+				"group": "2",
+				"order": 0
+			}
+		},
         {
             "name": "Web Server",
             "type": "node",
@@ -29,7 +79,11 @@
             "runtimeArgs": [
                 "run", "dev"
             ],
-            "console": "integratedTerminal"
+            "presentation": {
+				 "group": "2",
+			 },
+            "console": "integratedTerminal",
+            "consoleTitle": "Web Server Console"
         },
         {
             "name": "Model Server",
@@ -48,7 +102,11 @@
                 "--reload",
                 "--port",
                 "9000"
-            ]
+            ],
+            "presentation": {
+				 "group": "2",
+			},
+            "consoleTitle": "Model Server Console"
         },
         {
             "name": "API Server",
@@ -68,43 +126,13 @@
                 "--reload",
                 "--port",
                 "8080"
-            ]
+            ],
+            "presentation": {
+				 "group": "2",
+			 },
+            "consoleTitle": "API Server Console"
         },
-        {
-            "name": "Indexing",
-            "consoleName": "Indexing",
-            "type": "debugpy",
-            "request": "launch",
-            "program": "danswer/background/update.py",
-            "cwd": "${workspaceFolder}/backend",
-            "envFile": "${workspaceFolder}/.vscode/.env",
-            "env": {
-                "ENABLE_MULTIPASS_INDEXING": "false",
-                "LOG_LEVEL": "DEBUG",
-                "PYTHONUNBUFFERED": "1",
-                "PYTHONPATH": "."
-            }
-        },
-        // Celery and all async jobs, usually would include indexing as well but this is handled separately above for dev
-        {
-            "name": "Background Jobs",
-            "consoleName": "Background Jobs",
-            "type": "debugpy",
-            "request": "launch",
-            "program": "scripts/dev_run_background_jobs.py",
-            "cwd": "${workspaceFolder}/backend",
-            "envFile": "${workspaceFolder}/.vscode/.env",
-            "env": {
-                "LOG_DANSWER_MODEL_INTERACTIONS": "True",
-                "LOG_LEVEL": "DEBUG",
-                "PYTHONUNBUFFERED": "1",
-                "PYTHONPATH": "."
-            },
-            "args": [
-                "--no-indexing"
-            ]
-        },
-        // For the listner to access the Slack API,
+        // For the listener to access the Slack API,
         // DANSWER_BOT_SLACK_APP_TOKEN & DANSWER_BOT_SLACK_BOT_TOKEN need to be set in .env file located in the root of the project
         {
             "name": "Slack Bot",
@@ -118,7 +146,151 @@
                 "LOG_LEVEL": "DEBUG",
                 "PYTHONUNBUFFERED": "1",
                 "PYTHONPATH": "."
-            }
+            },
+            "presentation": {
+				 "group": "2",
+			},
+            "consoleTitle": "Slack Bot Console"
+        },
+        {
+            "name": "Celery primary",
+            "type": "debugpy",
+            "request": "launch",
+            "module": "celery",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {
+                "LOG_LEVEL": "INFO",
+                "PYTHONUNBUFFERED": "1",
+                "PYTHONPATH": "."
+            },
+            "args": [
+                "-A",
+                "danswer.background.celery.versioned_apps.primary",
+                "worker",
+                "--pool=threads",
+                "--concurrency=4",
+                "--prefetch-multiplier=1",
+                "--loglevel=INFO",
+                "--hostname=primary@%n",
+                "-Q",
+                "celery",
+            ],
+            "presentation": {
+				 "group": "2",
+			 },
+            "consoleTitle": "Celery primary Console"
+        },
+        {
+            "name": "Celery light",
+            "type": "debugpy",
+            "request": "launch",
+            "module": "celery",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {
+                "LOG_LEVEL": "INFO",
+                "PYTHONUNBUFFERED": "1",
+                "PYTHONPATH": "."
+            },
+            "args": [
+                "-A",
+                "danswer.background.celery.versioned_apps.light",
+                "worker",
+                "--pool=threads",
+                "--concurrency=64",
+                "--prefetch-multiplier=8",
+                "--loglevel=INFO",
+                "--hostname=light@%n",
+                "-Q",
+                "vespa_metadata_sync,connector_deletion,doc_permissions_upsert",
+            ],
+            "presentation": {
+				 "group": "2",
+			 },
+            "consoleTitle": "Celery light Console"
+        },
+        {
+            "name": "Celery heavy",
+            "type": "debugpy",
+            "request": "launch",
+            "module": "celery",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {
+                "LOG_LEVEL": "INFO",
+                "PYTHONUNBUFFERED": "1",
+                "PYTHONPATH": "."
+            },
+            "args": [
+                "-A",
+                "danswer.background.celery.versioned_apps.heavy",
+                "worker",
+                "--pool=threads",
+                "--concurrency=4",
+                "--prefetch-multiplier=1",
+                "--loglevel=INFO",
+                "--hostname=heavy@%n",
+                "-Q",
+                "connector_pruning,connector_doc_permissions_sync,connector_external_group_sync",
+            ],
+            "presentation": {
+				 "group": "2",
+			 },
+            "consoleTitle": "Celery heavy Console"
+        },
+        {
+            "name": "Celery indexing",
+            "type": "debugpy",
+            "request": "launch",
+            "module": "celery",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {
+                "ENABLE_MULTIPASS_INDEXING": "false",
+                "LOG_LEVEL": "DEBUG",
+                "PYTHONUNBUFFERED": "1",
+                "PYTHONPATH": "."
+            },
+            "args": [
+                "-A",
+                "danswer.background.celery.versioned_apps.indexing",
+                "worker",
+                "--pool=threads",
+                "--concurrency=1",
+                "--prefetch-multiplier=1",
+                "--loglevel=INFO",
+                "--hostname=indexing@%n",
+                "-Q",
+                "connector_indexing",
+            ],
+            "presentation": {
+				 "group": "2",
+			 },
+            "consoleTitle": "Celery indexing Console"
+        },
+        {
+            "name": "Celery beat",
+            "type": "debugpy",
+            "request": "launch",
+            "module": "celery",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {
+                "LOG_LEVEL": "DEBUG",
+                "PYTHONUNBUFFERED": "1",
+                "PYTHONPATH": "."
+            },
+            "args": [
+                "-A",
+                "danswer.background.celery.versioned_apps.beat",
+                "beat",
+                "--loglevel=INFO",
+            ],
+            "presentation": {
+				 "group": "2",
+			 },
+            "consoleTitle": "Celery beat Console"
         },
         {
             "name": "Pytest",
@@ -137,8 +309,22 @@
                 "-v"
                 // Specify a sepcific module/test to run or provide nothing to run all tests
                 //"tests/unit/danswer/llm/answering/test_prune_and_merge.py"
-            ]
+            ],
+            "presentation": {
+				 "group": "2",
+			 },
+            "consoleTitle": "Pytest Console"
         },
+		{
+	    	// Dummy entry used to label the group
+			"name": "--- Tasks ---",
+			"type": "node",
+			"request": "launch",
+			"presentation": {
+				"group": "3",
+				"order": 0
+			}
+		},
         {
             "name": "Clear and Restart External Volumes and Containers",
             "type": "node",
@@ -147,7 +333,27 @@
             "runtimeArgs": ["${workspaceFolder}/backend/scripts/restart_containers.sh"],
             "cwd": "${workspaceFolder}",
             "console": "integratedTerminal",
-            "stopOnEntry": true
-        }
+            "stopOnEntry": true,
+            "presentation": {
+				 "group": "3",
+			 },
+        },
+        {
+	        // Celery jobs launched through a single background script (legacy)
+	        // Recommend using the "Celery (all)" compound launch instead.
+            "name": "Background Jobs",
+            "consoleName": "Background Jobs",
+            "type": "debugpy",
+            "request": "launch",
+            "program": "scripts/dev_run_background_jobs.py",
+            "cwd": "${workspaceFolder}/backend",
+            "envFile": "${workspaceFolder}/.vscode/.env",
+            "env": {
+                "LOG_DANSWER_MODEL_INTERACTIONS": "True",
+                "LOG_LEVEL": "DEBUG",
+                "PYTHONUNBUFFERED": "1",
+                "PYTHONPATH": "."
+            },
+        },
     ]
 }

CONTRIBUTING.md

@@ -22,7 +22,7 @@ Your input is vital to making sure that Danswer moves in the right direction.
 Before starting on implementation, please raise a GitHub issue.
 
 And always feel free to message us (Chris Weaver / Yuhong Sun) on 
-[Slack](https://join.slack.com/t/danswer/shared_invite/zt-2afut44lv-Rw3kSWu6_OmdAXRpCv80DQ) / 
+[Slack](https://join.slack.com/t/danswer/shared_invite/zt-2lcmqw703-071hBuZBfNEOGUsLa5PXvQ) / 
 [Discord](https://discord.gg/TDJ59cGV2X) directly about anything at all. 
 
 
@@ -32,7 +32,7 @@ To contribute to this project, please follow the
 When opening a pull request, mention related issues and feel free to tag relevant maintainers.
 
 Before creating a pull request please make sure that the new changes conform to the formatting and linting requirements.
-See the [Formatting and Linting](#-formatting-and-linting) section for how to run these checks locally.
+See the [Formatting and Linting](#formatting-and-linting) section for how to run these checks locally.
 
 
 ### Getting Help 🙋
@@ -56,19 +56,18 @@ Danswer being a fully functional app, relies on some external software, specific
 
 
 > **Note:**
-> This guide provides instructions to set up the Danswer specific services outside of Docker because it's easier for
-> development purposes. However, you can also use the containers and update with local changes by providing the
-> `--build` flag.
+> This guide provides instructions to build and run Danswer locally from source with Docker containers providing the above external software. We believe this combination is easier for
+> development purposes. If you prefer to use pre-built container images, we provide instructions on running the full Danswer stack within Docker below.
 
 
 ### Local Set Up
-Be sure to use Python version 3.11.
+Be sure to use Python version 3.11. For instructions on installing Python 3.11 on macOS, refer to the [CONTRIBUTING_MACOS.md](./CONTRIBUTING_MACOS.md) readme.
 
 If using a lower version, modifications will have to be made to the code.
 If using a higher version, sometimes some libraries will not be available (i.e. we had problems with Tensorflow in the past with higher versions of python).
 
 
-#### Installing Requirements
+#### Backend: Python requirements
 Currently, we use pip and recommend creating a virtual environment.
 
 For convenience here's a command for it:
@@ -98,6 +97,16 @@ pip install -r danswer/backend/requirements/ee.txt
 pip install -r danswer/backend/requirements/model_server.txt
 ```
 
+Install Playwright for Python (headless browser required by the Web Connector)
+
+In the activated Python virtualenv, install Playwright for Python by running:
+```bash
+playwright install
+```
+
+You may have to deactivate and reactivate your virtualenv for `playwright` to appear on your path.
+
+#### Frontend: Node dependencies
 
 Install [Node.js and npm](https://docs.npmjs.com/downloading-and-installing-node-js-and-npm) for the frontend.
 Once the above is done, navigate to `danswer/web` run:
@@ -105,19 +114,7 @@ Once the above is done, navigate to `danswer/web` run:
 npm i
 ```
 
-Install Playwright (headless browser required by the Web Connector)
-
-> **Note:**
-> If you have just run the pip install, open a new terminal and source the python virtual-env again.
-> This will pull the updated PATH to include playwright
-
-Then install Playwright by running:
-```bash
-playwright install
-```
-
-
-#### Dependent Docker Containers
+#### Docker containers for external software
 You will need Docker installed to run these containers.
 
 First navigate to `danswer/deployment/docker_compose`, then start up Postgres/Vespa/Redis with:
@@ -127,7 +124,7 @@ docker compose -f docker-compose.dev.yml -p danswer-stack up -d index relational
 (index refers to Vespa, relational_db refers to Postgres, and cache refers to Redis)
 
 
-#### Running Danswer
+#### Running Danswer locally
 To start the frontend, navigate to `danswer/web` and run:
 ```bash
 npm run dev
@@ -176,6 +173,36 @@ powershell -Command "
 > **Note:**
 > If you need finer logging, add the additional environment variable `LOG_LEVEL=DEBUG` to the relevant services.
 
+#### Wrapping up
+
+You should now have 4 servers running:
+
+- Web server
+- Backend API
+- Model server
+- Background jobs
+
+Now, visit `http://localhost:3000` in your browser. You should see the Danswer onboarding wizard where you can connect your external LLM provider to Danswer.
+
+You've successfully set up a local Danswer instance! 🏁
+
+#### Running the Danswer application in a container
+
+You can run the full Danswer application stack from pre-built images including all external software dependencies.
+
+Navigate to `danswer/deployment/docker_compose` and run:
+
+```bash
+docker compose -f docker-compose.dev.yml -p danswer-stack up -d
+```
+
+After Docker pulls and starts these containers, navigate to `http://localhost:3000` to use Danswer.
+
+If you want to make changes to Danswer and run those changes in Docker, you can also build a local version of the Danswer container images that incorporates your changes like so:
+
+```bash
+docker compose -f docker-compose.dev.yml -p danswer-stack up -d --build
+```
 
 ### Formatting and Linting
 #### Backend

README.md

@@ -1,47 +1,48 @@
-<!-- DANSWER_METADATA={"link": "https://github.com/danswer-ai/danswer/blob/main/README.md"} -->
+<!-- DANSWER_METADATA={"link": "https://github.com/onyx-dot-app/onyx/blob/main/README.md"} -->
+<a name="readme-top"></a>
 
 <h2 align="center">
-<a href="https://www.danswer.ai/"> <img width="50%" src="https://github.com/danswer-owners/danswer/blob/1fabd9372d66cd54238847197c33f091a724803b/DanswerWithName.png?raw=true)" /></a>
+<a href="https://www.onyx.app/"> <img width="50%" src="https://github.com/onyx-dot-app/onyx/blob/logo/LogoOnyx.png?raw=true)" /></a>
 </h2>
 
 <p align="center">
-<p align="center">Open Source Gen-AI Chat + Unified Search.</p>
+<p align="center">Open Source Gen-AI + Enterprise Search.</p>
 
 <p align="center">
-<a href="https://docs.danswer.dev/" target="_blank">
+<a href="https://docs.onyx.app/" target="_blank">
     <img src="https://img.shields.io/badge/docs-view-blue" alt="Documentation">
 </a>
-<a href="https://join.slack.com/t/danswer/shared_invite/zt-2lcmqw703-071hBuZBfNEOGUsLa5PXvQ" target="_blank">
+<a href="https://join.slack.com/t/onyx-dot-app/shared_invite/zt-2sslpdbyq-iIbTaNIVPBw_i_4vrujLYQ" target="_blank">
     <img src="https://img.shields.io/badge/slack-join-blue.svg?logo=slack" alt="Slack">
 </a>
 <a href="https://discord.gg/TDJ59cGV2X" target="_blank">
     <img src="https://img.shields.io/badge/discord-join-blue.svg?logo=discord&logoColor=white" alt="Discord">
 </a>
-<a href="https://github.com/danswer-ai/danswer/blob/main/README.md" target="_blank">
+<a href="https://github.com/onyx-dot-app/onyx/blob/main/README.md" target="_blank">
     <img src="https://img.shields.io/static/v1?label=license&message=MIT&color=blue" alt="License">
 </a>
 </p>
 
-<strong>[Danswer](https://www.danswer.ai/)</strong> is the AI Assistant connected to your company's docs, apps, and people. 
-Danswer provides a Chat interface and plugs into any LLM of your choice. Danswer can be deployed anywhere and for any 
+<strong>[Onyx](https://www.onyx.app/)</strong> (Formerly Danswer) is the AI Assistant connected to your company's docs, apps, and people. 
+Onyx provides a Chat interface and plugs into any LLM of your choice. Onyx can be deployed anywhere and for any 
 scale - on a laptop, on-premise, or to cloud. Since you own the deployment, your user data and chats are fully in your 
-own control. Danswer is MIT licensed and designed to be modular and easily extensible. The system also comes fully ready 
+own control. Onyx is dual Licensed with most of it under MIT license and designed to be modular and easily extensible. The system also comes fully ready 
 for production usage with user authentication, role management (admin/basic users), chat persistence, and a UI for 
-configuring Personas (AI Assistants) and their Prompts.
+configuring AI Assistants.
 
-Danswer also serves as a Unified Search across all common workplace tools such as Slack, Google Drive, Confluence, etc.
-By combining LLMs and team specific knowledge, Danswer becomes a subject matter expert for the team. Imagine ChatGPT if
+Onyx also serves as a Enterprise Search across all common workplace tools such as Slack, Google Drive, Confluence, etc.
+By combining LLMs and team specific knowledge, Onyx becomes a subject matter expert for the team. Imagine ChatGPT if
 it had access to your team's unique knowledge! It enables questions such as "A customer wants feature X, is this already
 supported?" or "Where's the pull request for feature Y?"
 
 <h3>Usage</h3>
 
-Danswer Web App:
+Onyx Web App:
 
 https://github.com/danswer-ai/danswer/assets/32520769/563be14c-9304-47b5-bf0a-9049c2b6f410
 
 
-Or, plug Danswer into your existing Slack workflows (more integrations to come 😁):
+Or, plug Onyx into your existing Slack workflows (more integrations to come 😁):
 
 https://github.com/danswer-ai/danswer/assets/25087905/3e19739b-d178-4371-9a38-011430bdec1b
 
@@ -51,16 +52,16 @@ For more details on the Admin UI to manage connectors and users, check out our
 
 ## Deployment
 
-Danswer can easily be run locally (even on a laptop) or deployed on a virtual machine with a single
-`docker compose` command. Checkout our [docs](https://docs.danswer.dev/quickstart) to learn more.
+Onyx can easily be run locally (even on a laptop) or deployed on a virtual machine with a single
+`docker compose` command. Checkout our [docs](https://docs.onyx.app/quickstart) to learn more.
 
-We also have built-in support for deployment on Kubernetes. Files for that can be found [here](https://github.com/danswer-ai/danswer/tree/main/deployment/kubernetes).
+We also have built-in support for deployment on Kubernetes. Files for that can be found [here](https://github.com/onyx-dot-app/onyx/tree/main/deployment/kubernetes).
 
 
 ## 💃 Main Features 
 * Chat UI with the ability to select documents to chat with.
 * Create custom AI Assistants with different prompts and backing knowledge sets.
-* Connect Danswer with LLM of your choice (self-host for a fully airgapped solution).
+* Connect Onyx with LLM of your choice (self-host for a fully airgapped solution).
 * Document Search + AI Answers for natural language queries.
 * Connectors to all common workplace tools like Google Drive, Confluence, Slack, etc.
 * Slack integration to get answers and search results directly in Slack.
@@ -68,18 +69,18 @@ We also have built-in support for deployment on Kubernetes. Files for that can b
 
 ## 🚧 Roadmap
 * Chat/Prompt sharing with specific teammates and user groups.
-* Multi-Model model support, chat with images, video etc.
+* Multimodal model support, chat with images, video etc.
 * Choosing between LLMs and parameters during chat session.
 * Tool calling and agent configurations options.
 * Organizational understanding and ability to locate and suggest experts from your team.
 
 
-## Other Noteable Benefits of Danswer
+## Other Notable Benefits of Onyx
 * User Authentication with document level access management.
 * Best in class Hybrid Search across all sources (BM-25 + prefix aware embedding models).
 * Admin Dashboard to configure connectors, document-sets, access, etc.
 * Custom deep learning models + learn from user feedback.
-* Easy deployment and ability to host Danswer anywhere of your choosing.
+* Easy deployment and ability to host Onyx anywhere of your choosing.
 
 
 ## 🔌 Connectors
@@ -107,10 +108,10 @@ Efficiently pulls the latest changes from:
 
 ## 📚 Editions
 
-There are two editions of Danswer:
+There are two editions of Onyx:
 
-  * Danswer Community Edition (CE) is available freely under the MIT Expat license. This version has ALL the core features discussed above. This is the version of Danswer you will get if you follow the Deployment guide above.
-  * Danswer Enterprise Edition (EE) includes extra features that are primarily useful for larger organizations. Specifically, this includes:
+  * Onyx Community Edition (CE) is available freely under the MIT Expat license. This version has ALL the core features discussed above. This is the version of Onyx you will get if you follow the Deployment guide above.
+  * Onyx Enterprise Edition (EE) includes extra features that are primarily useful for larger organizations. Specifically, this includes:
     * Single Sign-On (SSO), with support for both SAML and OIDC
     * Role-based access control
     * Document permission inheritance from connected sources
@@ -118,12 +119,28 @@ There are two editions of Danswer:
     * Whitelabeling
     * API key authentication
     * Encryption of secrets
-    * Any many more! Checkout [our website](https://www.danswer.ai/) for the latest.
+    * Any many more! Checkout [our website](https://www.onyx.app/) for the latest.
 
-To try the Danswer Enterprise Edition: 
+To try the Onyx Enterprise Edition: 
 
-  1. Checkout our [Cloud product](https://app.danswer.ai/signup).
-  2. For self-hosting, contact us at [founders@danswer.ai](mailto:founders@danswer.ai) or book a call with us on our [Cal](https://cal.com/team/danswer/founders).
+  1. Checkout our [Cloud product](https://cloud.onyx.app/signup).
+  2. For self-hosting, contact us at [founders@onyx.app](mailto:founders@onyx.app) or book a call with us on our [Cal](https://cal.com/team/danswer/founders).
 
 ## 💡 Contributing
 Looking to contribute? Please check out the [Contribution Guide](CONTRIBUTING.md) for more details.
+
+## ⭐Star History
+
+[![Star History Chart](https://api.star-history.com/svg?repos=onyx-dot-app/onyx&type=Date)](https://star-history.com/#onyx-dot-app/onyx&Date)
+
+## ✨Contributors
+
+<a href="https://github.com/onyx-dot-app/onyx/graphs/contributors">
+  <img alt="contributors" src="https://contrib.rocks/image?repo=onyx-dot-app/onyx"/>
+</a>
+
+<p align="right" style="font-size: 14px; color: #555; margin-top: 20px;">
+    <a href="#readme-top" style="text-decoration: none; color: #007bff; font-weight: bold;">
+        ↑ Back to Top ↑
+    </a>
+</p>

backend/Dockerfile

@@ -8,8 +8,10 @@ Edition features outside of personal development or testing purposes. Please rea
 founders@danswer.ai for more information. Please visit https://github.com/danswer-ai/danswer"
 
 # Default DANSWER_VERSION, typically overriden during builds by GitHub Actions.
-ARG DANSWER_VERSION=0.3-dev
-ENV DANSWER_VERSION=${DANSWER_VERSION}
+ARG DANSWER_VERSION=0.8-dev
+ENV DANSWER_VERSION=${DANSWER_VERSION} \
+    DANSWER_RUNNING_IN_DOCKER="true"
+
 
 RUN echo "DANSWER_VERSION: ${DANSWER_VERSION}"
 # Install system dependencies
@@ -35,11 +37,15 @@ RUN apt-get update && \
     rm -rf /var/lib/apt/lists/* && \
     apt-get clean
 
+
+
 # Install Python dependencies
 # Remove py which is pulled in by retry, py is not needed and is a CVE
 COPY ./requirements/default.txt /tmp/requirements.txt
 COPY ./requirements/ee.txt /tmp/ee-requirements.txt
 RUN pip install --no-cache-dir --upgrade \
+        --retries 5 \
+        --timeout 30 \
         -r /tmp/requirements.txt \
         -r /tmp/ee-requirements.txt && \
     pip uninstall -y py && \
@@ -67,11 +73,11 @@ RUN apt-get update && \
     rm -rf /var/lib/apt/lists/* && \
     rm -f /usr/local/lib/python3.11/site-packages/tornado/test/test.key
 
+
 # Pre-downloading models for setups with limited egress
 RUN python -c "from tokenizers import Tokenizer; \
 Tokenizer.from_pretrained('nomic-ai/nomic-embed-text-v1')"
 
-
 # Pre-downloading NLTK for setups with limited egress
 RUN python -c "import nltk; \
 nltk.download('stopwords', quiet=True); \
@@ -89,6 +95,7 @@ COPY supervisord.conf /etc/supervisor/conf.d/supervisord.conf
 COPY ./danswer /app/danswer
 COPY ./shared_configs /app/shared_configs
 COPY ./alembic /app/alembic
+COPY ./alembic_tenants /app/alembic_tenants
 COPY ./alembic.ini /app/alembic.ini
 COPY supervisord.conf /usr/etc/supervisord.conf
 
@@ -98,7 +105,7 @@ COPY ./scripts/force_delete_connector_by_id.py /app/scripts/force_delete_connect
 # Put logo in assets
 COPY ./assets /app/assets
 
-ENV PYTHONPATH /app
+ENV PYTHONPATH=/app
 
 # Default command which does nothing
 # This container is used by api server and background which specify their own CMD

backend/Dockerfile.model_server

@@ -7,12 +7,18 @@ You can find it at https://hub.docker.com/r/danswer/danswer-model-server. For mo
 visit https://github.com/danswer-ai/danswer."
 
 # Default DANSWER_VERSION, typically overriden during builds by GitHub Actions.
-ARG DANSWER_VERSION=0.3-dev
-ENV DANSWER_VERSION=${DANSWER_VERSION}
+ARG DANSWER_VERSION=0.8-dev
+ENV DANSWER_VERSION=${DANSWER_VERSION} \
+    DANSWER_RUNNING_IN_DOCKER="true"
+
+
 RUN echo "DANSWER_VERSION: ${DANSWER_VERSION}"
 
 COPY ./requirements/model_server.txt /tmp/requirements.txt
-RUN pip install --no-cache-dir --upgrade -r /tmp/requirements.txt
+RUN pip install --no-cache-dir --upgrade \
+        --retries 5 \
+        --timeout 30 \
+        -r /tmp/requirements.txt
 
 RUN apt-get remove -y --allow-remove-essential perl-base && \
     apt-get autoremove -y
@@ -49,6 +55,6 @@ COPY ./shared_configs /app/shared_configs
 # Model Server main code
 COPY ./model_server /app/model_server
 
-ENV PYTHONPATH /app
+ENV PYTHONPATH=/app
 
 CMD ["uvicorn", "model_server.main:app", "--host", "0.0.0.0", "--port", "9000"]

backend/alembic.ini

@@ -1,6 +1,6 @@
 # A generic, single database configuration.
 
-[alembic]
+[DEFAULT]
 # path to migration scripts
 script_location = alembic
 
@@ -47,7 +47,8 @@ prepend_sys_path = .
 # version_path_separator = :
 # version_path_separator = ;
 # version_path_separator = space
-version_path_separator = os  # Use os.pathsep. Default configuration used for new projects.
+version_path_separator = os  
+# Use os.pathsep. Default configuration used for new projects.
 
 # set to 'true' to search source files recursively
 # in each "version_locations" directory
@@ -106,3 +107,12 @@ formatter = generic
 [formatter_generic]
 format = %(levelname)-5.5s [%(name)s] %(message)s
 datefmt = %H:%M:%S
+
+
+[alembic]
+script_location = alembic
+version_locations = %(script_location)s/versions
+
+[schema_private]
+script_location = alembic_tenants
+version_locations = %(script_location)s/versions

backend/alembic/env.py

@@ -1,107 +1,215 @@
+from sqlalchemy.engine.base import Connection
+from typing import Literal
 import asyncio
 from logging.config import fileConfig
+import logging
 
 from alembic import context
+from sqlalchemy import pool
+from sqlalchemy.ext.asyncio import create_async_engine
+from sqlalchemy.sql import text
+from sqlalchemy.sql.schema import SchemaItem
+
+from shared_configs.configs import MULTI_TENANT
 from danswer.db.engine import build_connection_string
 from danswer.db.models import Base
-from sqlalchemy import pool
-from sqlalchemy.engine import Connection
-from sqlalchemy.ext.asyncio import create_async_engine
 from celery.backends.database.session import ResultModelBase  # type: ignore
-from sqlalchemy.schema import SchemaItem
+from danswer.db.engine import get_all_tenant_ids
+from shared_configs.configs import POSTGRES_DEFAULT_SCHEMA
 
-# this is the Alembic Config object, which provides
-# access to the values within the .ini file in use.
+# Alembic Config object
 config = context.config
 
 # Interpret the config file for Python logging.
-# This line sets up loggers basically.
 if config.config_file_name is not None and config.attributes.get(
     "configure_logger", True
 ):
     fileConfig(config.config_file_name)
 
-# add your model's MetaData object here
-# for 'autogenerate' support
-# from myapp import mymodel
-# target_metadata = mymodel.Base.metadata
+# Add your model's MetaData object here for 'autogenerate' support
 target_metadata = [Base.metadata, ResultModelBase.metadata]
 
-# other values from the config, defined by the needs of env.py,
-# can be acquired:
-# my_important_option = config.get_main_option("my_important_option")
-# ... etc.
-
 EXCLUDE_TABLES = {"kombu_queue", "kombu_message"}
 
+# Set up logging
+logger = logging.getLogger(__name__)
+
 
 def include_object(
     object: SchemaItem,
-    name: str,
-    type_: str,
+    name: str | None,
+    type_: Literal[
+        "schema",
+        "table",
+        "column",
+        "index",
+        "unique_constraint",
+        "foreign_key_constraint",
+    ],
     reflected: bool,
     compare_to: SchemaItem | None,
 ) -> bool:
+    """
+    Determines whether a database object should be included in migrations.
+    Excludes specified tables from migrations.
+    """
     if type_ == "table" and name in EXCLUDE_TABLES:
         return False
     return True
 
 
-def run_migrations_offline() -> None:
-    """Run migrations in 'offline' mode.
-
-    This configures the context with just a URL
-    and not an Engine, though an Engine is acceptable
-    here as well.  By skipping the Engine creation
-    we don't even need a DBAPI to be available.
-
-    Calls to context.execute() here emit the given string to the
-    script output.
-
+def get_schema_options() -> tuple[str, bool, bool]:
     """
-    url = build_connection_string()
-    context.configure(
-        url=url,
-        target_metadata=target_metadata,  # type: ignore
-        literal_binds=True,
-        dialect_opts={"paramstyle": "named"},
-    )
+    Parses command-line options passed via '-x' in Alembic commands.
+    Recognizes 'schema', 'create_schema', and 'upgrade_all_tenants' options.
+    """
+    x_args_raw = context.get_x_argument()
+    x_args = {}
+    for arg in x_args_raw:
+        for pair in arg.split(","):
+            if "=" in pair:
+                key, value = pair.split("=", 1)
+                x_args[key.strip()] = value.strip()
+    schema_name = x_args.get("schema", POSTGRES_DEFAULT_SCHEMA)
+    create_schema = x_args.get("create_schema", "true").lower() == "true"
+    upgrade_all_tenants = x_args.get("upgrade_all_tenants", "false").lower() == "true"
 
-    with context.begin_transaction():
-        context.run_migrations()
+    if (
+        MULTI_TENANT
+        and schema_name == POSTGRES_DEFAULT_SCHEMA
+        and not upgrade_all_tenants
+    ):
+        raise ValueError(
+            "Cannot run default migrations in public schema when multi-tenancy is enabled. "
+            "Please specify a tenant-specific schema."
+        )
+
+    return schema_name, create_schema, upgrade_all_tenants
 
 
-def do_run_migrations(connection: Connection) -> None:
+def do_run_migrations(
+    connection: Connection, schema_name: str, create_schema: bool
+) -> None:
+    """
+    Executes migrations in the specified schema.
+    """
+    logger.info(f"About to migrate schema: {schema_name}")
+
+    if create_schema:
+        connection.execute(text(f'CREATE SCHEMA IF NOT EXISTS "{schema_name}"'))
+        connection.execute(text("COMMIT"))
+
+    # Set search_path to the target schema
+    connection.execute(text(f'SET search_path TO "{schema_name}"'))
+
     context.configure(
         connection=connection,
         target_metadata=target_metadata,  # type: ignore
         include_object=include_object,
-    )  # type: ignore
+        version_table_schema=schema_name,
+        include_schemas=True,
+        compare_type=True,
+        compare_server_default=True,
+        script_location=config.get_main_option("script_location"),
+    )
 
     with context.begin_transaction():
         context.run_migrations()
 
 
 async def run_async_migrations() -> None:
-    """In this scenario we need to create an Engine
-    and associate a connection with the context.
-
     """
+    Determines whether to run migrations for a single schema or all schemas,
+    and executes migrations accordingly.
+    """
+    schema_name, create_schema, upgrade_all_tenants = get_schema_options()
 
-    connectable = create_async_engine(
+    engine = create_async_engine(
         build_connection_string(),
         poolclass=pool.NullPool,
     )
 
-    async with connectable.connect() as connection:
-        await connection.run_sync(do_run_migrations)
+    if upgrade_all_tenants:
+        # Run migrations for all tenant schemas sequentially
+        tenant_schemas = get_all_tenant_ids()
 
-    await connectable.dispose()
+        for schema in tenant_schemas:
+            try:
+                logger.info(f"Migrating schema: {schema}")
+                async with engine.connect() as connection:
+                    await connection.run_sync(
+                        do_run_migrations,
+                        schema_name=schema,
+                        create_schema=create_schema,
+                    )
+            except Exception as e:
+                logger.error(f"Error migrating schema {schema}: {e}")
+                raise
+    else:
+        try:
+            logger.info(f"Migrating schema: {schema_name}")
+            async with engine.connect() as connection:
+                await connection.run_sync(
+                    do_run_migrations,
+                    schema_name=schema_name,
+                    create_schema=create_schema,
+                )
+        except Exception as e:
+            logger.error(f"Error migrating schema {schema_name}: {e}")
+            raise
+
+    await engine.dispose()
+
+
+def run_migrations_offline() -> None:
+    """
+    Run migrations in 'offline' mode.
+    """
+    schema_name, _, upgrade_all_tenants = get_schema_options()
+    url = build_connection_string()
+
+    if upgrade_all_tenants:
+        # Run offline migrations for all tenant schemas
+        engine = create_async_engine(url)
+        tenant_schemas = get_all_tenant_ids()
+        engine.sync_engine.dispose()
+
+        for schema in tenant_schemas:
+            logger.info(f"Migrating schema: {schema}")
+            context.configure(
+                url=url,
+                target_metadata=target_metadata,  # type: ignore
+                literal_binds=True,
+                include_object=include_object,
+                version_table_schema=schema,
+                include_schemas=True,
+                script_location=config.get_main_option("script_location"),
+                dialect_opts={"paramstyle": "named"},
+            )
+
+            with context.begin_transaction():
+                context.run_migrations()
+    else:
+        logger.info(f"Migrating schema: {schema_name}")
+        context.configure(
+            url=url,
+            target_metadata=target_metadata,  # type: ignore
+            literal_binds=True,
+            include_object=include_object,
+            version_table_schema=schema_name,
+            include_schemas=True,
+            script_location=config.get_main_option("script_location"),
+            dialect_opts={"paramstyle": "named"},
+        )
+
+        with context.begin_transaction():
+            context.run_migrations()
 
 
 def run_migrations_online() -> None:
-    """Run migrations in 'online' mode."""
-
+    """
+    Runs migrations in 'online' mode using an asynchronous engine.
+    """
     asyncio.run(run_async_migrations())
 
 

backend/alembic/versions/0ebb1d516877_add_ccpair_deletion_failure_message.py

@@ -0,0 +1,27 @@
+"""add ccpair deletion failure message
+
+Revision ID: 0ebb1d516877
+Revises: 52a219fb5233
+Create Date: 2024-09-10 15:03:48.233926
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision = "0ebb1d516877"
+down_revision = "52a219fb5233"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "connector_credential_pair",
+        sa.Column("deletion_failure_message", sa.String(), nullable=True),
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("connector_credential_pair", "deletion_failure_message")

backend/alembic/versions/177de57c21c9_display_custom_llm_models.py

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

backend/alembic/versions/1b10e1fda030_add_additional_data_to_notifications.py

@@ -0,0 +1,26 @@
+"""add additional data to notifications
+
+Revision ID: 1b10e1fda030
+Revises: 6756efa39ada
+Create Date: 2024-10-15 19:26:44.071259
+
+"""
+from alembic import op
+import sqlalchemy as sa
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision = "1b10e1fda030"
+down_revision = "6756efa39ada"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "notification", sa.Column("additional_data", postgresql.JSONB(), nullable=True)
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("notification", "additional_data")

backend/alembic/versions/1b8206b29c5d_add_user_delete_cascades.py

@@ -0,0 +1,102 @@
+"""add_user_delete_cascades
+
+Revision ID: 1b8206b29c5d
+Revises: 35e6853a51d5
+Create Date: 2024-09-18 11:48:59.418726
+
+"""
+from alembic import op
+
+
+# revision identifiers, used by Alembic.
+revision = "1b8206b29c5d"
+down_revision = "35e6853a51d5"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.drop_constraint("credential_user_id_fkey", "credential", type_="foreignkey")
+    op.create_foreign_key(
+        "credential_user_id_fkey",
+        "credential",
+        "user",
+        ["user_id"],
+        ["id"],
+        ondelete="CASCADE",
+    )
+
+    op.drop_constraint("chat_session_user_id_fkey", "chat_session", type_="foreignkey")
+    op.create_foreign_key(
+        "chat_session_user_id_fkey",
+        "chat_session",
+        "user",
+        ["user_id"],
+        ["id"],
+        ondelete="CASCADE",
+    )
+
+    op.drop_constraint("chat_folder_user_id_fkey", "chat_folder", type_="foreignkey")
+    op.create_foreign_key(
+        "chat_folder_user_id_fkey",
+        "chat_folder",
+        "user",
+        ["user_id"],
+        ["id"],
+        ondelete="CASCADE",
+    )
+
+    op.drop_constraint("prompt_user_id_fkey", "prompt", type_="foreignkey")
+    op.create_foreign_key(
+        "prompt_user_id_fkey", "prompt", "user", ["user_id"], ["id"], ondelete="CASCADE"
+    )
+
+    op.drop_constraint("notification_user_id_fkey", "notification", type_="foreignkey")
+    op.create_foreign_key(
+        "notification_user_id_fkey",
+        "notification",
+        "user",
+        ["user_id"],
+        ["id"],
+        ondelete="CASCADE",
+    )
+
+    op.drop_constraint("inputprompt_user_id_fkey", "inputprompt", type_="foreignkey")
+    op.create_foreign_key(
+        "inputprompt_user_id_fkey",
+        "inputprompt",
+        "user",
+        ["user_id"],
+        ["id"],
+        ondelete="CASCADE",
+    )
+
+
+def downgrade() -> None:
+    op.drop_constraint("credential_user_id_fkey", "credential", type_="foreignkey")
+    op.create_foreign_key(
+        "credential_user_id_fkey", "credential", "user", ["user_id"], ["id"]
+    )
+
+    op.drop_constraint("chat_session_user_id_fkey", "chat_session", type_="foreignkey")
+    op.create_foreign_key(
+        "chat_session_user_id_fkey", "chat_session", "user", ["user_id"], ["id"]
+    )
+
+    op.drop_constraint("chat_folder_user_id_fkey", "chat_folder", type_="foreignkey")
+    op.create_foreign_key(
+        "chat_folder_user_id_fkey", "chat_folder", "user", ["user_id"], ["id"]
+    )
+
+    op.drop_constraint("prompt_user_id_fkey", "prompt", type_="foreignkey")
+    op.create_foreign_key("prompt_user_id_fkey", "prompt", "user", ["user_id"], ["id"])
+
+    op.drop_constraint("notification_user_id_fkey", "notification", type_="foreignkey")
+    op.create_foreign_key(
+        "notification_user_id_fkey", "notification", "user", ["user_id"], ["id"]
+    )
+
+    op.drop_constraint("inputprompt_user_id_fkey", "inputprompt", type_="foreignkey")
+    op.create_foreign_key(
+        "inputprompt_user_id_fkey", "inputprompt", "user", ["user_id"], ["id"]
+    )

backend/alembic/versions/1f60f60c3401_embedding_model_search_settings.py

@@ -30,7 +30,7 @@ def upgrade() -> None:
     op.add_column(
         "search_settings",
         sa.Column(
-            "multipass_indexing", sa.Boolean(), nullable=False, server_default="true"
+            "multipass_indexing", sa.Boolean(), nullable=False, server_default="false"
         ),
     )
     op.add_column(

backend/alembic/versions/26b931506ecb_default_chosen_assistants_to_none.py

@@ -0,0 +1,68 @@
+"""default chosen assistants to none
+
+Revision ID: 26b931506ecb
+Revises: 2daa494a0851
+Create Date: 2024-11-12 13:23:29.858995
+
+"""
+from alembic import op
+import sqlalchemy as sa
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision = "26b931506ecb"
+down_revision = "2daa494a0851"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "user", sa.Column("chosen_assistants_new", postgresql.JSONB(), nullable=True)
+    )
+
+    op.execute(
+        """
+    UPDATE "user"
+    SET chosen_assistants_new =
+        CASE
+            WHEN chosen_assistants = '[-2, -1, 0]' THEN NULL
+            ELSE chosen_assistants
+        END
+    """
+    )
+
+    op.drop_column("user", "chosen_assistants")
+
+    op.alter_column(
+        "user", "chosen_assistants_new", new_column_name="chosen_assistants"
+    )
+
+
+def downgrade() -> None:
+    op.add_column(
+        "user",
+        sa.Column(
+            "chosen_assistants_old",
+            postgresql.JSONB(),
+            nullable=False,
+            server_default="[-2, -1, 0]",
+        ),
+    )
+
+    op.execute(
+        """
+    UPDATE "user"
+    SET chosen_assistants_old =
+        CASE
+            WHEN chosen_assistants IS NULL THEN '[-2, -1, 0]'::jsonb
+            ELSE chosen_assistants
+        END
+    """
+    )
+
+    op.drop_column("user", "chosen_assistants")
+
+    op.alter_column(
+        "user", "chosen_assistants_old", new_column_name="chosen_assistants"
+    )

backend/alembic/versions/2daa494a0851_add_group_sync_time.py

@@ -0,0 +1,30 @@
+"""add-group-sync-time
+
+Revision ID: 2daa494a0851
+Revises: c0fd6e4da83a
+Create Date: 2024-11-11 10:57:22.991157
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+# revision identifiers, used by Alembic.
+revision = "2daa494a0851"
+down_revision = "c0fd6e4da83a"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "connector_credential_pair",
+        sa.Column(
+            "last_time_external_group_sync",
+            sa.DateTime(timezone=True),
+            nullable=True,
+        ),
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("connector_credential_pair", "last_time_external_group_sync")

backend/alembic/versions/33cb72ea4d80_single_tool_call_per_message.py

@@ -0,0 +1,50 @@
+"""single tool call per message
+
+Revision ID: 33cb72ea4d80
+Revises: 5b29123cd710
+Create Date: 2024-11-01 12:51:01.535003
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision = "33cb72ea4d80"
+down_revision = "5b29123cd710"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # Step 1: Delete extraneous ToolCall entries
+    # Keep only the ToolCall with the smallest 'id' for each 'message_id'
+    op.execute(
+        sa.text(
+            """
+            DELETE FROM tool_call
+            WHERE id NOT IN (
+                SELECT MIN(id)
+                FROM tool_call
+                WHERE message_id IS NOT NULL
+                GROUP BY message_id
+            );
+        """
+        )
+    )
+
+    # Step 2: Add a unique constraint on message_id
+    op.create_unique_constraint(
+        constraint_name="uq_tool_call_message_id",
+        table_name="tool_call",
+        columns=["message_id"],
+    )
+
+
+def downgrade() -> None:
+    # Step 1: Drop the unique constraint on message_id
+    op.drop_constraint(
+        constraint_name="uq_tool_call_message_id",
+        table_name="tool_call",
+        type_="unique",
+    )

backend/alembic/versions/35e6853a51d5_server_default_chosen_assistants.py

@@ -0,0 +1,64 @@
+"""server default chosen assistants
+
+Revision ID: 35e6853a51d5
+Revises: c99d76fcd298
+Create Date: 2024-09-13 13:20:32.885317
+
+"""
+from alembic import op
+import sqlalchemy as sa
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision = "35e6853a51d5"
+down_revision = "c99d76fcd298"
+branch_labels = None
+depends_on = None
+
+DEFAULT_ASSISTANTS = [-2, -1, 0]
+
+
+def upgrade() -> None:
+    # Step 1: Update any NULL values to the default value
+    # This upgrades existing users without ordered assistant
+    # to have default assistants set to visible assistants which are
+    # accessible by them.
+    op.execute(
+        """
+        UPDATE "user" u
+        SET chosen_assistants = (
+            SELECT jsonb_agg(
+                p.id ORDER BY
+                    COALESCE(p.display_priority, 2147483647) ASC,
+                    p.id ASC
+            )
+            FROM persona p
+            LEFT JOIN persona__user pu ON p.id = pu.persona_id AND pu.user_id = u.id
+            WHERE p.is_visible = true
+            AND (p.is_public = true OR pu.user_id IS NOT NULL)
+        )
+        WHERE chosen_assistants IS NULL
+        OR chosen_assistants = 'null'
+        OR jsonb_typeof(chosen_assistants) = 'null'
+        OR (jsonb_typeof(chosen_assistants) = 'string' AND chosen_assistants = '"null"')
+    """
+    )
+
+    # Step 2: Alter the column to make it non-nullable
+    op.alter_column(
+        "user",
+        "chosen_assistants",
+        type_=postgresql.JSONB(astext_type=sa.Text()),
+        nullable=False,
+        server_default=sa.text(f"'{DEFAULT_ASSISTANTS}'::jsonb"),
+    )
+
+
+def downgrade() -> None:
+    op.alter_column(
+        "user",
+        "chosen_assistants",
+        type_=postgresql.JSONB(astext_type=sa.Text()),
+        nullable=True,
+        server_default=None,
+    )

backend/alembic/versions/46b7a812670f_fix_user__external_user_group_id_fk.py

@@ -0,0 +1,46 @@
+"""fix_user__external_user_group_id_fk
+
+Revision ID: 46b7a812670f
+Revises: f32615f71aeb
+Create Date: 2024-09-23 12:58:03.894038
+
+"""
+from alembic import op
+
+# revision identifiers, used by Alembic.
+revision = "46b7a812670f"
+down_revision = "f32615f71aeb"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # Drop the existing primary key
+    op.drop_constraint(
+        "user__external_user_group_id_pkey",
+        "user__external_user_group_id",
+        type_="primary",
+    )
+
+    # Add the new composite primary key
+    op.create_primary_key(
+        "user__external_user_group_id_pkey",
+        "user__external_user_group_id",
+        ["user_id", "external_user_group_id", "cc_pair_id"],
+    )
+
+
+def downgrade() -> None:
+    # Drop the composite primary key
+    op.drop_constraint(
+        "user__external_user_group_id_pkey",
+        "user__external_user_group_id",
+        type_="primary",
+    )
+    # Delete all entries from the table
+    op.execute("DELETE FROM user__external_user_group_id")
+
+    # Recreate the original primary key on user_id
+    op.create_primary_key(
+        "user__external_user_group_id_pkey", "user__external_user_group_id", ["user_id"]
+    )

backend/alembic/versions/47e5bef3a1d7_add_persona_categories.py

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

backend/alembic/versions/4ee1287bd26a_add_multiple_slack_bot_support.py

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

backend/alembic/versions/52a219fb5233_add_last_synced_and_last_modified_to_document_table.py

@@ -1,7 +1,7 @@
 """Add last synced and last modified to document table
 
 Revision ID: 52a219fb5233
-Revises: f17bf3b0d9f1
+Revises: f7e58d357687
 Create Date: 2024-08-28 17:40:46.077470
 
 """

backend/alembic/versions/55546a7967ee_assistant_rework.py

@@ -0,0 +1,79 @@
+"""assistant_rework
+
+Revision ID: 55546a7967ee
+Revises: 61ff3651add4
+Create Date: 2024-09-18 17:00:23.755399
+
+"""
+from alembic import op
+import sqlalchemy as sa
+from sqlalchemy.dialects import postgresql
+
+
+# revision identifiers, used by Alembic.
+revision = "55546a7967ee"
+down_revision = "61ff3651add4"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # Reworking persona and user tables for new assistant features
+    # keep track of user's chosen assistants separate from their `ordering`
+    op.add_column("persona", sa.Column("builtin_persona", sa.Boolean(), nullable=True))
+    op.execute("UPDATE persona SET builtin_persona = default_persona")
+    op.alter_column("persona", "builtin_persona", nullable=False)
+    op.drop_index("_default_persona_name_idx", table_name="persona")
+    op.create_index(
+        "_builtin_persona_name_idx",
+        "persona",
+        ["name"],
+        unique=True,
+        postgresql_where=sa.text("builtin_persona = true"),
+    )
+
+    op.add_column(
+        "user", sa.Column("visible_assistants", postgresql.JSONB(), nullable=True)
+    )
+    op.add_column(
+        "user", sa.Column("hidden_assistants", postgresql.JSONB(), nullable=True)
+    )
+    op.execute(
+        "UPDATE \"user\" SET visible_assistants = '[]'::jsonb, hidden_assistants = '[]'::jsonb"
+    )
+    op.alter_column(
+        "user",
+        "visible_assistants",
+        nullable=False,
+        server_default=sa.text("'[]'::jsonb"),
+    )
+    op.alter_column(
+        "user",
+        "hidden_assistants",
+        nullable=False,
+        server_default=sa.text("'[]'::jsonb"),
+    )
+    op.drop_column("persona", "default_persona")
+    op.add_column(
+        "persona", sa.Column("is_default_persona", sa.Boolean(), nullable=True)
+    )
+
+
+def downgrade() -> None:
+    # Reverting changes made in upgrade
+    op.drop_column("user", "hidden_assistants")
+    op.drop_column("user", "visible_assistants")
+    op.drop_index("_builtin_persona_name_idx", table_name="persona")
+
+    op.drop_column("persona", "is_default_persona")
+    op.add_column("persona", sa.Column("default_persona", sa.Boolean(), nullable=True))
+    op.execute("UPDATE persona SET default_persona = builtin_persona")
+    op.alter_column("persona", "default_persona", nullable=False)
+    op.drop_column("persona", "builtin_persona")
+    op.create_index(
+        "_default_persona_name_idx",
+        "persona",
+        ["name"],
+        unique=True,
+        postgresql_where=sa.text("default_persona = true"),
+    )

backend/alembic/versions/5b29123cd710_nullable_search_settings_for_historic_.py

@@ -0,0 +1,70 @@
+"""nullable search settings for historic index attempts
+
+Revision ID: 5b29123cd710
+Revises: 949b4a92a401
+Create Date: 2024-10-30 19:37:59.630704
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision = "5b29123cd710"
+down_revision = "949b4a92a401"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # Drop the existing foreign key constraint
+    op.drop_constraint(
+        "fk_index_attempt_search_settings", "index_attempt", type_="foreignkey"
+    )
+
+    # Modify the column to be nullable
+    op.alter_column(
+        "index_attempt", "search_settings_id", existing_type=sa.INTEGER(), nullable=True
+    )
+
+    # Add back the foreign key with ON DELETE SET NULL
+    op.create_foreign_key(
+        "fk_index_attempt_search_settings",
+        "index_attempt",
+        "search_settings",
+        ["search_settings_id"],
+        ["id"],
+        ondelete="SET NULL",
+    )
+
+
+def downgrade() -> None:
+    # Warning: This will delete all index attempts that don't have search settings
+    op.execute(
+        """
+        DELETE FROM index_attempt
+        WHERE search_settings_id IS NULL
+    """
+    )
+
+    # Drop foreign key constraint
+    op.drop_constraint(
+        "fk_index_attempt_search_settings", "index_attempt", type_="foreignkey"
+    )
+
+    # Modify the column to be not nullable
+    op.alter_column(
+        "index_attempt",
+        "search_settings_id",
+        existing_type=sa.INTEGER(),
+        nullable=False,
+    )
+
+    # Add back the foreign key without ON DELETE SET NULL
+    op.create_foreign_key(
+        "fk_index_attempt_search_settings",
+        "index_attempt",
+        "search_settings",
+        ["search_settings_id"],
+        ["id"],
+    )

backend/alembic/versions/5c7fdadae813_match_any_keywords_flag_for_standard_.py

@@ -0,0 +1,35 @@
+"""match_any_keywords flag for standard answers
+
+Revision ID: 5c7fdadae813
+Revises: efb35676026c
+Create Date: 2024-09-13 18:52:59.256478
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+# revision identifiers, used by Alembic.
+revision = "5c7fdadae813"
+down_revision = "efb35676026c"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.add_column(
+        "standard_answer",
+        sa.Column(
+            "match_any_keywords",
+            sa.Boolean(),
+            nullable=False,
+            server_default=sa.false(),
+        ),
+    )
+    # ### end Alembic commands ###
+
+
+def downgrade() -> None:
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_column("standard_answer", "match_any_keywords")
+    # ### end Alembic commands ###

backend/alembic/versions/5d12a446f5c0_add_api_version_and_deployment_name_to_.py

@@ -0,0 +1,30 @@
+"""add api_version and deployment_name to search settings
+
+Revision ID: 5d12a446f5c0
+Revises: e4334d5b33ba
+Create Date: 2024-10-08 15:56:07.975636
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision = "5d12a446f5c0"
+down_revision = "e4334d5b33ba"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "embedding_provider", sa.Column("api_version", sa.String(), nullable=True)
+    )
+    op.add_column(
+        "embedding_provider", sa.Column("deployment_name", sa.String(), nullable=True)
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("embedding_provider", "deployment_name")
+    op.drop_column("embedding_provider", "api_version")

backend/alembic/versions/61ff3651add4_add_permission_syncing.py

@@ -0,0 +1,162 @@
+"""Add Permission Syncing
+
+Revision ID: 61ff3651add4
+Revises: 1b8206b29c5d
+Create Date: 2024-09-05 13:57:11.770413
+
+"""
+import fastapi_users_db_sqlalchemy
+
+from alembic import op
+import sqlalchemy as sa
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision = "61ff3651add4"
+down_revision = "1b8206b29c5d"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # Admin user who set up connectors will lose access to the docs temporarily
+    # only way currently to give back access is to rerun from beginning
+    op.add_column(
+        "connector_credential_pair",
+        sa.Column(
+            "access_type",
+            sa.String(),
+            nullable=True,
+        ),
+    )
+    op.execute(
+        "UPDATE connector_credential_pair SET access_type = 'PUBLIC' WHERE is_public = true"
+    )
+    op.execute(
+        "UPDATE connector_credential_pair SET access_type = 'PRIVATE' WHERE is_public = false"
+    )
+    op.alter_column("connector_credential_pair", "access_type", nullable=False)
+
+    op.add_column(
+        "connector_credential_pair",
+        sa.Column(
+            "auto_sync_options",
+            postgresql.JSONB(astext_type=sa.Text()),
+            nullable=True,
+        ),
+    )
+    op.add_column(
+        "connector_credential_pair",
+        sa.Column("last_time_perm_sync", sa.DateTime(timezone=True), nullable=True),
+    )
+    op.drop_column("connector_credential_pair", "is_public")
+
+    op.add_column(
+        "document",
+        sa.Column("external_user_emails", postgresql.ARRAY(sa.String()), nullable=True),
+    )
+    op.add_column(
+        "document",
+        sa.Column(
+            "external_user_group_ids", postgresql.ARRAY(sa.String()), nullable=True
+        ),
+    )
+    op.add_column(
+        "document",
+        sa.Column("is_public", sa.Boolean(), nullable=True),
+    )
+
+    op.create_table(
+        "user__external_user_group_id",
+        sa.Column(
+            "user_id", fastapi_users_db_sqlalchemy.generics.GUID(), nullable=False
+        ),
+        sa.Column("external_user_group_id", sa.String(), nullable=False),
+        sa.Column("cc_pair_id", sa.Integer(), nullable=False),
+        sa.PrimaryKeyConstraint("user_id"),
+    )
+
+    op.drop_column("external_permission", "user_id")
+    op.drop_column("email_to_external_user_cache", "user_id")
+    op.drop_table("permission_sync_run")
+    op.drop_table("external_permission")
+    op.drop_table("email_to_external_user_cache")
+
+
+def downgrade() -> None:
+    op.add_column(
+        "connector_credential_pair",
+        sa.Column("is_public", sa.BOOLEAN(), nullable=True),
+    )
+    op.execute(
+        "UPDATE connector_credential_pair SET is_public = (access_type = 'PUBLIC')"
+    )
+    op.alter_column("connector_credential_pair", "is_public", nullable=False)
+
+    op.drop_column("connector_credential_pair", "auto_sync_options")
+    op.drop_column("connector_credential_pair", "access_type")
+    op.drop_column("connector_credential_pair", "last_time_perm_sync")
+    op.drop_column("document", "external_user_emails")
+    op.drop_column("document", "external_user_group_ids")
+    op.drop_column("document", "is_public")
+
+    op.drop_table("user__external_user_group_id")
+
+    # Drop the enum type at the end of the downgrade
+    op.create_table(
+        "permission_sync_run",
+        sa.Column("id", sa.Integer(), nullable=False),
+        sa.Column(
+            "source_type",
+            sa.String(),
+            nullable=False,
+        ),
+        sa.Column("update_type", sa.String(), nullable=False),
+        sa.Column("cc_pair_id", sa.Integer(), nullable=True),
+        sa.Column(
+            "status",
+            sa.String(),
+            nullable=False,
+        ),
+        sa.Column("error_msg", sa.Text(), nullable=True),
+        sa.Column(
+            "updated_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.text("now()"),
+            nullable=False,
+        ),
+        sa.ForeignKeyConstraint(
+            ["cc_pair_id"],
+            ["connector_credential_pair.id"],
+        ),
+        sa.PrimaryKeyConstraint("id"),
+    )
+    op.create_table(
+        "external_permission",
+        sa.Column("id", sa.Integer(), nullable=False),
+        sa.Column("user_id", sa.UUID(), nullable=True),
+        sa.Column("user_email", sa.String(), nullable=False),
+        sa.Column(
+            "source_type",
+            sa.String(),
+            nullable=False,
+        ),
+        sa.Column("external_permission_group", sa.String(), nullable=False),
+        sa.ForeignKeyConstraint(
+            ["user_id"],
+            ["user.id"],
+        ),
+        sa.PrimaryKeyConstraint("id"),
+    )
+    op.create_table(
+        "email_to_external_user_cache",
+        sa.Column("id", sa.Integer(), nullable=False),
+        sa.Column("external_user_id", sa.String(), nullable=False),
+        sa.Column("user_id", sa.UUID(), nullable=True),
+        sa.Column("user_email", sa.String(), nullable=False),
+        sa.ForeignKeyConstraint(
+            ["user_id"],
+            ["user.id"],
+        ),
+        sa.PrimaryKeyConstraint("id"),
+    )

backend/alembic/versions/6756efa39ada_id_uuid_for_chat_session.py

@@ -0,0 +1,153 @@
+"""Migrate chat_session and chat_message tables to use UUID primary keys
+
+Revision ID: 6756efa39ada
+Revises: 5d12a446f5c0
+Create Date: 2024-10-15 17:47:44.108537
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+revision = "6756efa39ada"
+down_revision = "5d12a446f5c0"
+branch_labels = None
+depends_on = None
+
+"""
+This script:
+1. Adds UUID columns to chat_session and chat_message
+2. Populates new columns with UUIDs
+3. Updates foreign key relationships
+4. Removes old integer ID columns
+
+Note: Downgrade will assign new integer IDs, not restore original ones.
+"""
+
+
+def upgrade() -> None:
+    op.execute("CREATE EXTENSION IF NOT EXISTS pgcrypto;")
+
+    op.add_column(
+        "chat_session",
+        sa.Column(
+            "new_id",
+            sa.UUID(as_uuid=True),
+            server_default=sa.text("gen_random_uuid()"),
+            nullable=False,
+        ),
+    )
+
+    op.execute("UPDATE chat_session SET new_id = gen_random_uuid();")
+
+    op.add_column(
+        "chat_message",
+        sa.Column("new_chat_session_id", sa.UUID(as_uuid=True), nullable=True),
+    )
+
+    op.execute(
+        """
+        UPDATE chat_message
+        SET new_chat_session_id = cs.new_id
+        FROM chat_session cs
+        WHERE chat_message.chat_session_id = cs.id;
+        """
+    )
+
+    op.drop_constraint(
+        "chat_message_chat_session_id_fkey", "chat_message", type_="foreignkey"
+    )
+
+    op.drop_column("chat_message", "chat_session_id")
+    op.alter_column(
+        "chat_message", "new_chat_session_id", new_column_name="chat_session_id"
+    )
+
+    op.drop_constraint("chat_session_pkey", "chat_session", type_="primary")
+    op.drop_column("chat_session", "id")
+    op.alter_column("chat_session", "new_id", new_column_name="id")
+
+    op.create_primary_key("chat_session_pkey", "chat_session", ["id"])
+
+    op.create_foreign_key(
+        "chat_message_chat_session_id_fkey",
+        "chat_message",
+        "chat_session",
+        ["chat_session_id"],
+        ["id"],
+        ondelete="CASCADE",
+    )
+
+
+def downgrade() -> None:
+    op.drop_constraint(
+        "chat_message_chat_session_id_fkey", "chat_message", type_="foreignkey"
+    )
+
+    op.add_column(
+        "chat_session",
+        sa.Column("old_id", sa.Integer, autoincrement=True, nullable=True),
+    )
+
+    op.execute("CREATE SEQUENCE chat_session_old_id_seq OWNED BY chat_session.old_id;")
+    op.execute(
+        "ALTER TABLE chat_session ALTER COLUMN old_id SET DEFAULT nextval('chat_session_old_id_seq');"
+    )
+
+    op.execute(
+        "UPDATE chat_session SET old_id = nextval('chat_session_old_id_seq') WHERE old_id IS NULL;"
+    )
+
+    op.alter_column("chat_session", "old_id", nullable=False)
+
+    op.drop_constraint("chat_session_pkey", "chat_session", type_="primary")
+    op.create_primary_key("chat_session_pkey", "chat_session", ["old_id"])
+
+    op.add_column(
+        "chat_message",
+        sa.Column("old_chat_session_id", sa.Integer, nullable=True),
+    )
+
+    op.execute(
+        """
+        UPDATE chat_message
+        SET old_chat_session_id = cs.old_id
+        FROM chat_session cs
+        WHERE chat_message.chat_session_id = cs.id;
+        """
+    )
+
+    op.drop_column("chat_message", "chat_session_id")
+    op.alter_column(
+        "chat_message", "old_chat_session_id", new_column_name="chat_session_id"
+    )
+
+    op.create_foreign_key(
+        "chat_message_chat_session_id_fkey",
+        "chat_message",
+        "chat_session",
+        ["chat_session_id"],
+        ["old_id"],
+        ondelete="CASCADE",
+    )
+
+    op.drop_column("chat_session", "id")
+    op.alter_column("chat_session", "old_id", new_column_name="id")
+
+    op.alter_column(
+        "chat_session",
+        "id",
+        type_=sa.Integer(),
+        existing_type=sa.Integer(),
+        existing_nullable=False,
+        existing_server_default=False,
+    )
+
+    # Rename the sequence
+    op.execute("ALTER SEQUENCE chat_session_old_id_seq RENAME TO chat_session_id_seq;")
+
+    # Update the default value to use the renamed sequence
+    op.alter_column(
+        "chat_session",
+        "id",
+        server_default=sa.text("nextval('chat_session_id_seq'::regclass)"),
+    )

backend/alembic/versions/6d562f86c78b_remove_default_bot.py

@@ -0,0 +1,45 @@
+"""remove default bot
+
+Revision ID: 6d562f86c78b
+Revises: 177de57c21c9
+Create Date: 2024-11-22 11:51:29.331336
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+# revision identifiers, used by Alembic.
+revision = "6d562f86c78b"
+down_revision = "177de57c21c9"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.execute(
+        sa.text(
+            """
+            DELETE FROM slack_bot
+            WHERE name = 'Default Bot'
+            AND bot_token = ''
+            AND app_token = ''
+            AND NOT EXISTS (
+                SELECT 1 FROM slack_channel_config
+                WHERE slack_channel_config.slack_bot_id = slack_bot.id
+            )
+            """
+        )
+    )
+
+
+def downgrade() -> None:
+    op.execute(
+        sa.text(
+            """
+            INSERT INTO slack_bot (name, enabled, bot_token, app_token)
+            SELECT 'Default Bot', true, '', ''
+            WHERE NOT EXISTS (SELECT 1 FROM slack_bot)
+            RETURNING id;
+            """
+        )
+    )

backend/alembic/versions/703313b75876_add_tokenratelimit_tables.py

@@ -9,7 +9,7 @@ import json
 from typing import cast
 from alembic import op
 import sqlalchemy as sa
-from danswer.dynamic_configs.factory import get_dynamic_config_store
+from danswer.key_value_store.factory import get_kv_store
 
 # revision identifiers, used by Alembic.
 revision = "703313b75876"
@@ -54,9 +54,7 @@ def upgrade() -> None:
     )
 
     try:
-        settings_json = cast(
-            str, get_dynamic_config_store().load("token_budget_settings")
-        )
+        settings_json = cast(str, get_kv_store().load("token_budget_settings"))
         settings = json.loads(settings_json)
 
         is_enabled = settings.get("enable_token_budget", False)
@@ -71,7 +69,7 @@ def upgrade() -> None:
             )
 
         # Delete the dynamic config
-        get_dynamic_config_store().delete("token_budget_settings")
+        get_kv_store().delete("token_budget_settings")
 
     except Exception:
         # Ignore if the dynamic config is not found

backend/alembic/versions/776b3bbe9092_remove_remaining_enums.py

@@ -9,8 +9,8 @@ from alembic import op
 import sqlalchemy as sa
 
 from danswer.db.models import IndexModelStatus
-from danswer.search.enums import RecencyBiasSetting
-from danswer.search.enums import SearchType
+from danswer.context.search.enums import RecencyBiasSetting
+from danswer.context.search.enums import SearchType
 
 # revision identifiers, used by Alembic.
 revision = "776b3bbe9092"

backend/alembic/versions/797089dfb4d2_persona_start_date.py

@@ -0,0 +1,27 @@
+"""persona_start_date
+
+Revision ID: 797089dfb4d2
+Revises: 55546a7967ee
+Create Date: 2024-09-11 14:51:49.785835
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision = "797089dfb4d2"
+down_revision = "55546a7967ee"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "persona",
+        sa.Column("search_start_date", sa.DateTime(timezone=True), nullable=True),
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("persona", "search_start_date")

backend/alembic/versions/93560ba1b118_add_web_ui_option_to_slack_config.py

@@ -0,0 +1,35 @@
+"""add web ui option to slack config
+
+Revision ID: 93560ba1b118
+Revises: 6d562f86c78b
+Create Date: 2024-11-24 06:36:17.490612
+
+"""
+from alembic import op
+
+# revision identifiers, used by Alembic.
+revision = "93560ba1b118"
+down_revision = "6d562f86c78b"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # Add show_continue_in_web_ui with default False to all existing channel_configs
+    op.execute(
+        """
+        UPDATE slack_channel_config
+        SET channel_config = channel_config || '{"show_continue_in_web_ui": false}'::jsonb
+        WHERE NOT channel_config ? 'show_continue_in_web_ui'
+        """
+    )
+
+
+def downgrade() -> None:
+    # Remove show_continue_in_web_ui from all channel_configs
+    op.execute(
+        """
+        UPDATE slack_channel_config
+        SET channel_config = channel_config - 'show_continue_in_web_ui'
+        """
+    )

backend/alembic/versions/949b4a92a401_remove_rt.py

@@ -0,0 +1,72 @@
+"""remove rt
+
+Revision ID: 949b4a92a401
+Revises: 1b10e1fda030
+Create Date: 2024-10-26 13:06:06.937969
+
+"""
+from alembic import op
+from sqlalchemy.orm import Session
+from sqlalchemy import text
+
+# Import your models and constants
+from danswer.db.models import (
+    Connector,
+    ConnectorCredentialPair,
+    Credential,
+    IndexAttempt,
+)
+
+
+# revision identifiers, used by Alembic.
+revision = "949b4a92a401"
+down_revision = "1b10e1fda030"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # Deletes all RequestTracker connectors and associated data
+    bind = op.get_bind()
+    session = Session(bind=bind)
+
+    # Get connectors using raw SQL
+    result = bind.execute(
+        text("SELECT id FROM connector WHERE source = 'requesttracker'")
+    )
+    connector_ids = [row[0] for row in result]
+
+    if connector_ids:
+        cc_pairs_to_delete = (
+            session.query(ConnectorCredentialPair)
+            .filter(ConnectorCredentialPair.connector_id.in_(connector_ids))
+            .all()
+        )
+
+        cc_pair_ids = [cc_pair.id for cc_pair in cc_pairs_to_delete]
+
+        if cc_pair_ids:
+            session.query(IndexAttempt).filter(
+                IndexAttempt.connector_credential_pair_id.in_(cc_pair_ids)
+            ).delete(synchronize_session=False)
+
+            session.query(ConnectorCredentialPair).filter(
+                ConnectorCredentialPair.id.in_(cc_pair_ids)
+            ).delete(synchronize_session=False)
+
+        credential_ids = [cc_pair.credential_id for cc_pair in cc_pairs_to_delete]
+        if credential_ids:
+            session.query(Credential).filter(Credential.id.in_(credential_ids)).delete(
+                synchronize_session=False
+            )
+
+        session.query(Connector).filter(Connector.id.in_(connector_ids)).delete(
+            synchronize_session=False
+        )
+
+    session.commit()
+
+
+def downgrade() -> None:
+    # No-op downgrade as we cannot restore deleted data
+    pass

backend/alembic/versions/9cf5c00f72fe_add_creator_to_cc_pair.py

@@ -0,0 +1,30 @@
+"""add creator to cc pair
+
+Revision ID: 9cf5c00f72fe
+Revises: 26b931506ecb
+Create Date: 2024-11-12 15:16:42.682902
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+# revision identifiers, used by Alembic.
+revision = "9cf5c00f72fe"
+down_revision = "26b931506ecb"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "connector_credential_pair",
+        sa.Column(
+            "creator_id",
+            sa.UUID(as_uuid=True),
+            nullable=True,
+        ),
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("connector_credential_pair", "creator_id")

backend/alembic/versions/9f696734098f_combine_search_and_chat.py

@@ -0,0 +1,36 @@
+"""Combine Search and Chat
+
+Revision ID: 9f696734098f
+Revises: a8c2065484e6
+Create Date: 2024-11-27 15:32:19.694972
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+# revision identifiers, used by Alembic.
+revision = "9f696734098f"
+down_revision = "a8c2065484e6"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.alter_column("chat_session", "description", nullable=True)
+    op.drop_column("chat_session", "one_shot")
+    op.drop_column("slack_channel_config", "response_type")
+
+
+def downgrade() -> None:
+    op.execute("UPDATE chat_session SET description = '' WHERE description IS NULL")
+    op.alter_column("chat_session", "description", nullable=False)
+    op.add_column(
+        "chat_session",
+        sa.Column("one_shot", sa.Boolean(), nullable=False, server_default=sa.false()),
+    )
+    op.add_column(
+        "slack_channel_config",
+        sa.Column(
+            "response_type", sa.String(), nullable=False, server_default="citations"
+        ),
+    )

backend/alembic/versions/a8c2065484e6_add_auto_scroll_to_user_model.py

@@ -0,0 +1,27 @@
+"""add auto scroll to user model
+
+Revision ID: a8c2065484e6
+Revises: abe7378b8217
+Create Date: 2024-11-22 17:34:09.690295
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision = "a8c2065484e6"
+down_revision = "abe7378b8217"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "user",
+        sa.Column("auto_scroll", sa.Boolean(), nullable=True, server_default=None),
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("user", "auto_scroll")

backend/alembic/versions/abe7378b8217_add_indexing_trigger_to_cc_pair.py

@@ -0,0 +1,30 @@
+"""add indexing trigger to cc_pair
+
+Revision ID: abe7378b8217
+Revises: 6d562f86c78b
+Create Date: 2024-11-26 19:09:53.481171
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+# revision identifiers, used by Alembic.
+revision = "abe7378b8217"
+down_revision = "93560ba1b118"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "connector_credential_pair",
+        sa.Column(
+            "indexing_trigger",
+            sa.Enum("UPDATE", "REINDEX", name="indexingmode", native_enum=False),
+            nullable=True,
+        ),
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("connector_credential_pair", "indexing_trigger")

backend/alembic/versions/ac5eaac849f9_add_last_pruned_to_connector_table.py

@@ -0,0 +1,27 @@
+"""add last_pruned to the connector_credential_pair table
+
+Revision ID: ac5eaac849f9
+Revises: 52a219fb5233
+Create Date: 2024-09-10 15:04:26.437118
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+# revision identifiers, used by Alembic.
+revision = "ac5eaac849f9"
+down_revision = "46b7a812670f"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # last pruned represents the last time the connector was pruned
+    op.add_column(
+        "connector_credential_pair",
+        sa.Column("last_pruned", sa.DateTime(timezone=True), nullable=True),
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("connector_credential_pair", "last_pruned")

backend/alembic/versions/b082fec533f0_make_last_attempt_status_nullable.py

@@ -31,6 +31,12 @@ def upgrade() -> None:
 
 
 def downgrade() -> None:
+    # First, update any null values to a default value
+    op.execute(
+        "UPDATE connector_credential_pair SET last_attempt_status = 'NOT_STARTED' WHERE last_attempt_status IS NULL"
+    )
+
+    # Then, make the column non-nullable
     op.alter_column(
         "connector_credential_pair",
         "last_attempt_status",

backend/alembic/versions/b156fa702355_chat_reworked.py

@@ -288,6 +288,15 @@ def upgrade() -> None:
 
 
 def downgrade() -> None:
+    # NOTE: you will lose all chat history. This is to satisfy the non-nullable constraints
+    # below
+    op.execute("DELETE FROM chat_feedback")
+    op.execute("DELETE FROM chat_message__search_doc")
+    op.execute("DELETE FROM document_retrieval_feedback")
+    op.execute("DELETE FROM document_retrieval_feedback")
+    op.execute("DELETE FROM chat_message")
+    op.execute("DELETE FROM chat_session")
+
     op.drop_constraint(
         "chat_feedback__chat_message_fk", "chat_feedback", type_="foreignkey"
     )

backend/alembic/versions/b72ed7a5db0e_remove_description_from_starter_messages.py

@@ -0,0 +1,48 @@
+"""remove description from starter messages
+
+Revision ID: b72ed7a5db0e
+Revises: 33cb72ea4d80
+Create Date: 2024-11-03 15:55:28.944408
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision = "b72ed7a5db0e"
+down_revision = "33cb72ea4d80"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.execute(
+        sa.text(
+            """
+            UPDATE persona
+            SET starter_messages = (
+                SELECT jsonb_agg(elem - 'description')
+                FROM jsonb_array_elements(starter_messages) elem
+            )
+            WHERE starter_messages IS NOT NULL
+              AND jsonb_typeof(starter_messages) = 'array'
+            """
+        )
+    )
+
+
+def downgrade() -> None:
+    op.execute(
+        sa.text(
+            """
+            UPDATE persona
+            SET starter_messages = (
+                SELECT jsonb_agg(elem || '{"description": ""}')
+                FROM jsonb_array_elements(starter_messages) elem
+            )
+            WHERE starter_messages IS NOT NULL
+              AND jsonb_typeof(starter_messages) = 'array'
+            """
+        )
+    )

backend/alembic/versions/bd2921608c3a_non_nullable_default_persona.py

@@ -0,0 +1,43 @@
+"""non nullable default persona
+
+Revision ID: bd2921608c3a
+Revises: 797089dfb4d2
+Create Date: 2024-09-20 10:28:37.992042
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision = "bd2921608c3a"
+down_revision = "797089dfb4d2"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # Set existing NULL values to False
+    op.execute(
+        "UPDATE persona SET is_default_persona = FALSE WHERE is_default_persona IS NULL"
+    )
+
+    # Alter the column to be not nullable with a default value of False
+    op.alter_column(
+        "persona",
+        "is_default_persona",
+        existing_type=sa.Boolean(),
+        nullable=False,
+        server_default=sa.text("false"),
+    )
+
+
+def downgrade() -> None:
+    # Revert the changes
+    op.alter_column(
+        "persona",
+        "is_default_persona",
+        existing_type=sa.Boolean(),
+        nullable=True,
+        server_default=None,
+    )

backend/alembic/versions/bf7a81109301_delete_input_prompts.py

@@ -0,0 +1,57 @@
+"""delete_input_prompts
+
+Revision ID: bf7a81109301
+Revises: f7a894b06d02
+Create Date: 2024-12-09 12:00:49.884228
+
+"""
+from alembic import op
+import sqlalchemy as sa
+import fastapi_users_db_sqlalchemy
+
+
+# revision identifiers, used by Alembic.
+revision = "bf7a81109301"
+down_revision = "f7a894b06d02"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.drop_table("inputprompt__user")
+    op.drop_table("inputprompt")
+
+
+def downgrade() -> None:
+    op.create_table(
+        "inputprompt",
+        sa.Column("id", sa.Integer(), autoincrement=True, nullable=False),
+        sa.Column("prompt", sa.String(), nullable=False),
+        sa.Column("content", sa.String(), nullable=False),
+        sa.Column("active", sa.Boolean(), nullable=False),
+        sa.Column("is_public", sa.Boolean(), nullable=False),
+        sa.Column(
+            "user_id",
+            fastapi_users_db_sqlalchemy.generics.GUID(),
+            nullable=True,
+        ),
+        sa.ForeignKeyConstraint(
+            ["user_id"],
+            ["user.id"],
+        ),
+        sa.PrimaryKeyConstraint("id"),
+    )
+    op.create_table(
+        "inputprompt__user",
+        sa.Column("input_prompt_id", sa.Integer(), nullable=False),
+        sa.Column("user_id", sa.Integer(), nullable=False),
+        sa.ForeignKeyConstraint(
+            ["input_prompt_id"],
+            ["inputprompt.id"],
+        ),
+        sa.ForeignKeyConstraint(
+            ["user_id"],
+            ["inputprompt.id"],
+        ),
+        sa.PrimaryKeyConstraint("input_prompt_id", "user_id"),
+    )

backend/alembic/versions/c0fd6e4da83a_add_recent_assistants.py

@@ -0,0 +1,29 @@
+"""add recent assistants
+
+Revision ID: c0fd6e4da83a
+Revises: b72ed7a5db0e
+Create Date: 2024-11-03 17:28:54.916618
+
+"""
+from alembic import op
+import sqlalchemy as sa
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision = "c0fd6e4da83a"
+down_revision = "b72ed7a5db0e"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "user",
+        sa.Column(
+            "recent_assistants", postgresql.JSONB(), server_default="[]", nullable=False
+        ),
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("user", "recent_assistants")

backend/alembic/versions/c99d76fcd298_add_nullable_to_persona_id_in_chat_.py

@@ -0,0 +1,81 @@
+"""add nullable to persona id in Chat Session
+
+Revision ID: c99d76fcd298
+Revises: 5c7fdadae813
+Create Date: 2024-07-09 19:27:01.579697
+
+"""
+
+from alembic import op
+import sqlalchemy as sa
+
+# revision identifiers, used by Alembic.
+revision = "c99d76fcd298"
+down_revision = "5c7fdadae813"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.alter_column(
+        "chat_session", "persona_id", existing_type=sa.INTEGER(), nullable=True
+    )
+
+
+def downgrade() -> None:
+    # Delete chat messages and feedback first since they reference chat sessions
+    # Get chat messages from sessions with null persona_id
+    chat_messages_query = """
+        SELECT id
+        FROM chat_message
+        WHERE chat_session_id IN (
+            SELECT id
+            FROM chat_session
+            WHERE persona_id IS NULL
+        )
+    """
+
+    # Delete dependent records first
+    op.execute(
+        f"""
+        DELETE FROM document_retrieval_feedback
+        WHERE chat_message_id IN (
+            {chat_messages_query}
+        )
+    """
+    )
+    op.execute(
+        f"""
+        DELETE FROM chat_message__search_doc
+        WHERE chat_message_id IN (
+            {chat_messages_query}
+        )
+    """
+    )
+
+    # Delete chat messages
+    op.execute(
+        """
+        DELETE FROM chat_message
+        WHERE chat_session_id IN (
+            SELECT id
+            FROM chat_session
+            WHERE persona_id IS NULL
+        )
+    """
+    )
+
+    # Now we can safely delete the chat sessions
+    op.execute(
+        """
+        DELETE FROM chat_session
+        WHERE persona_id IS NULL
+    """
+    )
+
+    op.alter_column(
+        "chat_session",
+        "persona_id",
+        existing_type=sa.INTEGER(),
+        nullable=False,
+    )

backend/alembic/versions/da4c21c69164_chosen_assistants_changed_to_jsonb.py

@@ -20,7 +20,7 @@ depends_on: None = None
 def upgrade() -> None:
     conn = op.get_bind()
     existing_ids_and_chosen_assistants = conn.execute(
-        sa.text("select id, chosen_assistants from public.user")
+        sa.text('select id, chosen_assistants from "user"')
     )
     op.drop_column(
         "user",
@@ -37,7 +37,7 @@ def upgrade() -> None:
     for id, chosen_assistants in existing_ids_and_chosen_assistants:
         conn.execute(
             sa.text(
-                "update public.user set chosen_assistants = :chosen_assistants where id = :id"
+                'update "user" set chosen_assistants = :chosen_assistants where id = :id'
             ),
             {"chosen_assistants": json.dumps(chosen_assistants), "id": id},
         )
@@ -46,7 +46,7 @@ def upgrade() -> None:
 def downgrade() -> None:
     conn = op.get_bind()
     existing_ids_and_chosen_assistants = conn.execute(
-        sa.text("select id, chosen_assistants from public.user")
+        sa.text('select id, chosen_assistants from "user"')
     )
     op.drop_column(
         "user",
@@ -59,7 +59,7 @@ def downgrade() -> None:
     for id, chosen_assistants in existing_ids_and_chosen_assistants:
         conn.execute(
             sa.text(
-                "update public.user set chosen_assistants = :chosen_assistants where id = :id"
+                'update "user" set chosen_assistants = :chosen_assistants where id = :id'
             ),
             {"chosen_assistants": chosen_assistants, "id": id},
         )

backend/alembic/versions/dfbe9e93d3c7_extended_role_for_non_web.py

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

backend/alembic/versions/e4334d5b33ba_add_deployment_name_to_llmprovider.py

@@ -0,0 +1,26 @@
+"""add_deployment_name_to_llmprovider
+
+Revision ID: e4334d5b33ba
+Revises: ac5eaac849f9
+Create Date: 2024-10-04 09:52:34.896867
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision = "e4334d5b33ba"
+down_revision = "ac5eaac849f9"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "llm_provider", sa.Column("deployment_name", sa.String(), nullable=True)
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("llm_provider", "deployment_name")

backend/alembic/versions/efb35676026c_standard_answer_match_regex_flag.py

@@ -0,0 +1,32 @@
+"""standard answer match_regex flag
+
+Revision ID: efb35676026c
+Revises: 0ebb1d516877
+Create Date: 2024-09-11 13:55:46.101149
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+# revision identifiers, used by Alembic.
+revision = "efb35676026c"
+down_revision = "0ebb1d516877"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.add_column(
+        "standard_answer",
+        sa.Column(
+            "match_regex", sa.Boolean(), nullable=False, server_default=sa.false()
+        ),
+    )
+    # ### end Alembic commands ###
+
+
+def downgrade() -> None:
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_column("standard_answer", "match_regex")
+    # ### end Alembic commands ###

backend/alembic/versions/f32615f71aeb_add_custom_headers_to_tools.py

@@ -0,0 +1,26 @@
+"""add custom headers to tools
+
+Revision ID: f32615f71aeb
+Revises: bd2921608c3a
+Create Date: 2024-09-12 20:26:38.932377
+
+"""
+from alembic import op
+import sqlalchemy as sa
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision = "f32615f71aeb"
+down_revision = "bd2921608c3a"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "tool", sa.Column("custom_headers", postgresql.JSONB(), nullable=True)
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("tool", "custom_headers")

backend/alembic/versions/f7a894b06d02_non_nullbale_slack_bot_id_in_channel_.py

@@ -0,0 +1,40 @@
+"""non-nullbale slack bot id in channel config
+
+Revision ID: f7a894b06d02
+Revises: 9f696734098f
+Create Date: 2024-12-06 12:55:42.845723
+
+"""
+
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision = "f7a894b06d02"
+down_revision = "9f696734098f"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # Delete all rows with null slack_bot_id
+    op.execute("DELETE FROM slack_channel_config WHERE slack_bot_id IS NULL")
+
+    # Make slack_bot_id non-nullable
+    op.alter_column(
+        "slack_channel_config",
+        "slack_bot_id",
+        existing_type=sa.Integer(),
+        nullable=False,
+    )
+
+
+def downgrade() -> None:
+    # Make slack_bot_id nullable again
+    op.alter_column(
+        "slack_channel_config",
+        "slack_bot_id",
+        existing_type=sa.Integer(),
+        nullable=True,
+    )

backend/alembic/versions/f7e58d357687_add_has_web_column_to_user.py

@@ -1,7 +1,7 @@
 """add has_web_login column to user
 
 Revision ID: f7e58d357687
-Revises: bceb1e139447
+Revises: ba98eba0f66a
 Create Date: 2024-09-07 20:20:54.522620
 
 """

backend/alembic_tenants/README.md

@@ -0,0 +1,3 @@
+These files are for public table migrations when operating with multi tenancy.
+
+If you are not a Danswer developer, you can ignore this directory entirely.

backend/alembic_tenants/env.py

@@ -0,0 +1,119 @@
+import asyncio
+from logging.config import fileConfig
+from typing import Literal
+
+from sqlalchemy import pool
+from sqlalchemy.engine import Connection
+from sqlalchemy.ext.asyncio import create_async_engine
+from sqlalchemy.schema import SchemaItem
+
+from alembic import context
+from danswer.db.engine import build_connection_string
+from danswer.db.models import PublicBase
+
+# this is the Alembic Config object, which provides
+# access to the values within the .ini file in use.
+config = context.config
+
+# Interpret the config file for Python logging.
+# This line sets up loggers basically.
+if config.config_file_name is not None and config.attributes.get(
+    "configure_logger", True
+):
+    fileConfig(config.config_file_name)
+
+# add your model's MetaData object here
+# for 'autogenerate' support
+# from myapp import mymodel
+# target_metadata = mymodel.Base.metadata
+target_metadata = [PublicBase.metadata]
+
+# other values from the config, defined by the needs of env.py,
+# can be acquired:
+# my_important_option = config.get_main_option("my_important_option")
+# ... etc.
+
+EXCLUDE_TABLES = {"kombu_queue", "kombu_message"}
+
+
+def include_object(
+    object: SchemaItem,
+    name: str | None,
+    type_: Literal[
+        "schema",
+        "table",
+        "column",
+        "index",
+        "unique_constraint",
+        "foreign_key_constraint",
+    ],
+    reflected: bool,
+    compare_to: SchemaItem | None,
+) -> bool:
+    if type_ == "table" and name in EXCLUDE_TABLES:
+        return False
+    return True
+
+
+def run_migrations_offline() -> None:
+    """Run migrations in 'offline' mode.
+
+    This configures the context with just a URL
+    and not an Engine, though an Engine is acceptable
+    here as well.  By skipping the Engine creation
+    we don't even need a DBAPI to be available.
+
+    Calls to context.execute() here emit the given string to the
+    script output.
+
+    """
+    url = build_connection_string()
+    context.configure(
+        url=url,
+        target_metadata=target_metadata,  # type: ignore
+        literal_binds=True,
+        dialect_opts={"paramstyle": "named"},
+    )
+
+    with context.begin_transaction():
+        context.run_migrations()
+
+
+def do_run_migrations(connection: Connection) -> None:
+    context.configure(
+        connection=connection,
+        target_metadata=target_metadata,  # type: ignore
+        include_object=include_object,
+    )  # type: ignore
+
+    with context.begin_transaction():
+        context.run_migrations()
+
+
+async def run_async_migrations() -> None:
+    """In this scenario we need to create an Engine
+    and associate a connection with the context.
+
+    """
+
+    connectable = create_async_engine(
+        build_connection_string(),
+        poolclass=pool.NullPool,
+    )
+
+    async with connectable.connect() as connection:
+        await connection.run_sync(do_run_migrations)
+
+    await connectable.dispose()
+
+
+def run_migrations_online() -> None:
+    """Run migrations in 'online' mode."""
+
+    asyncio.run(run_async_migrations())
+
+
+if context.is_offline_mode():
+    run_migrations_offline()
+else:
+    run_migrations_online()

backend/alembic_tenants/script.py.mako

@@ -0,0 +1,24 @@
+"""${message}
+
+Revision ID: ${up_revision}
+Revises: ${down_revision | comma,n}
+Create Date: ${create_date}
+
+"""
+from alembic import op
+import sqlalchemy as sa
+${imports if imports else ""}
+
+# revision identifiers, used by Alembic.
+revision = ${repr(up_revision)}
+down_revision = ${repr(down_revision)}
+branch_labels = ${repr(branch_labels)}
+depends_on = ${repr(depends_on)}
+
+
+def upgrade() -> None:
+    ${upgrades if upgrades else "pass"}
+
+
+def downgrade() -> None:
+    ${downgrades if downgrades else "pass"}

backend/alembic_tenants/versions/14a83a331951_create_usertenantmapping_table.py

@@ -0,0 +1,24 @@
+import sqlalchemy as sa
+
+from alembic import op
+
+# revision identifiers, used by Alembic.
+revision = "14a83a331951"
+down_revision = None
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.create_table(
+        "user_tenant_mapping",
+        sa.Column("email", sa.String(), nullable=False),
+        sa.Column("tenant_id", sa.String(), nullable=False),
+        sa.UniqueConstraint("email", "tenant_id", name="uq_user_tenant"),
+        sa.UniqueConstraint("email", name="uq_email"),
+        schema="public",
+    )
+
+
+def downgrade() -> None:
+    op.drop_table("user_tenant_mapping", schema="public")

backend/danswer/__init__.py

@@ -1,3 +1,3 @@
 import os
 
-__version__ = os.environ.get("DANSWER_VERSION", "") or "0.3-dev"
+__version__ = os.environ.get("DANSWER_VERSION", "") or "Development"

backend/danswer/access/access.py

@@ -1,7 +1,7 @@
 from sqlalchemy.orm import Session
 
 from danswer.access.models import DocumentAccess
-from danswer.access.utils import prefix_user
+from danswer.access.utils import prefix_user_email
 from danswer.configs.constants import PUBLIC_DOC_PAT
 from danswer.db.document import get_access_info_for_document
 from danswer.db.document import get_access_info_for_documents
@@ -18,10 +18,13 @@ def _get_access_for_document(
         document_id=document_id,
     )
 
-    if not info:
-        return DocumentAccess.build(user_ids=[], user_groups=[], is_public=False)
-
-    return DocumentAccess.build(user_ids=info[1], user_groups=[], is_public=info[2])
+    return DocumentAccess.build(
+        user_emails=info[1] if info and info[1] else [],
+        user_groups=[],
+        external_user_emails=[],
+        external_user_group_ids=[],
+        is_public=info[2] if info else False,
+    )
 
 
 def get_access_for_document(
@@ -34,6 +37,16 @@ def get_access_for_document(
     return versioned_get_access_for_document_fn(document_id, db_session)  # type: ignore
 
 
+def get_null_document_access() -> DocumentAccess:
+    return DocumentAccess(
+        user_emails=set(),
+        user_groups=set(),
+        is_public=False,
+        external_user_emails=set(),
+        external_user_group_ids=set(),
+    )
+
+
 def _get_access_for_documents(
     document_ids: list[str],
     db_session: Session,
@@ -42,13 +55,27 @@ def _get_access_for_documents(
         db_session=db_session,
         document_ids=document_ids,
     )
-    return {
-        document_id: DocumentAccess.build(
-            user_ids=user_ids, user_groups=[], is_public=is_public
+    doc_access = {
+        document_id: DocumentAccess(
+            user_emails=set([email for email in user_emails if email]),
+            # MIT version will wipe all groups and external groups on update
+            user_groups=set(),
+            is_public=is_public,
+            external_user_emails=set(),
+            external_user_group_ids=set(),
         )
-        for document_id, user_ids, is_public in document_access_info
+        for document_id, user_emails, is_public in document_access_info
     }
 
+    # Sometimes the document has not be indexed by the indexing job yet, in those cases
+    # the document does not exist and so we use least permissive. Specifically the EE version
+    # checks the MIT version permissions and creates a superset. This ensures that this flow
+    # does not fail even if the Document has not yet been indexed.
+    for doc_id in document_ids:
+        if doc_id not in doc_access:
+            doc_access[doc_id] = get_null_document_access()
+    return doc_access
+
 
 def get_access_for_documents(
     document_ids: list[str],
@@ -70,7 +97,7 @@ def _get_acl_for_user(user: User | None, db_session: Session) -> set[str]:
     matches one entry in the returned set.
     """
     if user:
-        return {prefix_user(str(user.id)), PUBLIC_DOC_PAT}
+        return {prefix_user_email(user.email), PUBLIC_DOC_PAT}
     return {PUBLIC_DOC_PAT}
 
 

backend/danswer/access/models.py

@@ -1,30 +1,121 @@
 from dataclasses import dataclass
-from uuid import UUID
 
-from danswer.access.utils import prefix_user
+from danswer.access.utils import prefix_external_group
+from danswer.access.utils import prefix_user_email
 from danswer.access.utils import prefix_user_group
 from danswer.configs.constants import PUBLIC_DOC_PAT
 
 
 @dataclass(frozen=True)
-class DocumentAccess:
-    user_ids: set[str]  # stringified UUIDs
-    user_groups: set[str]  # names of user groups associated with this document
+class ExternalAccess:
+    # Emails of external users with access to the doc externally
+    external_user_emails: set[str]
+    # Names or external IDs of groups with access to the doc
+    external_user_group_ids: set[str]
+    # Whether the document is public in the external system or Danswer
     is_public: bool
 
-    def to_acl(self) -> list[str]:
-        return (
-            [prefix_user(user_id) for user_id in self.user_ids]
+
+@dataclass(frozen=True)
+class DocExternalAccess:
+    """
+    This is just a class to wrap the external access and the document ID
+    together. It's used for syncing document permissions to Redis.
+    """
+
+    external_access: ExternalAccess
+    # The document ID
+    doc_id: str
+
+    def to_dict(self) -> dict:
+        return {
+            "external_access": {
+                "external_user_emails": list(self.external_access.external_user_emails),
+                "external_user_group_ids": list(
+                    self.external_access.external_user_group_ids
+                ),
+                "is_public": self.external_access.is_public,
+            },
+            "doc_id": self.doc_id,
+        }
+
+    @classmethod
+    def from_dict(cls, data: dict) -> "DocExternalAccess":
+        external_access = ExternalAccess(
+            external_user_emails=set(
+                data["external_access"].get("external_user_emails", [])
+            ),
+            external_user_group_ids=set(
+                data["external_access"].get("external_user_group_ids", [])
+            ),
+            is_public=data["external_access"]["is_public"],
+        )
+        return cls(
+            external_access=external_access,
+            doc_id=data["doc_id"],
+        )
+
+
+@dataclass(frozen=True)
+class DocumentAccess(ExternalAccess):
+    # User emails for Danswer users, None indicates admin
+    user_emails: set[str | None]
+    # Names of user groups associated with this document
+    user_groups: set[str]
+
+    def to_acl(self) -> set[str]:
+        return set(
+            [
+                prefix_user_email(user_email)
+                for user_email in self.user_emails
+                if user_email
+            ]
             + [prefix_user_group(group_name) for group_name in self.user_groups]
+            + [
+                prefix_user_email(user_email)
+                for user_email in self.external_user_emails
+            ]
+            + [
+                # The group names are already prefixed by the source type
+                # This adds an additional prefix of "external_group:"
+                prefix_external_group(group_name)
+                for group_name in self.external_user_group_ids
+            ]
             + ([PUBLIC_DOC_PAT] if self.is_public else [])
         )
 
     @classmethod
     def build(
-        cls, user_ids: list[UUID | None], user_groups: list[str], is_public: bool
+        cls,
+        user_emails: list[str | None],
+        user_groups: list[str],
+        external_user_emails: list[str],
+        external_user_group_ids: list[str],
+        is_public: bool,
     ) -> "DocumentAccess":
         return cls(
-            user_ids={str(user_id) for user_id in user_ids if user_id},
+            external_user_emails={
+                prefix_user_email(external_email)
+                for external_email in external_user_emails
+            },
+            external_user_group_ids={
+                prefix_external_group(external_group_id)
+                for external_group_id in external_user_group_ids
+            },
+            user_emails={
+                prefix_user_email(user_email)
+                for user_email in user_emails
+                if user_email
+            },
             user_groups=set(user_groups),
             is_public=is_public,
         )
+
+
+default_public_access = DocumentAccess(
+    external_user_emails=set(),
+    external_user_group_ids=set(),
+    user_emails=set(),
+    user_groups=set(),
+    is_public=True,
+)

backend/danswer/access/utils.py

@@ -1,10 +1,24 @@
-def prefix_user(user_id: str) -> str:
-    """Prefixes a user ID to eliminate collision with group names.
-    This assumes that groups are prefixed with a different prefix."""
-    return f"user_id:{user_id}"
+from danswer.configs.constants import DocumentSource
+
+
+def prefix_user_email(user_email: str) -> str:
+    """Prefixes a user email to eliminate collision with group names.
+    This applies to both a Danswer user and an External user, this is to make the query time
+    more efficient"""
+    return f"user_email:{user_email}"
 
 
 def prefix_user_group(user_group_name: str) -> str:
-    """Prefixes a user group name to eliminate collision with user IDs.
+    """Prefixes a user group name to eliminate collision with user emails.
     This assumes that user ids are prefixed with a different prefix."""
     return f"group:{user_group_name}"
+
+
+def prefix_external_group(ext_group_name: str) -> str:
+    """Prefixes an external group name to eliminate collision with user emails / Danswer groups."""
+    return f"external_group:{ext_group_name}"
+
+
+def prefix_group_w_source(ext_group_name: str, source: DocumentSource) -> str:
+    """External groups may collide across sources, every source needs its own prefix."""
+    return f"{source.value.upper()}_{ext_group_name}"

backend/danswer/auth/api_key.py

@@ -0,0 +1,102 @@
+import hashlib
+import secrets
+import uuid
+from urllib.parse import quote
+from urllib.parse import unquote
+
+from fastapi import Request
+from passlib.hash import sha256_crypt
+from pydantic import BaseModel
+
+from danswer.auth.schemas import UserRole
+from danswer.configs.app_configs import API_KEY_HASH_ROUNDS
+
+
+_API_KEY_HEADER_NAME = "Authorization"
+# NOTE for others who are curious: In the context of a header, "X-" often refers
+# to non-standard, experimental, or custom headers in HTTP or other protocols. It
+# indicates that the header is not part of the official standards defined by
+# organizations like the Internet Engineering Task Force (IETF).
+_API_KEY_HEADER_ALTERNATIVE_NAME = "X-Danswer-Authorization"
+_BEARER_PREFIX = "Bearer "
+_API_KEY_PREFIX = "on_"
+_DEPRECATED_API_KEY_PREFIX = "dn_"
+_API_KEY_LEN = 192
+
+
+class ApiKeyDescriptor(BaseModel):
+    api_key_id: int
+    api_key_display: str
+    api_key: str | None = None  # only present on initial creation
+    api_key_name: str | None = None
+    api_key_role: UserRole
+
+    user_id: uuid.UUID
+
+
+def generate_api_key(tenant_id: str | None = None) -> str:
+    # For backwards compatibility, if no tenant_id, generate old style key
+    if not tenant_id:
+        return _API_KEY_PREFIX + secrets.token_urlsafe(_API_KEY_LEN)
+
+    encoded_tenant = quote(tenant_id)  # URL encode the tenant ID
+    return f"{_API_KEY_PREFIX}{encoded_tenant}.{secrets.token_urlsafe(_API_KEY_LEN)}"
+
+
+def extract_tenant_from_api_key_header(request: Request) -> str | None:
+    """Extract tenant ID from request. Returns None if auth is disabled or invalid format."""
+    raw_api_key_header = request.headers.get(
+        _API_KEY_HEADER_ALTERNATIVE_NAME
+    ) or request.headers.get(_API_KEY_HEADER_NAME)
+
+    if not raw_api_key_header or not raw_api_key_header.startswith(_BEARER_PREFIX):
+        return None
+
+    api_key = raw_api_key_header[len(_BEARER_PREFIX) :].strip()
+
+    if not api_key.startswith(_API_KEY_PREFIX) and not api_key.startswith(
+        _DEPRECATED_API_KEY_PREFIX
+    ):
+        return None
+
+    parts = api_key[len(_API_KEY_PREFIX) :].split(".", 1)
+    if len(parts) != 2:
+        return None
+
+    tenant_id = parts[0]
+    return unquote(tenant_id) if tenant_id else None
+
+
+def _deprecated_hash_api_key(api_key: str) -> str:
+    return sha256_crypt.hash(api_key, salt="", rounds=API_KEY_HASH_ROUNDS)
+
+
+def hash_api_key(api_key: str) -> str:
+    # NOTE: no salt is needed, as the API key is randomly generated
+    # and overlaps are impossible
+    if api_key.startswith(_API_KEY_PREFIX):
+        return hashlib.sha256(api_key.encode("utf-8")).hexdigest()
+    elif api_key.startswith(_DEPRECATED_API_KEY_PREFIX):
+        return _deprecated_hash_api_key(api_key)
+    else:
+        raise ValueError(f"Invalid API key prefix: {api_key[:3]}")
+
+
+def build_displayable_api_key(api_key: str) -> str:
+    if api_key.startswith(_API_KEY_PREFIX):
+        api_key = api_key[len(_API_KEY_PREFIX) :]
+
+    return _API_KEY_PREFIX + api_key[:4] + "********" + api_key[-4:]
+
+
+def get_hashed_api_key_from_request(request: Request) -> str | None:
+    raw_api_key_header = request.headers.get(
+        _API_KEY_HEADER_ALTERNATIVE_NAME
+    ) or request.headers.get(_API_KEY_HEADER_NAME)
+    if raw_api_key_header is None:
+        return None
+
+    if raw_api_key_header.startswith(_BEARER_PREFIX):
+        raw_api_key_header = raw_api_key_header[len(_BEARER_PREFIX) :].strip()
+
+    return hash_api_key(raw_api_key_header)

backend/danswer/auth/invited_users.py

@@ -1,20 +1,20 @@
 from typing import cast
 
 from danswer.configs.constants import KV_USER_STORE_KEY
-from danswer.dynamic_configs.factory import get_dynamic_config_store
-from danswer.dynamic_configs.interface import ConfigNotFoundError
-from danswer.dynamic_configs.interface import JSON_ro
+from danswer.key_value_store.factory import get_kv_store
+from danswer.key_value_store.interface import KvKeyNotFoundError
+from danswer.utils.special_types import JSON_ro
 
 
 def get_invited_users() -> list[str]:
     try:
-        store = get_dynamic_config_store()
+        store = get_kv_store()
         return cast(list, store.load(KV_USER_STORE_KEY))
-    except ConfigNotFoundError:
+    except KvKeyNotFoundError:
         return list()
 
 
 def write_invited_users(emails: list[str]) -> int:
-    store = get_dynamic_config_store()
+    store = get_kv_store()
     store.store(KV_USER_STORE_KEY, cast(JSON_ro, emails))
     return len(emails)

backend/danswer/auth/noauth_user.py

@@ -4,29 +4,31 @@ from typing import cast
 
 from danswer.auth.schemas import UserRole
 from danswer.configs.constants import KV_NO_AUTH_USER_PREFERENCES_KEY
-from danswer.dynamic_configs.store import ConfigNotFoundError
-from danswer.dynamic_configs.store import DynamicConfigStore
+from danswer.key_value_store.store import KeyValueStore
+from danswer.key_value_store.store import KvKeyNotFoundError
 from danswer.server.manage.models import UserInfo
 from danswer.server.manage.models import UserPreferences
 
 
 def set_no_auth_user_preferences(
-    store: DynamicConfigStore, preferences: UserPreferences
+    store: KeyValueStore, preferences: UserPreferences
 ) -> None:
     store.store(KV_NO_AUTH_USER_PREFERENCES_KEY, preferences.model_dump())
 
 
-def load_no_auth_user_preferences(store: DynamicConfigStore) -> UserPreferences:
+def load_no_auth_user_preferences(store: KeyValueStore) -> UserPreferences:
     try:
         preferences_data = cast(
             Mapping[str, Any], store.load(KV_NO_AUTH_USER_PREFERENCES_KEY)
         )
         return UserPreferences(**preferences_data)
-    except ConfigNotFoundError:
-        return UserPreferences(chosen_assistants=None, default_model=None)
+    except KvKeyNotFoundError:
+        return UserPreferences(
+            chosen_assistants=None, default_model=None, auto_scroll=True
+        )
 
 
-def fetch_no_auth_user(store: DynamicConfigStore) -> UserInfo:
+def fetch_no_auth_user(store: KeyValueStore) -> UserInfo:
     return UserInfo(
         id="__no_auth_user__",
         email="anonymous@danswer.ai",

backend/danswer/auth/schemas.py

@@ -13,12 +13,24 @@ class UserRole(str, Enum):
         groups they are curators of
     - Global Curator can perform admin actions
         for all groups they are a member of
+    - Limited can access a limited set of basic api endpoints
+    - Slack are users that have used danswer via slack but dont have a web login
+    - External permissioned users that have been picked up during the external permissions sync process but don't have a web login
     """
 
+    LIMITED = "limited"
     BASIC = "basic"
     ADMIN = "admin"
     CURATOR = "curator"
     GLOBAL_CURATOR = "global_curator"
+    SLACK_USER = "slack_user"
+    EXT_PERM_USER = "ext_perm_user"
+
+    def is_web_login(self) -> bool:
+        return self not in [
+            UserRole.SLACK_USER,
+            UserRole.EXT_PERM_USER,
+        ]
 
 
 class UserStatus(str, Enum):
@@ -33,9 +45,8 @@ class UserRead(schemas.BaseUser[uuid.UUID]):
 
 class UserCreate(schemas.BaseUserCreate):
     role: UserRole = UserRole.BASIC
-    has_web_login: bool | None = True
+    tenant_id: str | None = None
 
 
 class UserUpdate(schemas.BaseUserUpdate):
     role: UserRole
-    has_web_login: bool | None = True

backend/danswer/auth/users.py

@@ -5,17 +5,23 @@ from datetime import datetime
 from datetime import timezone
 from email.mime.multipart import MIMEMultipart
 from email.mime.text import MIMEText
+from typing import Dict
+from typing import List
 from typing import Optional
 from typing import Tuple
 
+import jwt
 from email_validator import EmailNotValidError
+from email_validator import EmailUndeliverableError
 from email_validator import validate_email
 from fastapi import APIRouter
 from fastapi import Depends
 from fastapi import HTTPException
+from fastapi import Query
 from fastapi import Request
 from fastapi import Response
 from fastapi import status
+from fastapi.responses import RedirectResponse
 from fastapi.security import OAuth2PasswordRequestForm
 from fastapi_users import BaseUserManager
 from fastapi_users import exceptions
@@ -25,13 +31,27 @@ from fastapi_users import schemas
 from fastapi_users import UUIDIDMixin
 from fastapi_users.authentication import AuthenticationBackend
 from fastapi_users.authentication import CookieTransport
+from fastapi_users.authentication import JWTStrategy
 from fastapi_users.authentication import Strategy
 from fastapi_users.authentication.strategy.db import AccessTokenDatabase
 from fastapi_users.authentication.strategy.db import DatabaseStrategy
+from fastapi_users.exceptions import UserAlreadyExists
+from fastapi_users.jwt import decode_jwt
+from fastapi_users.jwt import generate_jwt
+from fastapi_users.jwt import SecretType
+from fastapi_users.manager import UserManagerDependency
 from fastapi_users.openapi import OpenAPIResponseType
+from fastapi_users.router.common import ErrorCode
+from fastapi_users.router.common import ErrorModel
 from fastapi_users_db_sqlalchemy import SQLAlchemyUserDatabase
-from sqlalchemy.orm import Session
+from httpx_oauth.integrations.fastapi import OAuth2AuthorizeCallback
+from httpx_oauth.oauth2 import BaseOAuth2
+from httpx_oauth.oauth2 import OAuth2Token
+from pydantic import BaseModel
+from sqlalchemy import text
+from sqlalchemy.ext.asyncio import AsyncSession
 
+from danswer.auth.api_key import get_hashed_api_key_from_request
 from danswer.auth.invited_users import get_invited_users
 from danswer.auth.schemas import UserCreate
 from danswer.auth.schemas import UserRole
@@ -53,19 +73,28 @@ from danswer.configs.constants import AuthType
 from danswer.configs.constants import DANSWER_API_KEY_DUMMY_EMAIL_DOMAIN
 from danswer.configs.constants import DANSWER_API_KEY_PREFIX
 from danswer.configs.constants import UNNAMED_KEY_PLACEHOLDER
+from danswer.db.api_key import fetch_user_for_api_key
 from danswer.db.auth import get_access_token_db
 from danswer.db.auth import get_default_admin_user_emails
 from danswer.db.auth import get_user_count
 from danswer.db.auth import get_user_db
-from danswer.db.engine import get_session
-from danswer.db.engine import get_sqlalchemy_engine
+from danswer.db.auth import SQLAlchemyUserAdminDB
+from danswer.db.engine import get_async_session
+from danswer.db.engine import get_async_session_with_tenant
+from danswer.db.engine import get_session_with_tenant
 from danswer.db.models import AccessToken
+from danswer.db.models import OAuthAccount
 from danswer.db.models import User
 from danswer.db.users import get_user_by_email
+from danswer.server.utils import BasicAuthenticationError
 from danswer.utils.logger import setup_logger
 from danswer.utils.telemetry import optional_telemetry
 from danswer.utils.telemetry import RecordType
+from danswer.utils.variable_functionality import fetch_ee_implementation_or_noop
 from danswer.utils.variable_functionality import fetch_versioned_implementation
+from shared_configs.configs import async_return_default_schema
+from shared_configs.configs import MULTI_TENANT
+from shared_configs.contextvars import CURRENT_TENANT_ID_CONTEXTVAR
 
 logger = setup_logger()
 
@@ -102,9 +131,12 @@ def get_display_email(email: str | None, space_less: bool = False) -> str:
 
 
 def user_needs_to_be_verified() -> bool:
-    # all other auth types besides basic should require users to be
-    # verified
-    return AUTH_TYPE != AuthType.BASIC or REQUIRE_EMAIL_VERIFICATION
+    if AUTH_TYPE == AuthType.BASIC or AUTH_TYPE == AuthType.CLOUD:
+        return REQUIRE_EMAIL_VERIFICATION
+
+    # For other auth types, if the user is authenticated it's assumed that
+    # the user is already verified via the external IDP
+    return False
 
 
 def verify_email_is_invited(email: str) -> None:
@@ -115,7 +147,10 @@ def verify_email_is_invited(email: str) -> None:
     if not email:
         raise PermissionError("Email must be specified")
 
-    email_info = validate_email(email)  # can raise EmailNotValidError
+    try:
+        email_info = validate_email(email)
+    except EmailUndeliverableError:
+        raise PermissionError("Email is not valid")
 
     for email_whitelist in whitelist:
         try:
@@ -133,8 +168,8 @@ def verify_email_is_invited(email: str) -> None:
     raise PermissionError("User not on allowed user whitelist")
 
 
-def verify_email_in_whitelist(email: str) -> None:
-    with Session(get_sqlalchemy_engine()) as db_session:
+def verify_email_in_whitelist(email: str, tenant_id: str | None = None) -> None:
+    with get_session_with_tenant(tenant_id) as db_session:
         if not get_user_by_email(email, db_session):
             verify_email_is_invited(email)
 
@@ -182,44 +217,71 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
     reset_password_token_secret = USER_AUTH_SECRET
     verification_token_secret = USER_AUTH_SECRET
 
+    user_db: SQLAlchemyUserDatabase[User, uuid.UUID]
+
     async def create(
         self,
         user_create: schemas.UC | UserCreate,
         safe: bool = False,
         request: Optional[Request] = None,
     ) -> User:
-        verify_email_is_invited(user_create.email)
-        verify_email_domain(user_create.email)
-        if hasattr(user_create, "role"):
-            user_count = await get_user_count()
-            if user_count == 0 or user_create.email in get_default_admin_user_emails():
-                user_create.role = UserRole.ADMIN
-            else:
-                user_create.role = UserRole.BASIC
-        user = None
-        try:
-            user = await super().create(user_create, safe=safe, request=request)  # type: ignore
-        except exceptions.UserAlreadyExists:
-            user = await self.get_by_email(user_create.email)
-            # Handle case where user has used product outside of web and is now creating an account through web
-            if (
-                not user.has_web_login
-                and hasattr(user_create, "has_web_login")
-                and user_create.has_web_login
-            ):
-                user_update = UserUpdate(
-                    password=user_create.password,
-                    has_web_login=True,
-                    role=user_create.role,
-                    is_verified=user_create.is_verified,
+        referral_source = None
+        if request is not None:
+            referral_source = request.cookies.get("referral_source", None)
+
+        tenant_id = await fetch_ee_implementation_or_noop(
+            "danswer.server.tenants.provisioning",
+            "get_or_create_tenant_id",
+            async_return_default_schema,
+        )(
+            email=user_create.email,
+            referral_source=referral_source,
+        )
+
+        async with get_async_session_with_tenant(tenant_id) as db_session:
+            token = CURRENT_TENANT_ID_CONTEXTVAR.set(tenant_id)
+
+            verify_email_is_invited(user_create.email)
+            verify_email_domain(user_create.email)
+            if MULTI_TENANT:
+                tenant_user_db = SQLAlchemyUserAdminDB[User, uuid.UUID](
+                    db_session, User, OAuthAccount
                 )
-                user = await self.update(user_update, user)
-            else:
-                raise exceptions.UserAlreadyExists()
-        return user
+                self.user_db = tenant_user_db
+                self.database = tenant_user_db
+
+            if hasattr(user_create, "role"):
+                user_count = await get_user_count()
+                if (
+                    user_count == 0
+                    or user_create.email in get_default_admin_user_emails()
+                ):
+                    user_create.role = UserRole.ADMIN
+                else:
+                    user_create.role = UserRole.BASIC
+
+            try:
+                user = await super().create(user_create, safe=safe, request=request)  # type: ignore
+            except exceptions.UserAlreadyExists:
+                user = await self.get_by_email(user_create.email)
+                # Handle case where user has used product outside of web and is now creating an account through web
+                if not user.role.is_web_login() and user_create.role.is_web_login():
+                    user_update = UserUpdate(
+                        password=user_create.password,
+                        role=user_create.role,
+                        is_verified=user_create.is_verified,
+                    )
+                    user = await self.update(user_update, user)
+                else:
+                    raise exceptions.UserAlreadyExists()
+
+            finally:
+                CURRENT_TENANT_ID_CONTEXTVAR.reset(token)
+
+            return user
 
     async def oauth_callback(
-        self: "BaseUserManager[models.UOAP, models.ID]",
+        self,
         oauth_name: str,
         access_token: str,
         account_id: str,
@@ -230,45 +292,128 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
         *,
         associate_by_email: bool = False,
         is_verified_by_default: bool = False,
-    ) -> models.UOAP:
-        verify_email_in_whitelist(account_email)
-        verify_email_domain(account_email)
+    ) -> User:
+        referral_source = None
+        if request:
+            referral_source = getattr(request.state, "referral_source", None)
 
-        user = await super().oauth_callback(  # type: ignore
-            oauth_name=oauth_name,
-            access_token=access_token,
-            account_id=account_id,
-            account_email=account_email,
-            expires_at=expires_at,
-            refresh_token=refresh_token,
-            request=request,
-            associate_by_email=associate_by_email,
-            is_verified_by_default=is_verified_by_default,
+        tenant_id = await fetch_ee_implementation_or_noop(
+            "danswer.server.tenants.provisioning",
+            "get_or_create_tenant_id",
+            async_return_default_schema,
+        )(
+            email=account_email,
+            referral_source=referral_source,
         )
 
-        # NOTE: Most IdPs have very short expiry times, and we don't want to force the user to
-        # re-authenticate that frequently, so by default this is disabled
-        if expires_at and TRACK_EXTERNAL_IDP_EXPIRY:
-            oidc_expiry = datetime.fromtimestamp(expires_at, tz=timezone.utc)
-            await self.user_db.update(user, update_dict={"oidc_expiry": oidc_expiry})
+        if not tenant_id:
+            raise HTTPException(status_code=401, detail="User not found")
 
-        # this is needed if an organization goes from `TRACK_EXTERNAL_IDP_EXPIRY=true` to `false`
-        # otherwise, the oidc expiry will always be old, and the user will never be able to login
-        if user.oidc_expiry and not TRACK_EXTERNAL_IDP_EXPIRY:
-            await self.user_db.update(user, update_dict={"oidc_expiry": None})
+        # Proceed with the tenant context
+        token = None
+        async with get_async_session_with_tenant(tenant_id) as db_session:
+            token = CURRENT_TENANT_ID_CONTEXTVAR.set(tenant_id)
 
-        # Handle case where user has used product outside of web and is now creating an account through web
-        if not user.has_web_login:
-            await self.user_db.update(
-                user,
-                update_dict={
-                    "is_verified": is_verified_by_default,
-                    "has_web_login": True,
-                },
-            )
-            user.is_verified = is_verified_by_default
-            user.has_web_login = True
-        return user
+            verify_email_in_whitelist(account_email, tenant_id)
+            verify_email_domain(account_email)
+
+            if MULTI_TENANT:
+                tenant_user_db = SQLAlchemyUserAdminDB[User, uuid.UUID](
+                    db_session, User, OAuthAccount
+                )
+                self.user_db = tenant_user_db
+                self.database = tenant_user_db
+
+            oauth_account_dict = {
+                "oauth_name": oauth_name,
+                "access_token": access_token,
+                "account_id": account_id,
+                "account_email": account_email,
+                "expires_at": expires_at,
+                "refresh_token": refresh_token,
+            }
+
+            try:
+                # Attempt to get user by OAuth account
+                user = await self.get_by_oauth_account(oauth_name, account_id)
+
+            except exceptions.UserNotExists:
+                try:
+                    # Attempt to get user by email
+                    user = await self.get_by_email(account_email)
+                    if not associate_by_email:
+                        raise exceptions.UserAlreadyExists()
+
+                    user = await self.user_db.add_oauth_account(
+                        user, oauth_account_dict
+                    )
+
+                    # If user not found by OAuth account or email, create a new user
+                except exceptions.UserNotExists:
+                    password = self.password_helper.generate()
+                    user_dict = {
+                        "email": account_email,
+                        "hashed_password": self.password_helper.hash(password),
+                        "is_verified": is_verified_by_default,
+                    }
+
+                    user = await self.user_db.create(user_dict)
+
+                    # Explicitly set the Postgres schema for this session to ensure
+                    # OAuth account creation happens in the correct tenant schema
+                    await db_session.execute(text(f'SET search_path = "{tenant_id}"'))
+
+                    # Add OAuth account
+                    await self.user_db.add_oauth_account(user, oauth_account_dict)
+                    await self.on_after_register(user, request)
+
+            else:
+                for existing_oauth_account in user.oauth_accounts:
+                    if (
+                        existing_oauth_account.account_id == account_id
+                        and existing_oauth_account.oauth_name == oauth_name
+                    ):
+                        user = await self.user_db.update_oauth_account(
+                            user,
+                            # NOTE: OAuthAccount DOES implement the OAuthAccountProtocol
+                            # but the type checker doesn't know that :(
+                            existing_oauth_account,  # type: ignore
+                            oauth_account_dict,
+                        )
+
+            # NOTE: Most IdPs have very short expiry times, and we don't want to force the user to
+            # re-authenticate that frequently, so by default this is disabled
+
+            if expires_at and TRACK_EXTERNAL_IDP_EXPIRY:
+                oidc_expiry = datetime.fromtimestamp(expires_at, tz=timezone.utc)
+                await self.user_db.update(
+                    user, update_dict={"oidc_expiry": oidc_expiry}
+                )
+
+            # Handle case where user has used product outside of web and is now creating an account through web
+            if not user.role.is_web_login():
+                await self.user_db.update(
+                    user,
+                    {
+                        "is_verified": is_verified_by_default,
+                        "role": UserRole.BASIC,
+                    },
+                )
+                user.is_verified = is_verified_by_default
+
+            # this is needed if an organization goes from `TRACK_EXTERNAL_IDP_EXPIRY=true` to `false`
+            # otherwise, the oidc expiry will always be old, and the user will never be able to login
+            if (
+                user.oidc_expiry is not None  # type: ignore
+                and not TRACK_EXTERNAL_IDP_EXPIRY
+            ):
+                await self.user_db.update(user, {"oidc_expiry": None})
+                user.oidc_expiry = None  # type: ignore
+
+            if token:
+                CURRENT_TENANT_ID_CONTEXTVAR.reset(token)
+
+            return user
 
     async def on_after_register(
         self, user: User, request: Optional[Request] = None
@@ -299,18 +444,53 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
     async def authenticate(
         self, credentials: OAuth2PasswordRequestForm
     ) -> Optional[User]:
-        user = await super().authenticate(credentials)
-        if user is None:
+        email = credentials.username
+
+        # Get tenant_id from mapping table
+        tenant_id = await fetch_ee_implementation_or_noop(
+            "danswer.server.tenants.provisioning",
+            "get_or_create_tenant_id",
+            async_return_default_schema,
+        )(
+            email=email,
+        )
+        if not tenant_id:
+            # User not found in mapping
+            self.password_helper.hash(credentials.password)
+            return None
+
+        # Create a tenant-specific session
+        async with get_async_session_with_tenant(tenant_id) as tenant_session:
+            tenant_user_db: SQLAlchemyUserDatabase = SQLAlchemyUserDatabase(
+                tenant_session, User
+            )
+            self.user_db = tenant_user_db
+
+            # Proceed with authentication
             try:
-                user = await self.get_by_email(credentials.username)
-                if not user.has_web_login:
-                    raise HTTPException(
-                        status_code=status.HTTP_403_FORBIDDEN,
-                        detail="NO_WEB_LOGIN_AND_HAS_NO_PASSWORD",
-                    )
+                user = await self.get_by_email(email)
+
             except exceptions.UserNotExists:
-                pass
-        return user
+                self.password_helper.hash(credentials.password)
+                return None
+
+            if not user.role.is_web_login():
+                raise BasicAuthenticationError(
+                    detail="NO_WEB_LOGIN_AND_HAS_NO_PASSWORD",
+                )
+
+            verified, updated_password_hash = self.password_helper.verify_and_update(
+                credentials.password, user.hashed_password
+            )
+            if not verified:
+                return None
+
+            if updated_password_hash is not None:
+                await self.user_db.update(
+                    user, {"hashed_password": updated_password_hash}
+                )
+
+            return user
 
 
 async def get_user_manager(
@@ -325,21 +505,51 @@ cookie_transport = CookieTransport(
 )
 
 
+# This strategy is used to add tenant_id to the JWT token
+class TenantAwareJWTStrategy(JWTStrategy):
+    async def _create_token_data(self, user: User, impersonate: bool = False) -> dict:
+        tenant_id = await fetch_ee_implementation_or_noop(
+            "danswer.server.tenants.provisioning",
+            "get_or_create_tenant_id",
+            async_return_default_schema,
+        )(
+            email=user.email,
+        )
+
+        data = {
+            "sub": str(user.id),
+            "aud": self.token_audience,
+            "tenant_id": tenant_id,
+        }
+        return data
+
+    async def write_token(self, user: User) -> str:
+        data = await self._create_token_data(user)
+        return generate_jwt(
+            data, self.encode_key, self.lifetime_seconds, algorithm=self.algorithm
+        )
+
+
+def get_jwt_strategy() -> TenantAwareJWTStrategy:
+    return TenantAwareJWTStrategy(
+        secret=USER_AUTH_SECRET,
+        lifetime_seconds=SESSION_EXPIRE_TIME_SECONDS,
+    )
+
+
 def get_database_strategy(
     access_token_db: AccessTokenDatabase[AccessToken] = Depends(get_access_token_db),
 ) -> DatabaseStrategy:
-    strategy = DatabaseStrategy(
+    return DatabaseStrategy(
         access_token_db, lifetime_seconds=SESSION_EXPIRE_TIME_SECONDS  # type: ignore
     )
 
-    return strategy
-
 
 auth_backend = AuthenticationBackend(
-    name="database",
+    name="jwt" if MULTI_TENANT else "database",
     transport=cookie_transport,
-    get_strategy=get_database_strategy,
-)
+    get_strategy=get_jwt_strategy if MULTI_TENANT else get_database_strategy,  # type: ignore
+)  # type: ignore
 
 
 class FastAPIUserWithLogoutRouter(FastAPIUsers[models.UP, models.ID]):
@@ -353,9 +563,11 @@ class FastAPIUserWithLogoutRouter(FastAPIUsers[models.UP, models.ID]):
         This way the login router does not need to be included
         """
         router = APIRouter()
+
         get_current_user_token = self.authenticator.current_user_token(
             active=True, verified=requires_verification
         )
+
         logout_responses: OpenAPIResponseType = {
             **{
                 status.HTTP_401_UNAUTHORIZED: {
@@ -393,7 +605,7 @@ optional_fastapi_current_user = fastapi_users.current_user(active=True, optional
 async def optional_user_(
     request: Request,
     user: User | None,
-    db_session: Session,
+    async_db_session: AsyncSession,
 ) -> User | None:
     """NOTE: `request` and `db_session` are not used here, but are included
     for the EE version of this function."""
@@ -402,49 +614,79 @@ async def optional_user_(
 
 async def optional_user(
     request: Request,
+    async_db_session: AsyncSession = Depends(get_async_session),
     user: User | None = Depends(optional_fastapi_current_user),
-    db_session: Session = Depends(get_session),
 ) -> User | None:
     versioned_fetch_user = fetch_versioned_implementation(
         "danswer.auth.users", "optional_user_"
     )
-    return await versioned_fetch_user(request, user, db_session)
+    user = await versioned_fetch_user(request, user, async_db_session)
+
+    # check if an API key is present
+    if user is None:
+        hashed_api_key = get_hashed_api_key_from_request(request)
+        if hashed_api_key:
+            user = await fetch_user_for_api_key(hashed_api_key, async_db_session)
+
+    return user
 
 
 async def double_check_user(
     user: User | None,
     optional: bool = DISABLE_AUTH,
+    include_expired: bool = False,
 ) -> User | None:
     if optional:
         return None
 
     if user is None:
-        raise HTTPException(
-            status_code=status.HTTP_403_FORBIDDEN,
+        raise BasicAuthenticationError(
             detail="Access denied. User is not authenticated.",
         )
 
     if user_needs_to_be_verified() and not user.is_verified:
-        raise HTTPException(
-            status_code=status.HTTP_403_FORBIDDEN,
+        raise BasicAuthenticationError(
             detail="Access denied. User is not verified.",
         )
 
-    if user.oidc_expiry and user.oidc_expiry < datetime.now(timezone.utc):
-        raise HTTPException(
-            status_code=status.HTTP_403_FORBIDDEN,
+    if (
+        user.oidc_expiry
+        and user.oidc_expiry < datetime.now(timezone.utc)
+        and not include_expired
+    ):
+        raise BasicAuthenticationError(
             detail="Access denied. User's OIDC token has expired.",
         )
 
     return user
 
 
-async def current_user(
+async def current_user_with_expired_token(
+    user: User | None = Depends(optional_user),
+) -> User | None:
+    return await double_check_user(user, include_expired=True)
+
+
+async def current_limited_user(
     user: User | None = Depends(optional_user),
 ) -> User | None:
     return await double_check_user(user)
 
 
+async def current_user(
+    user: User | None = Depends(optional_user),
+) -> User | None:
+    user = await double_check_user(user)
+    if not user:
+        return None
+
+    if user.role == UserRole.LIMITED:
+        raise BasicAuthenticationError(
+            detail="Access denied. User role is LIMITED. BASIC or higher permissions are required.",
+        )
+    return user
+
+
 async def current_curator_or_admin_user(
     user: User | None = Depends(current_user),
 ) -> User | None:
@@ -452,15 +694,13 @@ async def current_curator_or_admin_user(
         return None
 
     if not user or not hasattr(user, "role"):
-        raise HTTPException(
-            status_code=status.HTTP_403_FORBIDDEN,
+        raise BasicAuthenticationError(
             detail="Access denied. User is not authenticated or lacks role information.",
         )
 
     allowed_roles = {UserRole.GLOBAL_CURATOR, UserRole.CURATOR, UserRole.ADMIN}
     if user.role not in allowed_roles:
-        raise HTTPException(
-            status_code=status.HTTP_403_FORBIDDEN,
+        raise BasicAuthenticationError(
             detail="Access denied. User is not a curator or admin.",
         )
 
@@ -472,8 +712,7 @@ async def current_admin_user(user: User | None = Depends(current_user)) -> User
         return None
 
     if not user or not hasattr(user, "role") or user.role != UserRole.ADMIN:
-        raise HTTPException(
-            status_code=status.HTTP_403_FORBIDDEN,
+        raise BasicAuthenticationError(
             detail="Access denied. User must be an admin to perform this action.",
         )
 
@@ -483,3 +722,212 @@ async def current_admin_user(user: User | None = Depends(current_user)) -> User
 def get_default_admin_user_emails_() -> list[str]:
     # No default seeding available for Danswer MIT
     return []
+
+
+STATE_TOKEN_AUDIENCE = "fastapi-users:oauth-state"
+
+
+class OAuth2AuthorizeResponse(BaseModel):
+    authorization_url: str
+
+
+def generate_state_token(
+    data: Dict[str, str], secret: SecretType, lifetime_seconds: int = 3600
+) -> str:
+    data["aud"] = STATE_TOKEN_AUDIENCE
+
+    return generate_jwt(data, secret, lifetime_seconds)
+
+
+# refer to https://github.com/fastapi-users/fastapi-users/blob/42ddc241b965475390e2bce887b084152ae1a2cd/fastapi_users/fastapi_users.py#L91
+def create_danswer_oauth_router(
+    oauth_client: BaseOAuth2,
+    backend: AuthenticationBackend,
+    state_secret: SecretType,
+    redirect_url: Optional[str] = None,
+    associate_by_email: bool = False,
+    is_verified_by_default: bool = False,
+) -> APIRouter:
+    return get_oauth_router(
+        oauth_client,
+        backend,
+        get_user_manager,
+        state_secret,
+        redirect_url,
+        associate_by_email,
+        is_verified_by_default,
+    )
+
+
+def get_oauth_router(
+    oauth_client: BaseOAuth2,
+    backend: AuthenticationBackend,
+    get_user_manager: UserManagerDependency[models.UP, models.ID],
+    state_secret: SecretType,
+    redirect_url: Optional[str] = None,
+    associate_by_email: bool = False,
+    is_verified_by_default: bool = False,
+) -> APIRouter:
+    """Generate a router with the OAuth routes."""
+    router = APIRouter()
+    callback_route_name = f"oauth:{oauth_client.name}.{backend.name}.callback"
+
+    if redirect_url is not None:
+        oauth2_authorize_callback = OAuth2AuthorizeCallback(
+            oauth_client,
+            redirect_url=redirect_url,
+        )
+    else:
+        oauth2_authorize_callback = OAuth2AuthorizeCallback(
+            oauth_client,
+            route_name=callback_route_name,
+        )
+
+    @router.get(
+        "/authorize",
+        name=f"oauth:{oauth_client.name}.{backend.name}.authorize",
+        response_model=OAuth2AuthorizeResponse,
+    )
+    async def authorize(
+        request: Request,
+        scopes: List[str] = Query(None),
+    ) -> OAuth2AuthorizeResponse:
+        referral_source = request.cookies.get("referral_source", None)
+
+        if redirect_url is not None:
+            authorize_redirect_url = redirect_url
+        else:
+            authorize_redirect_url = str(request.url_for(callback_route_name))
+
+        next_url = request.query_params.get("next", "/")
+
+        state_data: Dict[str, str] = {
+            "next_url": next_url,
+            "referral_source": referral_source or "default_referral",
+        }
+        state = generate_state_token(state_data, state_secret)
+        authorization_url = await oauth_client.get_authorization_url(
+            authorize_redirect_url,
+            state,
+            scopes,
+        )
+
+        return OAuth2AuthorizeResponse(authorization_url=authorization_url)
+
+    @router.get(
+        "/callback",
+        name=callback_route_name,
+        description="The response varies based on the authentication backend used.",
+        responses={
+            status.HTTP_400_BAD_REQUEST: {
+                "model": ErrorModel,
+                "content": {
+                    "application/json": {
+                        "examples": {
+                            "INVALID_STATE_TOKEN": {
+                                "summary": "Invalid state token.",
+                                "value": None,
+                            },
+                            ErrorCode.LOGIN_BAD_CREDENTIALS: {
+                                "summary": "User is inactive.",
+                                "value": {"detail": ErrorCode.LOGIN_BAD_CREDENTIALS},
+                            },
+                        }
+                    }
+                },
+            },
+        },
+    )
+    async def callback(
+        request: Request,
+        access_token_state: Tuple[OAuth2Token, str] = Depends(
+            oauth2_authorize_callback
+        ),
+        user_manager: BaseUserManager[models.UP, models.ID] = Depends(get_user_manager),
+        strategy: Strategy[models.UP, models.ID] = Depends(backend.get_strategy),
+    ) -> RedirectResponse:
+        token, state = access_token_state
+        account_id, account_email = await oauth_client.get_id_email(
+            token["access_token"]
+        )
+
+        if account_email is None:
+            raise HTTPException(
+                status_code=status.HTTP_400_BAD_REQUEST,
+                detail=ErrorCode.OAUTH_NOT_AVAILABLE_EMAIL,
+            )
+
+        try:
+            state_data = decode_jwt(state, state_secret, [STATE_TOKEN_AUDIENCE])
+        except jwt.DecodeError:
+            raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST)
+
+        next_url = state_data.get("next_url", "/")
+        referral_source = state_data.get("referral_source", None)
+
+        request.state.referral_source = referral_source
+
+        # Proceed to authenticate or create the user
+        try:
+            user = await user_manager.oauth_callback(
+                oauth_client.name,
+                token["access_token"],
+                account_id,
+                account_email,
+                token.get("expires_at"),
+                token.get("refresh_token"),
+                request,
+                associate_by_email=associate_by_email,
+                is_verified_by_default=is_verified_by_default,
+            )
+        except UserAlreadyExists:
+            raise HTTPException(
+                status_code=status.HTTP_400_BAD_REQUEST,
+                detail=ErrorCode.OAUTH_USER_ALREADY_EXISTS,
+            )
+
+        if not user.is_active:
+            raise HTTPException(
+                status_code=status.HTTP_400_BAD_REQUEST,
+                detail=ErrorCode.LOGIN_BAD_CREDENTIALS,
+            )
+
+        # Login user
+        response = await backend.login(strategy, user)
+        await user_manager.on_after_login(user, request, response)
+
+        # Prepare redirect response
+        redirect_response = RedirectResponse(next_url, status_code=302)
+
+        # Copy headers and other attributes from 'response' to 'redirect_response'
+        for header_name, header_value in response.headers.items():
+            redirect_response.headers[header_name] = header_value
+
+        if hasattr(response, "body"):
+            redirect_response.body = response.body
+        if hasattr(response, "status_code"):
+            redirect_response.status_code = response.status_code
+        if hasattr(response, "media_type"):
+            redirect_response.media_type = response.media_type
+        return redirect_response
+
+    return router
+
+
+async def api_key_dep(
+    request: Request, async_db_session: AsyncSession = Depends(get_async_session)
+) -> User | None:
+    if AUTH_TYPE == AuthType.DISABLED:
+        return None
+
+    hashed_api_key = get_hashed_api_key_from_request(request)
+    if not hashed_api_key:
+        raise HTTPException(status_code=401, detail="Missing API key")
+
+    if hashed_api_key:
+        user = await fetch_user_for_api_key(hashed_api_key, async_db_session)
+
+    if user is None:
+        raise HTTPException(status_code=401, detail="Invalid API key")
+
+    return user

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

@@ -0,0 +1,403 @@
+import logging
+import multiprocessing
+import time
+from typing import Any
+
+import requests
+import sentry_sdk
+from celery import Task
+from celery.app import trace
+from celery.exceptions import WorkerShutdown
+from celery.states import READY_STATES
+from celery.utils.log import get_task_logger
+from celery.worker import strategy  # type: ignore
+from redis.lock import Lock as RedisLock
+from sentry_sdk.integrations.celery import CeleryIntegration
+from sqlalchemy import text
+from sqlalchemy.orm import Session
+
+from danswer.background.celery.apps.task_formatters import CeleryTaskColoredFormatter
+from danswer.background.celery.apps.task_formatters import CeleryTaskPlainFormatter
+from danswer.background.celery.celery_utils import celery_is_worker_primary
+from danswer.configs.constants import DanswerRedisLocks
+from danswer.db.engine import get_sqlalchemy_engine
+from danswer.document_index.vespa_constants import VESPA_CONFIG_SERVER_URL
+from danswer.redis.redis_connector import RedisConnector
+from danswer.redis.redis_connector_credential_pair import RedisConnectorCredentialPair
+from danswer.redis.redis_connector_delete import RedisConnectorDelete
+from danswer.redis.redis_connector_doc_perm_sync import RedisConnectorPermissionSync
+from danswer.redis.redis_connector_ext_group_sync import RedisConnectorExternalGroupSync
+from danswer.redis.redis_connector_prune import RedisConnectorPrune
+from danswer.redis.redis_document_set import RedisDocumentSet
+from danswer.redis.redis_pool import get_redis_client
+from danswer.redis.redis_usergroup import RedisUserGroup
+from danswer.utils.logger import ColoredFormatter
+from danswer.utils.logger import PlainFormatter
+from danswer.utils.logger import setup_logger
+from shared_configs.configs import SENTRY_DSN
+
+
+logger = setup_logger()
+
+task_logger = get_task_logger(__name__)
+
+if SENTRY_DSN:
+    sentry_sdk.init(
+        dsn=SENTRY_DSN,
+        integrations=[CeleryIntegration()],
+        traces_sample_rate=0.1,
+    )
+    logger.info("Sentry initialized")
+else:
+    logger.debug("Sentry DSN not provided, skipping Sentry initialization")
+
+
+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:
+    pass
+
+
+def on_task_postrun(
+    sender: Any | None = None,
+    task_id: str | None = None,
+    task: Task | None = None,
+    args: tuple | None = None,
+    kwargs: dict[str, Any] | None = None,
+    retval: Any | None = None,
+    state: str | None = None,
+    **kwds: Any,
+) -> None:
+    """We handle this signal in order to remove completed tasks
+    from their respective tasksets. This allows us to track the progress of document set
+    and user group syncs.
+
+    This function runs after any task completes (both success and failure)
+    Note that this signal does not fire on a task that failed to complete and is going
+    to be retried.
+
+    This also does not fire if a worker with acks_late=False crashes (which all of our
+    long running workers are)
+    """
+    if not task:
+        return
+
+    task_logger.debug(f"Task {task.name} (ID: {task_id}) completed with state: {state}")
+
+    if state not in READY_STATES:
+        return
+
+    if not task_id:
+        return
+
+    # Get tenant_id directly from kwargs- each celery task has a tenant_id kwarg
+    if not kwargs:
+        logger.error(f"Task {task.name} (ID: {task_id}) is missing kwargs")
+        tenant_id = None
+    else:
+        tenant_id = kwargs.get("tenant_id")
+
+    task_logger.debug(
+        f"Task {task.name} (ID: {task_id}) completed with state: {state} "
+        f"{f'for tenant_id={tenant_id}' if tenant_id else ''}"
+    )
+
+    r = get_redis_client(tenant_id=tenant_id)
+
+    if task_id.startswith(RedisConnectorCredentialPair.PREFIX):
+        r.srem(RedisConnectorCredentialPair.get_taskset_key(), task_id)
+        return
+
+    if task_id.startswith(RedisDocumentSet.PREFIX):
+        document_set_id = RedisDocumentSet.get_id_from_task_id(task_id)
+        if document_set_id is not None:
+            rds = RedisDocumentSet(tenant_id, int(document_set_id))
+            r.srem(rds.taskset_key, task_id)
+        return
+
+    if task_id.startswith(RedisUserGroup.PREFIX):
+        usergroup_id = RedisUserGroup.get_id_from_task_id(task_id)
+        if usergroup_id is not None:
+            rug = RedisUserGroup(tenant_id, int(usergroup_id))
+            r.srem(rug.taskset_key, task_id)
+        return
+
+    if task_id.startswith(RedisConnectorDelete.PREFIX):
+        cc_pair_id = RedisConnector.get_id_from_task_id(task_id)
+        if cc_pair_id is not None:
+            RedisConnectorDelete.remove_from_taskset(int(cc_pair_id), task_id, r)
+        return
+
+    if task_id.startswith(RedisConnectorPrune.SUBTASK_PREFIX):
+        cc_pair_id = RedisConnector.get_id_from_task_id(task_id)
+        if cc_pair_id is not None:
+            RedisConnectorPrune.remove_from_taskset(int(cc_pair_id), task_id, r)
+        return
+
+    if task_id.startswith(RedisConnectorPermissionSync.SUBTASK_PREFIX):
+        cc_pair_id = RedisConnector.get_id_from_task_id(task_id)
+        if cc_pair_id is not None:
+            RedisConnectorPermissionSync.remove_from_taskset(
+                int(cc_pair_id), task_id, r
+            )
+        return
+
+    if task_id.startswith(RedisConnectorExternalGroupSync.SUBTASK_PREFIX):
+        cc_pair_id = RedisConnector.get_id_from_task_id(task_id)
+        if cc_pair_id is not None:
+            RedisConnectorExternalGroupSync.remove_from_taskset(
+                int(cc_pair_id), task_id, r
+            )
+        return
+
+
+def on_celeryd_init(sender: Any = None, conf: Any = None, **kwargs: Any) -> None:
+    """The first signal sent on celery worker startup"""
+    multiprocessing.set_start_method("spawn")  # fork is unsafe, set to spawn
+
+
+def wait_for_redis(sender: Any, **kwargs: Any) -> None:
+    """Waits for redis to become ready subject to a hardcoded timeout.
+    Will raise WorkerShutdown to kill the celery worker if the timeout is reached."""
+
+    r = get_redis_client(tenant_id=None)
+
+    WAIT_INTERVAL = 5
+    WAIT_LIMIT = 60
+
+    ready = False
+    time_start = time.monotonic()
+    logger.info("Redis: Readiness probe starting.")
+    while True:
+        try:
+            if r.ping():
+                ready = True
+                break
+        except Exception:
+            pass
+
+        time_elapsed = time.monotonic() - time_start
+        if time_elapsed > WAIT_LIMIT:
+            break
+
+        logger.info(
+            f"Redis: Readiness probe ongoing. elapsed={time_elapsed:.1f} timeout={WAIT_LIMIT:.1f}"
+        )
+
+        time.sleep(WAIT_INTERVAL)
+
+    if not ready:
+        msg = (
+            f"Redis: Readiness probe did not succeed within the timeout "
+            f"({WAIT_LIMIT} seconds). Exiting..."
+        )
+        logger.error(msg)
+        raise WorkerShutdown(msg)
+
+    logger.info("Redis: Readiness probe succeeded. Continuing...")
+    return
+
+
+def wait_for_db(sender: Any, **kwargs: Any) -> None:
+    """Waits for the db to become ready subject to a hardcoded timeout.
+    Will raise WorkerShutdown to kill the celery worker if the timeout is reached."""
+
+    WAIT_INTERVAL = 5
+    WAIT_LIMIT = 60
+
+    ready = False
+    time_start = time.monotonic()
+    logger.info("Database: Readiness probe starting.")
+    while True:
+        try:
+            with Session(get_sqlalchemy_engine()) as db_session:
+                result = db_session.execute(text("SELECT NOW()")).scalar()
+                if result:
+                    ready = True
+                    break
+        except Exception:
+            pass
+
+        time_elapsed = time.monotonic() - time_start
+        if time_elapsed > WAIT_LIMIT:
+            break
+
+        logger.info(
+            f"Database: Readiness probe ongoing. elapsed={time_elapsed:.1f} timeout={WAIT_LIMIT:.1f}"
+        )
+
+        time.sleep(WAIT_INTERVAL)
+
+    if not ready:
+        msg = (
+            f"Database: Readiness probe did not succeed within the timeout "
+            f"({WAIT_LIMIT} seconds). Exiting..."
+        )
+        logger.error(msg)
+        raise WorkerShutdown(msg)
+
+    logger.info("Database: Readiness probe succeeded. Continuing...")
+    return
+
+
+def wait_for_vespa(sender: Any, **kwargs: Any) -> None:
+    """Waits for Vespa to become ready subject to a hardcoded timeout.
+    Will raise WorkerShutdown to kill the celery worker if the timeout is reached."""
+
+    WAIT_INTERVAL = 5
+    WAIT_LIMIT = 60
+
+    ready = False
+    time_start = time.monotonic()
+    logger.info("Vespa: Readiness probe starting.")
+    while True:
+        try:
+            response = requests.get(f"{VESPA_CONFIG_SERVER_URL}/state/v1/health")
+            response.raise_for_status()
+
+            response_dict = response.json()
+            if response_dict["status"]["code"] == "up":
+                ready = True
+                break
+        except Exception:
+            pass
+
+        time_elapsed = time.monotonic() - time_start
+        if time_elapsed > WAIT_LIMIT:
+            break
+
+        logger.info(
+            f"Vespa: Readiness probe ongoing. elapsed={time_elapsed:.1f} timeout={WAIT_LIMIT:.1f}"
+        )
+
+        time.sleep(WAIT_INTERVAL)
+
+    if not ready:
+        msg = (
+            f"Vespa: Readiness probe did not succeed within the timeout "
+            f"({WAIT_LIMIT} seconds). Exiting..."
+        )
+        logger.error(msg)
+        raise WorkerShutdown(msg)
+
+    logger.info("Vespa: Readiness probe succeeded. Continuing...")
+    return
+
+
+def on_secondary_worker_init(sender: Any, **kwargs: Any) -> None:
+    logger.info("Running as a secondary celery worker.")
+
+    # Set up variables for waiting on primary worker
+    WAIT_INTERVAL = 5
+    WAIT_LIMIT = 60
+    r = get_redis_client(tenant_id=None)
+    time_start = time.monotonic()
+
+    logger.info("Waiting for primary worker to be ready...")
+    while True:
+        if r.exists(DanswerRedisLocks.PRIMARY_WORKER):
+            break
+
+        time_elapsed = time.monotonic() - time_start
+        logger.info(
+            f"Primary worker is not ready yet. elapsed={time_elapsed:.1f} timeout={WAIT_LIMIT:.1f}"
+        )
+        if time_elapsed > WAIT_LIMIT:
+            msg = (
+                f"Primary worker was not ready within the timeout. "
+                f"({WAIT_LIMIT} seconds). Exiting..."
+            )
+            logger.error(msg)
+            raise WorkerShutdown(msg)
+
+        time.sleep(WAIT_INTERVAL)
+
+    logger.info("Wait for primary worker completed successfully. Continuing...")
+    return
+
+
+def on_worker_ready(sender: Any, **kwargs: Any) -> None:
+    task_logger.info("worker_ready signal received.")
+
+
+def on_worker_shutdown(sender: Any, **kwargs: Any) -> None:
+    if not celery_is_worker_primary(sender):
+        return
+
+    if not sender.primary_worker_lock:
+        return
+
+    logger.info("Releasing primary worker lock.")
+    lock: RedisLock = sender.primary_worker_lock
+    try:
+        if lock.owned():
+            try:
+                lock.release()
+                sender.primary_worker_lock = None
+            except Exception:
+                logger.exception("Failed to release primary worker lock")
+    except Exception:
+        logger.exception("Failed to check if primary worker lock is owned")
+
+
+def on_setup_logging(
+    loglevel: Any, logfile: Any, format: Any, colorize: Any, **kwargs: Any
+) -> None:
+    # TODO: could unhardcode format and colorize and accept these as options from
+    # celery's config
+
+    # reformats the root logger
+    root_logger = logging.getLogger()
+
+    root_handler = logging.StreamHandler()  # Set up a handler for the root logger
+    root_formatter = ColoredFormatter(
+        "%(asctime)s %(filename)30s %(lineno)4s: %(message)s",
+        datefmt="%m/%d/%Y %I:%M:%S %p",
+    )
+    root_handler.setFormatter(root_formatter)
+    root_logger.addHandler(root_handler)  # Apply the handler to the root logger
+
+    if logfile:
+        root_file_handler = logging.FileHandler(logfile)
+        root_file_formatter = PlainFormatter(
+            "%(asctime)s %(filename)30s %(lineno)4s: %(message)s",
+            datefmt="%m/%d/%Y %I:%M:%S %p",
+        )
+        root_file_handler.setFormatter(root_file_formatter)
+        root_logger.addHandler(root_file_handler)
+
+    root_logger.setLevel(loglevel)
+
+    # reformats celery's task logger
+    task_formatter = CeleryTaskColoredFormatter(
+        "%(asctime)s %(filename)30s %(lineno)4s: %(message)s",
+        datefmt="%m/%d/%Y %I:%M:%S %p",
+    )
+    task_handler = logging.StreamHandler()  # Set up a handler for the task logger
+    task_handler.setFormatter(task_formatter)
+    task_logger.addHandler(task_handler)  # Apply the handler to the task logger
+
+    if logfile:
+        task_file_handler = logging.FileHandler(logfile)
+        task_file_formatter = CeleryTaskPlainFormatter(
+            "%(asctime)s %(filename)30s %(lineno)4s: %(message)s",
+            datefmt="%m/%d/%Y %I:%M:%S %p",
+        )
+        task_file_handler.setFormatter(task_file_formatter)
+        task_logger.addHandler(task_file_handler)
+
+    task_logger.setLevel(loglevel)
+    task_logger.propagate = False
+
+    # hide celery task received spam
+    # e.g. "Task check_for_pruning[a1e96171-0ba8-4e00-887b-9fbf7442eab3] received"
+    strategy.logger.setLevel(logging.WARNING)
+
+    # hide celery task succeeded/failed spam
+    # e.g. "Task check_for_pruning[a1e96171-0ba8-4e00-887b-9fbf7442eab3] succeeded in 0.03137450001668185s: None"
+    trace.logger.setLevel(logging.WARNING)

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

@@ -0,0 +1,172 @@
+from datetime import timedelta
+from typing import Any
+
+from celery import Celery
+from celery import signals
+from celery.beat import PersistentScheduler  # type: ignore
+from celery.signals import beat_init
+
+import danswer.background.celery.apps.app_base as app_base
+from danswer.configs.constants import POSTGRES_CELERY_BEAT_APP_NAME
+from danswer.db.engine import get_all_tenant_ids
+from danswer.db.engine import SqlEngine
+from danswer.utils.logger import setup_logger
+from danswer.utils.variable_functionality import fetch_versioned_implementation
+from shared_configs.configs import IGNORED_SYNCING_TENANT_LIST
+from shared_configs.configs import MULTI_TENANT
+
+logger = setup_logger(__name__)
+
+celery_app = Celery(__name__)
+celery_app.config_from_object("danswer.background.celery.configs.beat")
+
+
+class DynamicTenantScheduler(PersistentScheduler):
+    def __init__(self, *args: Any, **kwargs: Any) -> None:
+        logger.info("Initializing DynamicTenantScheduler")
+        super().__init__(*args, **kwargs)
+        self._reload_interval = timedelta(minutes=2)
+        self._last_reload = self.app.now() - self._reload_interval
+        # Let the parent class handle store initialization
+        self.setup_schedule()
+        self._update_tenant_tasks()
+        logger.info(f"Set reload interval to {self._reload_interval}")
+
+    def setup_schedule(self) -> None:
+        logger.info("Setting up initial schedule")
+        super().setup_schedule()
+        logger.info("Initial schedule setup complete")
+
+    def tick(self) -> float:
+        retval = super().tick()
+        now = self.app.now()
+        if (
+            self._last_reload is None
+            or (now - self._last_reload) > self._reload_interval
+        ):
+            logger.info("Reload interval reached, initiating tenant task update")
+            self._update_tenant_tasks()
+            self._last_reload = now
+            logger.info("Tenant task update completed, reset reload timer")
+        return retval
+
+    def _update_tenant_tasks(self) -> None:
+        logger.info("Starting tenant task update process")
+        try:
+            logger.info("Fetching all tenant IDs")
+            tenant_ids = get_all_tenant_ids()
+            logger.info(f"Found {len(tenant_ids)} tenants")
+
+            logger.info("Fetching tasks to schedule")
+            tasks_to_schedule = fetch_versioned_implementation(
+                "danswer.background.celery.tasks.beat_schedule", "get_tasks_to_schedule"
+            )
+
+            new_beat_schedule: dict[str, dict[str, Any]] = {}
+
+            current_schedule = self.schedule.items()
+
+            existing_tenants = set()
+            for task_name, _ in current_schedule:
+                if "-" in task_name:
+                    existing_tenants.add(task_name.split("-")[-1])
+            logger.info(f"Found {len(existing_tenants)} existing tenants in schedule")
+
+            for tenant_id in tenant_ids:
+                if (
+                    IGNORED_SYNCING_TENANT_LIST
+                    and tenant_id in IGNORED_SYNCING_TENANT_LIST
+                ):
+                    logger.info(
+                        f"Skipping tenant {tenant_id} as it is in the ignored syncing list"
+                    )
+                    continue
+
+                if tenant_id not in existing_tenants:
+                    logger.info(f"Processing new tenant: {tenant_id}")
+
+                for task in tasks_to_schedule():
+                    task_name = f"{task['name']}-{tenant_id}"
+                    logger.debug(f"Creating task configuration for {task_name}")
+                    new_task = {
+                        "task": task["task"],
+                        "schedule": task["schedule"],
+                        "kwargs": {"tenant_id": tenant_id},
+                    }
+                    if options := task.get("options"):
+                        logger.debug(f"Adding options to task {task_name}: {options}")
+                        new_task["options"] = options
+                    new_beat_schedule[task_name] = new_task
+
+            if self._should_update_schedule(current_schedule, new_beat_schedule):
+                logger.info(
+                    "Schedule update required",
+                    extra={
+                        "new_tasks": len(new_beat_schedule),
+                        "current_tasks": len(current_schedule),
+                    },
+                )
+
+                # Create schedule entries
+                entries = {}
+                for name, entry in new_beat_schedule.items():
+                    entries[name] = self.Entry(
+                        name=name,
+                        app=self.app,
+                        task=entry["task"],
+                        schedule=entry["schedule"],
+                        options=entry.get("options", {}),
+                        kwargs=entry.get("kwargs", {}),
+                    )
+
+                # Update the schedule using the scheduler's methods
+                self.schedule.clear()
+                self.schedule.update(entries)
+
+                # Ensure changes are persisted
+                self.sync()
+
+                logger.info("Schedule update completed successfully")
+            else:
+                logger.info("Schedule is up to date, no changes needed")
+
+        except (AttributeError, KeyError):
+            logger.exception("Failed to process task configuration")
+        except Exception:
+            logger.exception("Unexpected error updating tenant tasks")
+
+    def _should_update_schedule(
+        self, current_schedule: dict, new_schedule: dict
+    ) -> bool:
+        """Compare schedules to determine if an update is needed."""
+        logger.debug("Comparing current and new schedules")
+        current_tasks = set(name for name, _ in current_schedule)
+        new_tasks = set(new_schedule.keys())
+        needs_update = current_tasks != new_tasks
+        logger.debug(f"Schedule update needed: {needs_update}")
+        return needs_update
+
+
+@beat_init.connect
+def on_beat_init(sender: Any, **kwargs: Any) -> None:
+    logger.info("beat_init signal received.")
+
+    # Celery beat shouldn't touch the db at all. But just setting a low minimum here.
+    SqlEngine.set_app_name(POSTGRES_CELERY_BEAT_APP_NAME)
+    SqlEngine.init_engine(pool_size=2, max_overflow=0)
+
+    # Startup checks are not needed in multi-tenant case
+    if MULTI_TENANT:
+        return
+
+    app_base.wait_for_redis(sender, **kwargs)
+
+
+@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)
+
+
+celery_app.conf.beat_scheduler = DynamicTenantScheduler

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

@@ -0,0 +1,97 @@
+import multiprocessing
+from typing import Any
+
+from celery import Celery
+from celery import signals
+from celery import Task
+from celery.signals import celeryd_init
+from celery.signals import worker_init
+from celery.signals import worker_ready
+from celery.signals import worker_shutdown
+
+import danswer.background.celery.apps.app_base as app_base
+from danswer.configs.constants import POSTGRES_CELERY_WORKER_HEAVY_APP_NAME
+from danswer.db.engine import SqlEngine
+from danswer.utils.logger import setup_logger
+from shared_configs.configs import MULTI_TENANT
+
+
+logger = setup_logger()
+
+celery_app = Celery(__name__)
+celery_app.config_from_object("danswer.background.celery.configs.heavy")
+
+
+@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: Any = None, conf: Any = None, **kwargs: Any) -> None:
+    app_base.on_celeryd_init(sender, conf, **kwargs)
+
+
+@worker_init.connect
+def on_worker_init(sender: Any, **kwargs: Any) -> None:
+    logger.info("worker_init signal received.")
+    logger.info(f"Multiprocessing start method: {multiprocessing.get_start_method()}")
+
+    SqlEngine.set_app_name(POSTGRES_CELERY_WORKER_HEAVY_APP_NAME)
+    SqlEngine.init_engine(pool_size=4, max_overflow=12)
+
+    # Startup checks are not needed in multi-tenant case
+    if MULTI_TENANT:
+        return
+
+    app_base.wait_for_redis(sender, **kwargs)
+    app_base.wait_for_db(sender, **kwargs)
+    app_base.wait_for_vespa(sender, **kwargs)
+    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)
+
+
+@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)
+
+
+celery_app.autodiscover_tasks(
+    [
+        "danswer.background.celery.tasks.pruning",
+        "danswer.background.celery.tasks.doc_permission_syncing",
+        "danswer.background.celery.tasks.external_group_syncing",
+    ]
+)

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

@@ -0,0 +1,101 @@
+import multiprocessing
+from typing import Any
+
+from celery import Celery
+from celery import signals
+from celery import Task
+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 danswer.background.celery.apps.app_base as app_base
+from danswer.configs.constants import POSTGRES_CELERY_WORKER_INDEXING_APP_NAME
+from danswer.db.engine import SqlEngine
+from danswer.utils.logger import setup_logger
+from shared_configs.configs import MULTI_TENANT
+
+
+logger = setup_logger()
+
+celery_app = Celery(__name__)
+celery_app.config_from_object("danswer.background.celery.configs.indexing")
+
+
+@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: Any = None, conf: Any = None, **kwargs: Any) -> None:
+    app_base.on_celeryd_init(sender, conf, **kwargs)
+
+
+@worker_init.connect
+def on_worker_init(sender: Any, **kwargs: Any) -> None:
+    logger.info("worker_init signal received.")
+    logger.info(f"Multiprocessing start method: {multiprocessing.get_start_method()}")
+
+    SqlEngine.set_app_name(POSTGRES_CELERY_WORKER_INDEXING_APP_NAME)
+    SqlEngine.init_engine(pool_size=sender.concurrency, max_overflow=sender.concurrency)
+
+    # Startup checks are not needed in multi-tenant case
+    if MULTI_TENANT:
+        return
+
+    app_base.wait_for_redis(sender, **kwargs)
+    app_base.wait_for_db(sender, **kwargs)
+    app_base.wait_for_vespa(sender, **kwargs)
+    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:
+    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)
+
+
+celery_app.autodiscover_tasks(
+    [
+        "danswer.background.celery.tasks.indexing",
+    ]
+)

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

@@ -0,0 +1,97 @@
+import multiprocessing
+from typing import Any
+
+from celery import Celery
+from celery import signals
+from celery import Task
+from celery.signals import celeryd_init
+from celery.signals import worker_init
+from celery.signals import worker_ready
+from celery.signals import worker_shutdown
+
+import danswer.background.celery.apps.app_base as app_base
+from danswer.configs.constants import POSTGRES_CELERY_WORKER_LIGHT_APP_NAME
+from danswer.db.engine import SqlEngine
+from danswer.utils.logger import setup_logger
+from shared_configs.configs import MULTI_TENANT
+
+
+logger = setup_logger()
+
+celery_app = Celery(__name__)
+celery_app.config_from_object("danswer.background.celery.configs.light")
+
+
+@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: Any = None, conf: Any = None, **kwargs: Any) -> None:
+    app_base.on_celeryd_init(sender, conf, **kwargs)
+
+
+@worker_init.connect
+def on_worker_init(sender: Any, **kwargs: Any) -> None:
+    logger.info("worker_init signal received.")
+    logger.info(f"Multiprocessing start method: {multiprocessing.get_start_method()}")
+
+    SqlEngine.set_app_name(POSTGRES_CELERY_WORKER_LIGHT_APP_NAME)
+    SqlEngine.init_engine(pool_size=sender.concurrency, max_overflow=8)
+    # Startup checks are not needed in multi-tenant case
+    if MULTI_TENANT:
+        return
+
+    app_base.wait_for_redis(sender, **kwargs)
+    app_base.wait_for_db(sender, **kwargs)
+    app_base.wait_for_vespa(sender, **kwargs)
+    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)
+
+
+@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)
+
+
+celery_app.autodiscover_tasks(
+    [
+        "danswer.background.celery.tasks.shared",
+        "danswer.background.celery.tasks.vespa",
+        "danswer.background.celery.tasks.connector_deletion",
+        "danswer.background.celery.tasks.doc_permission_syncing",
+    ]
+)

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

@@ -0,0 +1,285 @@
+import multiprocessing
+from typing import Any
+from typing import cast
+
+from celery import bootsteps  # type: ignore
+from celery import Celery
+from celery import signals
+from celery import Task
+from celery.exceptions import WorkerShutdown
+from celery.signals import celeryd_init
+from celery.signals import worker_init
+from celery.signals import worker_ready
+from celery.signals import worker_shutdown
+from redis.lock import Lock as RedisLock
+
+import danswer.background.celery.apps.app_base as app_base
+from danswer.background.celery.apps.app_base import task_logger
+from danswer.background.celery.celery_utils import celery_is_worker_primary
+from danswer.background.celery.tasks.indexing.tasks import (
+    get_unfenced_index_attempt_ids,
+)
+from danswer.configs.constants import CELERY_PRIMARY_WORKER_LOCK_TIMEOUT
+from danswer.configs.constants import DanswerRedisLocks
+from danswer.configs.constants import POSTGRES_CELERY_WORKER_PRIMARY_APP_NAME
+from danswer.db.engine import get_session_with_default_tenant
+from danswer.db.engine import SqlEngine
+from danswer.db.index_attempt import get_index_attempt
+from danswer.db.index_attempt import mark_attempt_canceled
+from danswer.redis.redis_connector_credential_pair import RedisConnectorCredentialPair
+from danswer.redis.redis_connector_delete import RedisConnectorDelete
+from danswer.redis.redis_connector_doc_perm_sync import RedisConnectorPermissionSync
+from danswer.redis.redis_connector_ext_group_sync import RedisConnectorExternalGroupSync
+from danswer.redis.redis_connector_index import RedisConnectorIndex
+from danswer.redis.redis_connector_prune import RedisConnectorPrune
+from danswer.redis.redis_connector_stop import RedisConnectorStop
+from danswer.redis.redis_document_set import RedisDocumentSet
+from danswer.redis.redis_pool import get_redis_client
+from danswer.redis.redis_usergroup import RedisUserGroup
+from danswer.utils.logger import setup_logger
+from shared_configs.configs import MULTI_TENANT
+
+logger = setup_logger()
+
+celery_app = Celery(__name__)
+celery_app.config_from_object("danswer.background.celery.configs.primary")
+
+
+@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: Any = None, conf: Any = None, **kwargs: Any) -> None:
+    app_base.on_celeryd_init(sender, conf, **kwargs)
+
+
+@worker_init.connect
+def on_worker_init(sender: Any, **kwargs: Any) -> None:
+    logger.info("worker_init signal received.")
+    logger.info(f"Multiprocessing start method: {multiprocessing.get_start_method()}")
+
+    SqlEngine.set_app_name(POSTGRES_CELERY_WORKER_PRIMARY_APP_NAME)
+    SqlEngine.init_engine(pool_size=8, max_overflow=0)
+
+    # Startup checks are not needed in multi-tenant case
+    if MULTI_TENANT:
+        return
+
+    app_base.wait_for_redis(sender, **kwargs)
+    app_base.wait_for_db(sender, **kwargs)
+    app_base.wait_for_vespa(sender, **kwargs)
+
+    logger.info("Running as the primary celery worker.")
+
+    # This is singleton work that should be done on startup exactly once
+    # by the primary worker. This is unnecessary in the multi tenant scenario
+    r = get_redis_client(tenant_id=None)
+
+    # Log the role and slave count - being connected to a slave or slave count > 0 could be problematic
+    info: dict[str, Any] = cast(dict, r.info("replication"))
+    role: str = cast(str, info.get("role"))
+    connected_slaves: int = info.get("connected_slaves", 0)
+
+    logger.info(
+        f"Redis INFO REPLICATION: role={role} connected_slaves={connected_slaves}"
+    )
+
+    # For the moment, we're assuming that we are the only primary worker
+    # that should be running.
+    # TODO: maybe check for or clean up another zombie primary worker if we detect it
+    r.delete(DanswerRedisLocks.PRIMARY_WORKER)
+
+    # this process wide lock is taken to help other workers start up in order.
+    # it is planned to use this lock to enforce singleton behavior on the primary
+    # worker, since the primary worker does redis cleanup on startup, but this isn't
+    # implemented yet.
+
+    # set thread_local=False since we don't control what thread the periodic task might
+    # reacquire the lock with
+    lock: RedisLock = r.lock(
+        DanswerRedisLocks.PRIMARY_WORKER,
+        timeout=CELERY_PRIMARY_WORKER_LOCK_TIMEOUT,
+        thread_local=False,
+    )
+
+    logger.info("Primary worker lock: Acquire starting.")
+    acquired = lock.acquire(blocking_timeout=CELERY_PRIMARY_WORKER_LOCK_TIMEOUT / 2)
+    if acquired:
+        logger.info("Primary worker lock: Acquire succeeded.")
+    else:
+        logger.error("Primary worker lock: Acquire failed!")
+        raise WorkerShutdown("Primary worker lock could not be acquired!")
+
+    # tacking on our own user data to the sender
+    sender.primary_worker_lock = lock
+
+    # As currently designed, when this worker starts as "primary", we reinitialize redis
+    # to a clean state (for our purposes, anyway)
+    r.delete(DanswerRedisLocks.CHECK_VESPA_SYNC_BEAT_LOCK)
+    r.delete(DanswerRedisLocks.MONITOR_VESPA_SYNC_BEAT_LOCK)
+
+    r.delete(RedisConnectorCredentialPair.get_taskset_key())
+    r.delete(RedisConnectorCredentialPair.get_fence_key())
+
+    RedisDocumentSet.reset_all(r)
+
+    RedisUserGroup.reset_all(r)
+
+    RedisConnectorDelete.reset_all(r)
+
+    RedisConnectorPrune.reset_all(r)
+
+    RedisConnectorIndex.reset_all(r)
+
+    RedisConnectorStop.reset_all(r)
+
+    RedisConnectorPermissionSync.reset_all(r)
+
+    RedisConnectorExternalGroupSync.reset_all(r)
+
+    # mark orphaned index attempts as failed
+    with get_session_with_default_tenant() as db_session:
+        unfenced_attempt_ids = get_unfenced_index_attempt_ids(db_session, r)
+        for attempt_id in unfenced_attempt_ids:
+            attempt = get_index_attempt(db_session, attempt_id)
+            if not attempt:
+                continue
+
+            failure_reason = (
+                f"Canceling leftover index attempt found on startup: "
+                f"index_attempt={attempt.id} "
+                f"cc_pair={attempt.connector_credential_pair_id} "
+                f"search_settings={attempt.search_settings_id}"
+            )
+            logger.warning(failure_reason)
+            mark_attempt_canceled(attempt.id, db_session, failure_reason)
+
+
+@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)
+
+
+@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)
+
+
+class HubPeriodicTask(bootsteps.StartStopStep):
+    """Regularly reacquires the primary worker lock outside of the task queue.
+    Use the task_logger in this class to avoid double logging.
+
+    This cannot be done inside a regular beat task because it must run on schedule and
+    a queue of existing work would starve the task from running.
+    """
+
+    # it's unclear to me whether using the hub's timer or the bootstep timer is better
+    requires = {"celery.worker.components:Hub"}
+
+    def __init__(self, worker: Any, **kwargs: Any) -> None:
+        self.interval = CELERY_PRIMARY_WORKER_LOCK_TIMEOUT / 8  # Interval in seconds
+        self.task_tref = None
+
+    def start(self, worker: Any) -> None:
+        if not celery_is_worker_primary(worker):
+            return
+
+        # Access the worker's event loop (hub)
+        hub = worker.consumer.controller.hub
+
+        # Schedule the periodic task
+        self.task_tref = hub.call_repeatedly(
+            self.interval, self.run_periodic_task, worker
+        )
+        task_logger.info("Scheduled periodic task with hub.")
+
+    def run_periodic_task(self, worker: Any) -> None:
+        try:
+            if not celery_is_worker_primary(worker):
+                return
+
+            if not hasattr(worker, "primary_worker_lock"):
+                return
+
+            lock: RedisLock = worker.primary_worker_lock
+
+            r = get_redis_client(tenant_id=None)
+
+            if lock.owned():
+                task_logger.debug("Reacquiring primary worker lock.")
+                lock.reacquire()
+            else:
+                task_logger.warning(
+                    "Full acquisition of primary worker lock. "
+                    "Reasons could be worker restart or lock expiration."
+                )
+                lock = r.lock(
+                    DanswerRedisLocks.PRIMARY_WORKER,
+                    timeout=CELERY_PRIMARY_WORKER_LOCK_TIMEOUT,
+                )
+
+                task_logger.info("Primary worker lock: Acquire starting.")
+                acquired = lock.acquire(
+                    blocking_timeout=CELERY_PRIMARY_WORKER_LOCK_TIMEOUT / 2
+                )
+                if acquired:
+                    task_logger.info("Primary worker lock: Acquire succeeded.")
+                    worker.primary_worker_lock = lock
+                else:
+                    task_logger.error("Primary worker lock: Acquire failed!")
+                    raise TimeoutError("Primary worker lock could not be acquired!")
+
+        except Exception:
+            task_logger.exception("Periodic task failed.")
+
+    def stop(self, worker: Any) -> None:
+        # Cancel the scheduled task when the worker stops
+        if self.task_tref:
+            self.task_tref.cancel()
+            task_logger.info("Canceled periodic task with hub.")
+
+
+celery_app.steps["worker"].add(HubPeriodicTask)
+
+celery_app.autodiscover_tasks(
+    [
+        "danswer.background.celery.tasks.connector_deletion",
+        "danswer.background.celery.tasks.indexing",
+        "danswer.background.celery.tasks.periodic",
+        "danswer.background.celery.tasks.doc_permission_syncing",
+        "danswer.background.celery.tasks.external_group_syncing",
+        "danswer.background.celery.tasks.pruning",
+        "danswer.background.celery.tasks.shared",
+        "danswer.background.celery.tasks.vespa",
+    ]
+)

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

@@ -0,0 +1,26 @@
+import logging
+
+from celery import current_task
+
+from danswer.utils.logger import ColoredFormatter
+from danswer.utils.logger import PlainFormatter
+
+
+class CeleryTaskPlainFormatter(PlainFormatter):
+    def format(self, record: logging.LogRecord) -> str:
+        task = current_task
+        if task and task.request:
+            record.__dict__.update(task_id=task.request.id, task_name=task.name)
+            record.msg = f"[{task.name}({task.request.id})] {record.msg}"
+
+        return super().format(record)
+
+
+class CeleryTaskColoredFormatter(ColoredFormatter):
+    def format(self, record: logging.LogRecord) -> str:
+        task = current_task
+        if task and task.request:
+            record.__dict__.update(task_id=task.request.id, task_name=task.name)
+            record.msg = f"[{task.name}({task.request.id})] {record.msg}"
+
+        return super().format(record)

backend/danswer/background/celery/celery_app.py

@@ -1,921 +0,0 @@
-import json
-from datetime import timedelta
-from typing import Any
-from typing import cast
-
-import redis
-from celery import Celery
-from celery import signals
-from celery import Task
-from celery.contrib.abortable import AbortableTask  # type: ignore
-from celery.exceptions import SoftTimeLimitExceeded
-from celery.exceptions import TaskRevokedError
-from celery.signals import beat_init
-from celery.signals import worker_init
-from celery.states import READY_STATES
-from celery.utils.log import get_task_logger
-from redis import Redis
-from sqlalchemy import inspect
-from sqlalchemy import text
-from sqlalchemy.orm import Session
-
-from danswer.access.access import get_access_for_document
-from danswer.background.celery.celery_redis import RedisConnectorCredentialPair
-from danswer.background.celery.celery_redis import RedisDocumentSet
-from danswer.background.celery.celery_redis import RedisUserGroup
-from danswer.background.celery.celery_utils import extract_ids_from_runnable_connector
-from danswer.background.celery.celery_utils import should_kick_off_deletion_of_cc_pair
-from danswer.background.celery.celery_utils import should_prune_cc_pair
-from danswer.background.connector_deletion import delete_connector_credential_pair
-from danswer.background.connector_deletion import delete_connector_credential_pair_batch
-from danswer.background.task_utils import build_celery_task_wrapper
-from danswer.background.task_utils import name_cc_cleanup_task
-from danswer.background.task_utils import name_cc_prune_task
-from danswer.configs.app_configs import JOB_TIMEOUT
-from danswer.configs.constants import CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT
-from danswer.configs.constants import DanswerCeleryPriority
-from danswer.configs.constants import DanswerRedisLocks
-from danswer.configs.constants import POSTGRES_CELERY_BEAT_APP_NAME
-from danswer.configs.constants import POSTGRES_CELERY_WORKER_APP_NAME
-from danswer.configs.constants import PostgresAdvisoryLocks
-from danswer.connectors.factory import instantiate_connector
-from danswer.connectors.models import InputType
-from danswer.db.connector_credential_pair import (
-    get_connector_credential_pair,
-)
-from danswer.db.connector_credential_pair import get_connector_credential_pairs
-from danswer.db.deletion_attempt import check_deletion_attempt_is_allowed
-from danswer.db.document import count_documents_by_needs_sync
-from danswer.db.document import get_document
-from danswer.db.document import get_documents_for_connector_credential_pair
-from danswer.db.document import mark_document_as_synced
-from danswer.db.document_set import delete_document_set
-from danswer.db.document_set import fetch_document_set_for_document
-from danswer.db.document_set import fetch_document_sets
-from danswer.db.document_set import get_document_set_by_id
-from danswer.db.document_set import mark_document_set_as_synced
-from danswer.db.engine import get_sqlalchemy_engine
-from danswer.db.engine import init_sqlalchemy_engine
-from danswer.db.models import DocumentSet
-from danswer.db.models import UserGroup
-from danswer.document_index.document_index_utils import get_both_index_names
-from danswer.document_index.factory import get_default_document_index
-from danswer.document_index.interfaces import UpdateRequest
-from danswer.redis.redis_pool import RedisPool
-from danswer.utils.logger import setup_logger
-from danswer.utils.variable_functionality import fetch_versioned_implementation
-from danswer.utils.variable_functionality import (
-    fetch_versioned_implementation_with_fallback,
-)
-from danswer.utils.variable_functionality import noop_fallback
-
-logger = setup_logger()
-
-# use this within celery tasks to get celery task specific logging
-task_logger = get_task_logger(__name__)
-
-redis_pool = RedisPool()
-
-celery_app = Celery(__name__)
-celery_app.config_from_object(
-    "danswer.background.celery.celeryconfig"
-)  # Load configuration from 'celeryconfig.py'
-
-
-#####
-# Tasks that need to be run in job queue, registered via APIs
-#
-# If imports from this module are needed, use local imports to avoid circular importing
-#####
-@build_celery_task_wrapper(name_cc_cleanup_task)
-@celery_app.task(soft_time_limit=JOB_TIMEOUT)
-def cleanup_connector_credential_pair_task(
-    connector_id: int,
-    credential_id: int,
-) -> int:
-    """Connector deletion task. This is run as an async task because it is a somewhat slow job.
-    Needs to potentially update a large number of Postgres and Vespa docs, including deleting them
-    or updating the ACL"""
-    engine = get_sqlalchemy_engine()
-    with Session(engine) as db_session:
-        # validate that the connector / credential pair is deletable
-        cc_pair = get_connector_credential_pair(
-            db_session=db_session,
-            connector_id=connector_id,
-            credential_id=credential_id,
-        )
-        if not cc_pair:
-            raise ValueError(
-                f"Cannot run deletion attempt - connector_credential_pair with Connector ID: "
-                f"{connector_id} and Credential ID: {credential_id} does not exist."
-            )
-
-        deletion_attempt_disallowed_reason = check_deletion_attempt_is_allowed(
-            connector_credential_pair=cc_pair, db_session=db_session
-        )
-        if deletion_attempt_disallowed_reason:
-            raise ValueError(deletion_attempt_disallowed_reason)
-
-        try:
-            # The bulk of the work is in here, updates Postgres and Vespa
-            curr_ind_name, sec_ind_name = get_both_index_names(db_session)
-            document_index = get_default_document_index(
-                primary_index_name=curr_ind_name, secondary_index_name=sec_ind_name
-            )
-            return delete_connector_credential_pair(
-                db_session=db_session,
-                document_index=document_index,
-                cc_pair=cc_pair,
-            )
-        except Exception as e:
-            task_logger.exception(
-                f"Failed to run connector_deletion. "
-                f"connector_id={connector_id} credential_id={credential_id}"
-            )
-            raise e
-
-
-@build_celery_task_wrapper(name_cc_prune_task)
-@celery_app.task(soft_time_limit=JOB_TIMEOUT)
-def prune_documents_task(connector_id: int, credential_id: int) -> None:
-    """connector pruning task. For a cc pair, this task pulls all document IDs from the source
-    and compares those IDs to locally stored documents and deletes all locally stored IDs missing
-    from the most recently pulled document ID list"""
-    with Session(get_sqlalchemy_engine()) as db_session:
-        try:
-            cc_pair = get_connector_credential_pair(
-                db_session=db_session,
-                connector_id=connector_id,
-                credential_id=credential_id,
-            )
-
-            if not cc_pair:
-                task_logger.warning(
-                    f"ccpair not found for {connector_id} {credential_id}"
-                )
-                return
-
-            runnable_connector = instantiate_connector(
-                cc_pair.connector.source,
-                InputType.PRUNE,
-                cc_pair.connector.connector_specific_config,
-                cc_pair.credential,
-                db_session,
-            )
-
-            all_connector_doc_ids: set[str] = extract_ids_from_runnable_connector(
-                runnable_connector
-            )
-
-            all_indexed_document_ids = {
-                doc.id
-                for doc in get_documents_for_connector_credential_pair(
-                    db_session=db_session,
-                    connector_id=connector_id,
-                    credential_id=credential_id,
-                )
-            }
-
-            doc_ids_to_remove = list(all_indexed_document_ids - all_connector_doc_ids)
-
-            curr_ind_name, sec_ind_name = get_both_index_names(db_session)
-            document_index = get_default_document_index(
-                primary_index_name=curr_ind_name, secondary_index_name=sec_ind_name
-            )
-
-            if len(doc_ids_to_remove) == 0:
-                task_logger.info(
-                    f"No docs to prune from {cc_pair.connector.source} connector"
-                )
-                return
-
-            task_logger.info(
-                f"pruning {len(doc_ids_to_remove)} doc(s) from {cc_pair.connector.source} connector"
-            )
-            delete_connector_credential_pair_batch(
-                document_ids=doc_ids_to_remove,
-                connector_id=connector_id,
-                credential_id=credential_id,
-                document_index=document_index,
-            )
-        except Exception as e:
-            task_logger.exception(
-                f"Failed to run pruning for connector id {connector_id}."
-            )
-            raise e
-
-
-def try_generate_stale_document_sync_tasks(
-    db_session: Session, r: Redis, lock_beat: redis.lock.Lock
-) -> int | None:
-    # the fence is up, do nothing
-    if r.exists(RedisConnectorCredentialPair.get_fence_key()):
-        return None
-
-    r.delete(RedisConnectorCredentialPair.get_taskset_key())  # delete the taskset
-
-    # add tasks to celery and build up the task set to monitor in redis
-    stale_doc_count = count_documents_by_needs_sync(db_session)
-    if stale_doc_count == 0:
-        return None
-
-    task_logger.info(
-        f"Stale documents found (at least {stale_doc_count}). Generating sync tasks by cc pair."
-    )
-
-    # rkuo: we could technically sync all stale docs in one big pass.
-    # but I feel it's more understandable to group the docs by cc_pair
-    total_tasks_generated = 0
-    cc_pairs = get_connector_credential_pairs(db_session)
-    for cc_pair in cc_pairs:
-        rc = RedisConnectorCredentialPair(cc_pair.id)
-        tasks_generated = rc.generate_tasks(celery_app, db_session, r, lock_beat)
-
-        if tasks_generated is None:
-            continue
-
-        if tasks_generated == 0:
-            continue
-
-        task_logger.info(
-            f"RedisConnector.generate_tasks finished. "
-            f"cc_pair_id={cc_pair.id} tasks_generated={tasks_generated}"
-        )
-
-        total_tasks_generated += tasks_generated
-
-    task_logger.info(
-        f"All per connector generate_tasks finished. total_tasks_generated={total_tasks_generated}"
-    )
-
-    r.set(RedisConnectorCredentialPair.get_fence_key(), total_tasks_generated)
-    return total_tasks_generated
-
-
-def try_generate_document_set_sync_tasks(
-    document_set: DocumentSet, db_session: Session, r: Redis, lock_beat: redis.lock.Lock
-) -> int | None:
-    lock_beat.reacquire()
-
-    rds = RedisDocumentSet(document_set.id)
-
-    # don't generate document set sync tasks if tasks are still pending
-    if r.exists(rds.fence_key):
-        return None
-
-    # don't generate sync tasks if we're up to date
-    if document_set.is_up_to_date:
-        return None
-
-    # add tasks to celery and build up the task set to monitor in redis
-    r.delete(rds.taskset_key)
-
-    task_logger.info(
-        f"RedisDocumentSet.generate_tasks starting. document_set_id={document_set.id}"
-    )
-
-    # Add all documents that need to be updated into the queue
-    tasks_generated = rds.generate_tasks(celery_app, db_session, r, lock_beat)
-    if tasks_generated is None:
-        return None
-
-    # Currently we are allowing the sync to proceed with 0 tasks.
-    # It's possible for sets/groups to be generated initially with no entries
-    # and they still need to be marked as up to date.
-    # if tasks_generated == 0:
-    #     return 0
-
-    task_logger.info(
-        f"RedisDocumentSet.generate_tasks finished. "
-        f"document_set_id={document_set.id} tasks_generated={tasks_generated}"
-    )
-
-    # set this only after all tasks have been added
-    r.set(rds.fence_key, tasks_generated)
-    return tasks_generated
-
-
-def try_generate_user_group_sync_tasks(
-    usergroup: UserGroup, db_session: Session, r: Redis, lock_beat: redis.lock.Lock
-) -> int | None:
-    lock_beat.reacquire()
-
-    rug = RedisUserGroup(usergroup.id)
-
-    # don't generate sync tasks if tasks are still pending
-    if r.exists(rug.fence_key):
-        return None
-
-    if usergroup.is_up_to_date:
-        return None
-
-    # add tasks to celery and build up the task set to monitor in redis
-    r.delete(rug.taskset_key)
-
-    # Add all documents that need to be updated into the queue
-    task_logger.info(f"generate_tasks starting. usergroup_id={usergroup.id}")
-    tasks_generated = rug.generate_tasks(celery_app, db_session, r, lock_beat)
-    if tasks_generated is None:
-        return None
-
-    # Currently we are allowing the sync to proceed with 0 tasks.
-    # It's possible for sets/groups to be generated initially with no entries
-    # and they still need to be marked as up to date.
-    # if tasks_generated == 0:
-    #     return 0
-
-    task_logger.info(
-        f"generate_tasks finished. "
-        f"usergroup_id={usergroup.id} tasks_generated={tasks_generated}"
-    )
-
-    # set this only after all tasks have been added
-    r.set(rug.fence_key, tasks_generated)
-    return tasks_generated
-
-
-#####
-# Periodic Tasks
-#####
-@celery_app.task(
-    name="check_for_vespa_sync_task",
-    soft_time_limit=JOB_TIMEOUT,
-)
-def check_for_vespa_sync_task() -> None:
-    """Runs periodically to check if any document needs syncing.
-    Generates sets of tasks for Celery if syncing is needed."""
-
-    r = redis_pool.get_client()
-
-    lock_beat = r.lock(
-        DanswerRedisLocks.CHECK_VESPA_SYNC_BEAT_LOCK,
-        timeout=CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT,
-    )
-
-    try:
-        # these tasks should never overlap
-        if not lock_beat.acquire(blocking=False):
-            return
-
-        with Session(get_sqlalchemy_engine()) as db_session:
-            try_generate_stale_document_sync_tasks(db_session, r, lock_beat)
-
-            # check if any document sets are not synced
-            document_set_info = fetch_document_sets(
-                user_id=None, db_session=db_session, include_outdated=True
-            )
-            for document_set, _ in document_set_info:
-                try_generate_document_set_sync_tasks(
-                    document_set, db_session, r, lock_beat
-                )
-
-            # check if any user groups are not synced
-            try:
-                fetch_user_groups = fetch_versioned_implementation(
-                    "danswer.db.user_group", "fetch_user_groups"
-                )
-
-                user_groups = fetch_user_groups(
-                    db_session=db_session, only_up_to_date=False
-                )
-                for usergroup in user_groups:
-                    try_generate_user_group_sync_tasks(
-                        usergroup, db_session, r, lock_beat
-                    )
-            except ModuleNotFoundError:
-                # Always exceptions on the MIT version, which is expected
-                pass
-    except SoftTimeLimitExceeded:
-        task_logger.info(
-            "Soft time limit exceeded, task is being terminated gracefully."
-        )
-    except Exception:
-        task_logger.exception("Unexpected exception")
-    finally:
-        if lock_beat.owned():
-            lock_beat.release()
-
-
-@celery_app.task(
-    name="check_for_cc_pair_deletion_task",
-    soft_time_limit=JOB_TIMEOUT,
-)
-def check_for_cc_pair_deletion_task() -> None:
-    """Runs periodically to check if any deletion tasks should be run"""
-    with Session(get_sqlalchemy_engine()) as db_session:
-        # check if any cc pairs are up for deletion
-        cc_pairs = get_connector_credential_pairs(db_session)
-        for cc_pair in cc_pairs:
-            if should_kick_off_deletion_of_cc_pair(cc_pair, db_session):
-                task_logger.info(
-                    f"Deleting the {cc_pair.name} connector credential pair"
-                )
-                cleanup_connector_credential_pair_task.apply_async(
-                    kwargs=dict(
-                        connector_id=cc_pair.connector.id,
-                        credential_id=cc_pair.credential.id,
-                    ),
-                )
-
-
-@celery_app.task(
-    name="kombu_message_cleanup_task",
-    soft_time_limit=JOB_TIMEOUT,
-    bind=True,
-    base=AbortableTask,
-)
-def kombu_message_cleanup_task(self: Any) -> int:
-    """Runs periodically to clean up the kombu_message table"""
-
-    # we will select messages older than this amount to clean up
-    KOMBU_MESSAGE_CLEANUP_AGE = 7  # days
-    KOMBU_MESSAGE_CLEANUP_PAGE_LIMIT = 1000
-
-    ctx = {}
-    ctx["last_processed_id"] = 0
-    ctx["deleted"] = 0
-    ctx["cleanup_age"] = KOMBU_MESSAGE_CLEANUP_AGE
-    ctx["page_limit"] = KOMBU_MESSAGE_CLEANUP_PAGE_LIMIT
-    with Session(get_sqlalchemy_engine()) as db_session:
-        # Exit the task if we can't take the advisory lock
-        result = db_session.execute(
-            text("SELECT pg_try_advisory_lock(:id)"),
-            {"id": PostgresAdvisoryLocks.KOMBU_MESSAGE_CLEANUP_LOCK_ID.value},
-        ).scalar()
-        if not result:
-            return 0
-
-        while True:
-            if self.is_aborted():
-                raise TaskRevokedError("kombu_message_cleanup_task was aborted.")
-
-            b = kombu_message_cleanup_task_helper(ctx, db_session)
-            if not b:
-                break
-
-            db_session.commit()
-
-    if ctx["deleted"] > 0:
-        task_logger.info(
-            f"Deleted {ctx['deleted']} orphaned messages from kombu_message."
-        )
-
-    return ctx["deleted"]
-
-
-def kombu_message_cleanup_task_helper(ctx: dict, db_session: Session) -> bool:
-    """
-    Helper function to clean up old messages from the `kombu_message` table that are no longer relevant.
-
-    This function retrieves messages from the `kombu_message` table that are no longer visible and
-    older than a specified interval. It checks if the corresponding task_id exists in the
-    `celery_taskmeta` table. If the task_id does not exist, the message is deleted.
-
-    Args:
-        ctx (dict): A context dictionary containing configuration parameters such as:
-            - 'cleanup_age' (int): The age in days after which messages are considered old.
-            - 'page_limit' (int): The maximum number of messages to process in one batch.
-            - 'last_processed_id' (int): The ID of the last processed message to handle pagination.
-            - 'deleted' (int): A counter to track the number of deleted messages.
-        db_session (Session): The SQLAlchemy database session for executing queries.
-
-    Returns:
-        bool: Returns True if there are more rows to process, False if not.
-    """
-
-    inspector = inspect(db_session.bind)
-    if not inspector:
-        return False
-
-    # With the move to redis as celery's broker and backend, kombu tables may not even exist.
-    # We can fail silently.
-    if not inspector.has_table("kombu_message"):
-        return False
-
-    query = text(
-        """
-    SELECT id, timestamp, payload
-    FROM kombu_message WHERE visible = 'false'
-    AND timestamp < CURRENT_TIMESTAMP - INTERVAL :interval_days
-    AND id > :last_processed_id
-    ORDER BY id
-    LIMIT :page_limit
-"""
-    )
-    kombu_messages = db_session.execute(
-        query,
-        {
-            "interval_days": f"{ctx['cleanup_age']} days",
-            "page_limit": ctx["page_limit"],
-            "last_processed_id": ctx["last_processed_id"],
-        },
-    ).fetchall()
-
-    if len(kombu_messages) == 0:
-        return False
-
-    for msg in kombu_messages:
-        payload = json.loads(msg[2])
-        task_id = payload["headers"]["id"]
-
-        # Check if task_id exists in celery_taskmeta
-        task_exists = db_session.execute(
-            text("SELECT 1 FROM celery_taskmeta WHERE task_id = :task_id"),
-            {"task_id": task_id},
-        ).fetchone()
-
-        # If task_id does not exist, delete the message
-        if not task_exists:
-            result = db_session.execute(
-                text("DELETE FROM kombu_message WHERE id = :message_id"),
-                {"message_id": msg[0]},
-            )
-            if result.rowcount > 0:  # type: ignore
-                ctx["deleted"] += 1
-
-        ctx["last_processed_id"] = msg[0]
-
-    return True
-
-
-@celery_app.task(
-    name="check_for_prune_task",
-    soft_time_limit=JOB_TIMEOUT,
-)
-def check_for_prune_task() -> None:
-    """Runs periodically to check if any prune tasks should be run and adds them
-    to the queue"""
-
-    with Session(get_sqlalchemy_engine()) as db_session:
-        all_cc_pairs = get_connector_credential_pairs(db_session)
-
-        for cc_pair in all_cc_pairs:
-            if should_prune_cc_pair(
-                connector=cc_pair.connector,
-                credential=cc_pair.credential,
-                db_session=db_session,
-            ):
-                task_logger.info(f"Pruning the {cc_pair.connector.name} connector")
-
-                prune_documents_task.apply_async(
-                    kwargs=dict(
-                        connector_id=cc_pair.connector.id,
-                        credential_id=cc_pair.credential.id,
-                    )
-                )
-
-
-@celery_app.task(
-    name="vespa_metadata_sync_task",
-    bind=True,
-    soft_time_limit=45,
-    time_limit=60,
-    max_retries=3,
-)
-def vespa_metadata_sync_task(self: Task, document_id: str) -> bool:
-    task_logger.info(f"document_id={document_id}")
-
-    try:
-        with Session(get_sqlalchemy_engine()) as db_session:
-            curr_ind_name, sec_ind_name = get_both_index_names(db_session)
-            document_index = get_default_document_index(
-                primary_index_name=curr_ind_name, secondary_index_name=sec_ind_name
-            )
-
-            doc = get_document(document_id, db_session)
-            if not doc:
-                return False
-
-            # document set sync
-            doc_sets = fetch_document_set_for_document(document_id, db_session)
-            update_doc_sets: set[str] = set(doc_sets)
-
-            # User group sync
-            doc_access = get_access_for_document(
-                document_id=document_id, db_session=db_session
-            )
-            update_request = UpdateRequest(
-                document_ids=[document_id],
-                document_sets=update_doc_sets,
-                access=doc_access,
-                boost=doc.boost,
-                hidden=doc.hidden,
-            )
-
-            # update Vespa
-            document_index.update(update_requests=[update_request])
-
-            # update db last. Worst case = we crash right before this and
-            # the sync might repeat again later
-            mark_document_as_synced(document_id, db_session)
-    except SoftTimeLimitExceeded:
-        task_logger.info(f"SoftTimeLimitExceeded exception. doc_id={document_id}")
-    except Exception as e:
-        task_logger.exception("Unexpected exception")
-
-        # Exponential backoff from 2^4 to 2^6 ... i.e. 16, 32, 64
-        countdown = 2 ** (self.request.retries + 4)
-        self.retry(exc=e, countdown=countdown)
-
-    return True
-
-
-@signals.task_postrun.connect
-def celery_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:
-    """We handle this signal in order to remove completed tasks
-    from their respective tasksets. This allows us to track the progress of document set
-    and user group syncs.
-
-    This function runs after any task completes (both success and failure)
-    Note that this signal does not fire on a task that failed to complete and is going
-    to be retried.
-    """
-    if not task:
-        return
-
-    task_logger.debug(f"Task {task.name} (ID: {task_id}) completed with state: {state}")
-    # logger.debug(f"Result: {retval}")
-
-    if state not in READY_STATES:
-        return
-
-    if not task_id:
-        return
-
-    if task_id.startswith(RedisConnectorCredentialPair.PREFIX):
-        r = redis_pool.get_client()
-        r.srem(RedisConnectorCredentialPair.get_taskset_key(), task_id)
-        return
-
-    if task_id.startswith(RedisDocumentSet.PREFIX):
-        r = redis_pool.get_client()
-        document_set_id = RedisDocumentSet.get_id_from_task_id(task_id)
-        if document_set_id is not None:
-            rds = RedisDocumentSet(document_set_id)
-            r.srem(rds.taskset_key, task_id)
-        return
-
-    if task_id.startswith(RedisUserGroup.PREFIX):
-        r = redis_pool.get_client()
-        usergroup_id = RedisUserGroup.get_id_from_task_id(task_id)
-        if usergroup_id is not None:
-            rug = RedisUserGroup(usergroup_id)
-            r.srem(rug.taskset_key, task_id)
-        return
-
-
-def monitor_connector_taskset(r: Redis) -> None:
-    fence_value = r.get(RedisConnectorCredentialPair.get_fence_key())
-    if fence_value is None:
-        return
-
-    try:
-        initial_count = int(cast(int, fence_value))
-    except ValueError:
-        task_logger.error("The value is not an integer.")
-        return
-
-    count = r.scard(RedisConnectorCredentialPair.get_taskset_key())
-    task_logger.info(f"Stale documents: remaining={count} initial={initial_count}")
-    if count == 0:
-        r.delete(RedisConnectorCredentialPair.get_taskset_key())
-        r.delete(RedisConnectorCredentialPair.get_fence_key())
-        task_logger.info(f"Successfully synced stale documents. count={initial_count}")
-
-
-def monitor_document_set_taskset(
-    key_bytes: bytes, r: Redis, db_session: Session
-) -> None:
-    fence_key = key_bytes.decode("utf-8")
-    document_set_id = RedisDocumentSet.get_id_from_fence_key(fence_key)
-    if document_set_id is None:
-        task_logger.warning("could not parse document set id from {key}")
-        return
-
-    rds = RedisDocumentSet(document_set_id)
-
-    fence_value = r.get(rds.fence_key)
-    if fence_value is None:
-        return
-
-    try:
-        initial_count = int(cast(int, fence_value))
-    except ValueError:
-        task_logger.error("The value is not an integer.")
-        return
-
-    count = cast(int, r.scard(rds.taskset_key))
-    task_logger.info(
-        f"document_set_id={document_set_id} remaining={count} initial={initial_count}"
-    )
-    if count > 0:
-        return
-
-    document_set = cast(
-        DocumentSet,
-        get_document_set_by_id(db_session=db_session, document_set_id=document_set_id),
-    )  # casting since we "know" a document set with this ID exists
-    if document_set:
-        if not document_set.connector_credential_pairs:
-            # if there are no connectors, then delete the document set.
-            delete_document_set(document_set_row=document_set, db_session=db_session)
-            task_logger.info(
-                f"Successfully deleted document set with ID: '{document_set_id}'!"
-            )
-        else:
-            mark_document_set_as_synced(document_set_id, db_session)
-            task_logger.info(
-                f"Successfully synced document set with ID: '{document_set_id}'!"
-            )
-
-    r.delete(rds.taskset_key)
-    r.delete(rds.fence_key)
-
-
-def monitor_usergroup_taskset(key_bytes: bytes, r: Redis, db_session: Session) -> None:
-    key = key_bytes.decode("utf-8")
-    usergroup_id = RedisUserGroup.get_id_from_fence_key(key)
-    if not usergroup_id:
-        task_logger.warning("Could not parse usergroup id from {key}")
-        return
-
-    rug = RedisUserGroup(usergroup_id)
-    fence_value = r.get(rug.fence_key)
-    if fence_value is None:
-        return
-
-    try:
-        initial_count = int(cast(int, fence_value))
-    except ValueError:
-        task_logger.error("The value is not an integer.")
-        return
-
-    count = cast(int, r.scard(rug.taskset_key))
-    task_logger.info(
-        f"usergroup_id={usergroup_id} remaining={count} initial={initial_count}"
-    )
-    if count > 0:
-        return
-
-    try:
-        fetch_user_group = fetch_versioned_implementation(
-            "danswer.db.user_group", "fetch_user_group"
-        )
-    except ModuleNotFoundError:
-        task_logger.exception(
-            "fetch_versioned_implementation failed to look up fetch_user_group."
-        )
-        return
-
-    user_group: UserGroup | None = fetch_user_group(
-        db_session=db_session, user_group_id=usergroup_id
-    )
-    if user_group:
-        if user_group.is_up_for_deletion:
-            delete_user_group = fetch_versioned_implementation_with_fallback(
-                "danswer.db.user_group", "delete_user_group", noop_fallback
-            )
-
-            delete_user_group(db_session=db_session, user_group=user_group)
-            task_logger.info(f" Deleted usergroup. id='{usergroup_id}'")
-        else:
-            mark_user_group_as_synced = fetch_versioned_implementation_with_fallback(
-                "danswer.db.user_group", "mark_user_group_as_synced", noop_fallback
-            )
-
-            mark_user_group_as_synced(db_session=db_session, user_group=user_group)
-            task_logger.info(f"Synced usergroup. id='{usergroup_id}'")
-
-    r.delete(rug.taskset_key)
-    r.delete(rug.fence_key)
-
-
-@celery_app.task(name="monitor_vespa_sync", soft_time_limit=300)
-def monitor_vespa_sync() -> None:
-    """This is a celery beat task that monitors and finalizes metadata sync tasksets.
-    It scans for fence values and then gets the counts of any associated tasksets.
-    If the count is 0, that means all tasks finished and we should clean up.
-
-    This task lock timeout is CELERY_METADATA_SYNC_BEAT_LOCK_TIMEOUT seconds, so don't
-    do anything too expensive in this function!
-    """
-    r = redis_pool.get_client()
-
-    lock_beat = r.lock(
-        DanswerRedisLocks.MONITOR_VESPA_SYNC_BEAT_LOCK,
-        timeout=CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT,
-    )
-
-    try:
-        # prevent overlapping tasks
-        if not lock_beat.acquire(blocking=False):
-            return
-
-        with Session(get_sqlalchemy_engine()) as db_session:
-            if r.exists(RedisConnectorCredentialPair.get_fence_key()):
-                monitor_connector_taskset(r)
-
-            for key_bytes in r.scan_iter(RedisDocumentSet.FENCE_PREFIX + "*"):
-                monitor_document_set_taskset(key_bytes, r, db_session)
-
-            for key_bytes in r.scan_iter(RedisUserGroup.FENCE_PREFIX + "*"):
-                monitor_usergroup_taskset(key_bytes, r, db_session)
-
-        #
-        # r_celery = celery_app.broker_connection().channel().client
-        # length = celery_get_queue_length(DanswerCeleryQueues.VESPA_METADATA_SYNC, r_celery)
-        # task_logger.warning(f"queue={DanswerCeleryQueues.VESPA_METADATA_SYNC} length={length}")
-    except SoftTimeLimitExceeded:
-        task_logger.info(
-            "Soft time limit exceeded, task is being terminated gracefully."
-        )
-    finally:
-        if lock_beat.owned():
-            lock_beat.release()
-
-
-@beat_init.connect
-def on_beat_init(sender: Any, **kwargs: Any) -> None:
-    init_sqlalchemy_engine(POSTGRES_CELERY_BEAT_APP_NAME)
-
-
-@worker_init.connect
-def on_worker_init(sender: Any, **kwargs: Any) -> None:
-    init_sqlalchemy_engine(POSTGRES_CELERY_WORKER_APP_NAME)
-
-    # TODO(rkuo): this is singleton work that should be done on startup exactly once
-    # if we run multiple workers, we'll need to centralize where this cleanup happens
-    r = redis_pool.get_client()
-
-    r.delete(DanswerRedisLocks.CHECK_VESPA_SYNC_BEAT_LOCK)
-    r.delete(DanswerRedisLocks.MONITOR_VESPA_SYNC_BEAT_LOCK)
-
-    r.delete(RedisConnectorCredentialPair.get_taskset_key())
-    r.delete(RedisConnectorCredentialPair.get_fence_key())
-
-    for key in r.scan_iter(RedisDocumentSet.TASKSET_PREFIX + "*"):
-        r.delete(key)
-
-    for key in r.scan_iter(RedisDocumentSet.FENCE_PREFIX + "*"):
-        r.delete(key)
-
-    for key in r.scan_iter(RedisUserGroup.TASKSET_PREFIX + "*"):
-        r.delete(key)
-
-    for key in r.scan_iter(RedisUserGroup.FENCE_PREFIX + "*"):
-        r.delete(key)
-
-
-#####
-# Celery Beat (Periodic Tasks) Settings
-#####
-celery_app.conf.beat_schedule = {
-    "check-for-vespa-sync": {
-        "task": "check_for_vespa_sync_task",
-        "schedule": timedelta(seconds=5),
-        "options": {"priority": DanswerCeleryPriority.HIGH},
-    },
-    "check-for-cc-pair-deletion": {
-        "task": "check_for_cc_pair_deletion_task",
-        # don't need to check too often, since we kick off a deletion initially
-        # during the API call that actually marks the CC pair for deletion
-        "schedule": timedelta(minutes=1),
-        "options": {"priority": DanswerCeleryPriority.HIGH},
-    },
-}
-celery_app.conf.beat_schedule.update(
-    {
-        "check-for-prune": {
-            "task": "check_for_prune_task",
-            "schedule": timedelta(seconds=5),
-            "options": {"priority": DanswerCeleryPriority.HIGH},
-        },
-    }
-)
-celery_app.conf.beat_schedule.update(
-    {
-        "kombu-message-cleanup": {
-            "task": "kombu_message_cleanup_task",
-            "schedule": timedelta(seconds=3600),
-            "options": {"priority": DanswerCeleryPriority.LOWEST},
-        },
-    }
-)
-celery_app.conf.beat_schedule.update(
-    {
-        "monitor-vespa-sync": {
-            "task": "monitor_vespa_sync",
-            "schedule": timedelta(seconds=5),
-            "options": {"priority": DanswerCeleryPriority.HIGH},
-        },
-    }
-)

backend/danswer/background/celery/celery_redis.py

@@ -1,284 +1,10 @@
 # These are helper objects for tracking the keys we need to write in redis
-import time
-from abc import ABC
-from abc import abstractmethod
 from typing import cast
-from uuid import uuid4
 
-import redis
-from celery import Celery
 from redis import Redis
-from sqlalchemy.orm import Session
 
-from danswer.background.celery.celeryconfig import CELERY_SEPARATOR
-from danswer.configs.constants import CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT
+from danswer.background.celery.configs.base import CELERY_SEPARATOR
 from danswer.configs.constants import DanswerCeleryPriority
-from danswer.configs.constants import DanswerCeleryQueues
-from danswer.db.connector_credential_pair import get_connector_credential_pair_from_id
-from danswer.db.document import (
-    construct_document_select_for_connector_credential_pair_by_needs_sync,
-)
-from danswer.db.document_set import construct_document_select_by_docset
-from danswer.utils.variable_functionality import fetch_versioned_implementation
-
-
-class RedisObjectHelper(ABC):
-    PREFIX = "base"
-    FENCE_PREFIX = PREFIX + "_fence"
-    TASKSET_PREFIX = PREFIX + "_taskset"
-
-    def __init__(self, id: int):
-        self._id: int = id
-
-    @property
-    def task_id_prefix(self) -> str:
-        return f"{self.PREFIX}_{self._id}"
-
-    @property
-    def fence_key(self) -> str:
-        # example: documentset_fence_1
-        return f"{self.FENCE_PREFIX}_{self._id}"
-
-    @property
-    def taskset_key(self) -> str:
-        # example: documentset_taskset_1
-        return f"{self.TASKSET_PREFIX}_{self._id}"
-
-    @staticmethod
-    def get_id_from_fence_key(key: str) -> int | None:
-        """
-        Extracts the object ID from a fence key in the format `PREFIX_fence_X`.
-
-        Args:
-            key (str): The fence key string.
-
-        Returns:
-            Optional[int]: The extracted ID if the key is in the correct format, otherwise None.
-        """
-        parts = key.split("_")
-        if len(parts) != 3:
-            return None
-
-        try:
-            object_id = int(parts[2])
-        except ValueError:
-            return None
-
-        return object_id
-
-    @staticmethod
-    def get_id_from_task_id(task_id: str) -> int | None:
-        """
-        Extracts the object ID from a task ID string.
-
-        This method assumes the task ID is formatted as `prefix_objectid_suffix`, where:
-        - `prefix` is an arbitrary string (e.g., the name of the task or entity),
-        - `objectid` is the ID you want to extract,
-        - `suffix` is another arbitrary string (e.g., a UUID).
-
-        Example:
-            If the input `task_id` is `documentset_1_cbfdc96a-80ca-4312-a242-0bb68da3c1dc`,
-            this method will return the string `"1"`.
-
-        Args:
-            task_id (str): The task ID string from which to extract the object ID.
-
-        Returns:
-            str | None: The extracted object ID if the task ID is in the correct format, otherwise None.
-        """
-        # example: task_id=documentset_1_cbfdc96a-80ca-4312-a242-0bb68da3c1dc
-        parts = task_id.split("_")
-        if len(parts) != 3:
-            return None
-
-        try:
-            object_id = int(parts[1])
-        except ValueError:
-            return None
-
-        return object_id
-
-    @abstractmethod
-    def generate_tasks(
-        self,
-        celery_app: Celery,
-        db_session: Session,
-        redis_client: Redis,
-        lock: redis.lock.Lock,
-    ) -> int | None:
-        pass
-
-
-class RedisDocumentSet(RedisObjectHelper):
-    PREFIX = "documentset"
-    FENCE_PREFIX = PREFIX + "_fence"
-    TASKSET_PREFIX = PREFIX + "_taskset"
-
-    def generate_tasks(
-        self,
-        celery_app: Celery,
-        db_session: Session,
-        redis_client: Redis,
-        lock: redis.lock.Lock,
-    ) -> int | None:
-        last_lock_time = time.monotonic()
-
-        async_results = []
-        stmt = construct_document_select_by_docset(self._id)
-        for doc in db_session.scalars(stmt).yield_per(1):
-            current_time = time.monotonic()
-            if current_time - last_lock_time >= (
-                CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT / 4
-            ):
-                lock.reacquire()
-                last_lock_time = current_time
-
-            # celery's default task id format is "dd32ded3-00aa-4884-8b21-42f8332e7fac"
-            # the actual redis key is "celery-task-meta-dd32ded3-00aa-4884-8b21-42f8332e7fac"
-            # we prefix the task id so it's easier to keep track of who created the task
-            # aka "documentset_1_6dd32ded3-00aa-4884-8b21-42f8332e7fac"
-            custom_task_id = f"{self.task_id_prefix}_{uuid4()}"
-
-            # add to the set BEFORE creating the task.
-            redis_client.sadd(self.taskset_key, custom_task_id)
-
-            result = celery_app.send_task(
-                "vespa_metadata_sync_task",
-                kwargs=dict(document_id=doc.id),
-                queue=DanswerCeleryQueues.VESPA_METADATA_SYNC,
-                task_id=custom_task_id,
-                priority=DanswerCeleryPriority.LOW,
-            )
-
-            async_results.append(result)
-
-        return len(async_results)
-
-
-class RedisUserGroup(RedisObjectHelper):
-    PREFIX = "usergroup"
-    FENCE_PREFIX = PREFIX + "_fence"
-    TASKSET_PREFIX = PREFIX + "_taskset"
-
-    def generate_tasks(
-        self,
-        celery_app: Celery,
-        db_session: Session,
-        redis_client: Redis,
-        lock: redis.lock.Lock,
-    ) -> int | None:
-        last_lock_time = time.monotonic()
-
-        async_results = []
-
-        try:
-            construct_document_select_by_usergroup = fetch_versioned_implementation(
-                "danswer.db.user_group",
-                "construct_document_select_by_usergroup",
-            )
-        except ModuleNotFoundError:
-            return 0
-
-        stmt = construct_document_select_by_usergroup(self._id)
-        for doc in db_session.scalars(stmt).yield_per(1):
-            current_time = time.monotonic()
-            if current_time - last_lock_time >= (
-                CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT / 4
-            ):
-                lock.reacquire()
-                last_lock_time = current_time
-
-            # celery's default task id format is "dd32ded3-00aa-4884-8b21-42f8332e7fac"
-            # the actual redis key is "celery-task-meta-dd32ded3-00aa-4884-8b21-42f8332e7fac"
-            # we prefix the task id so it's easier to keep track of who created the task
-            # aka "documentset_1_6dd32ded3-00aa-4884-8b21-42f8332e7fac"
-            custom_task_id = f"{self.task_id_prefix}_{uuid4()}"
-
-            # add to the set BEFORE creating the task.
-            redis_client.sadd(self.taskset_key, custom_task_id)
-
-            result = celery_app.send_task(
-                "vespa_metadata_sync_task",
-                kwargs=dict(document_id=doc.id),
-                queue=DanswerCeleryQueues.VESPA_METADATA_SYNC,
-                task_id=custom_task_id,
-                priority=DanswerCeleryPriority.LOW,
-            )
-
-            async_results.append(result)
-
-        return len(async_results)
-
-
-class RedisConnectorCredentialPair(RedisObjectHelper):
-    PREFIX = "connectorsync"
-    FENCE_PREFIX = PREFIX + "_fence"
-    TASKSET_PREFIX = PREFIX + "_taskset"
-
-    @classmethod
-    def get_fence_key(cls) -> str:
-        return RedisConnectorCredentialPair.FENCE_PREFIX
-
-    @classmethod
-    def get_taskset_key(cls) -> str:
-        return RedisConnectorCredentialPair.TASKSET_PREFIX
-
-    @property
-    def taskset_key(self) -> str:
-        """Notice that this is intentionally reusing the same taskset for all
-        connector syncs"""
-        # example: connector_taskset
-        return f"{self.TASKSET_PREFIX}"
-
-    def generate_tasks(
-        self,
-        celery_app: Celery,
-        db_session: Session,
-        redis_client: Redis,
-        lock: redis.lock.Lock,
-    ) -> int | None:
-        last_lock_time = time.monotonic()
-
-        async_results = []
-        cc_pair = get_connector_credential_pair_from_id(self._id, db_session)
-        if not cc_pair:
-            return None
-
-        stmt = construct_document_select_for_connector_credential_pair_by_needs_sync(
-            cc_pair.connector_id, cc_pair.credential_id
-        )
-        for doc in db_session.scalars(stmt).yield_per(1):
-            current_time = time.monotonic()
-            if current_time - last_lock_time >= (
-                CELERY_VESPA_SYNC_BEAT_LOCK_TIMEOUT / 4
-            ):
-                lock.reacquire()
-                last_lock_time = current_time
-
-            # celery's default task id format is "dd32ded3-00aa-4884-8b21-42f8332e7fac"
-            # the actual redis key is "celery-task-meta-dd32ded3-00aa-4884-8b21-42f8332e7fac"
-            # we prefix the task id so it's easier to keep track of who created the task
-            # aka "documentset_1_6dd32ded3-00aa-4884-8b21-42f8332e7fac"
-            custom_task_id = f"{self.task_id_prefix}_{uuid4()}"
-
-            # add to the tracking taskset in redis BEFORE creating the celery task.
-            # note that for the moment we are using a single taskset key, not differentiated by cc_pair id
-            redis_client.sadd(
-                RedisConnectorCredentialPair.get_taskset_key(), custom_task_id
-            )
-
-            # Priority on sync's triggered by new indexing should be medium
-            result = celery_app.send_task(
-                "vespa_metadata_sync_task",
-                kwargs=dict(document_id=doc.id),
-                queue=DanswerCeleryQueues.VESPA_METADATA_SYNC,
-                task_id=custom_task_id,
-                priority=DanswerCeleryPriority.MEDIUM,
-            )
-
-            async_results.append(result)
-
-        return len(async_results)
 
 
 def celery_get_queue_length(queue: str, r: Redis) -> int:

backend/danswer/background/celery/celery_utils.py

@@ -1,49 +1,65 @@
 from datetime import datetime
 from datetime import timezone
+from typing import Any
 
 from sqlalchemy.orm import Session
 
-from danswer.background.task_utils import name_cc_cleanup_task
-from danswer.background.task_utils import name_cc_prune_task
-from danswer.configs.app_configs import ALLOW_SIMULTANEOUS_PRUNING
 from danswer.configs.app_configs import MAX_PRUNING_DOCUMENT_RETRIEVAL_PER_MINUTE
 from danswer.connectors.cross_connector_utils.rate_limit_wrapper import (
     rate_limit_builder,
 )
 from danswer.connectors.interfaces import BaseConnector
-from danswer.connectors.interfaces import IdConnector
 from danswer.connectors.interfaces import LoadConnector
 from danswer.connectors.interfaces import PollConnector
+from danswer.connectors.interfaces import SlimConnector
 from danswer.connectors.models import Document
-from danswer.db.deletion_attempt import check_deletion_attempt_is_allowed
-from danswer.db.engine import get_db_current_time
-from danswer.db.enums import ConnectorCredentialPairStatus
-from danswer.db.models import Connector
-from danswer.db.models import ConnectorCredentialPair
-from danswer.db.models import Credential
+from danswer.db.connector_credential_pair import get_connector_credential_pair
+from danswer.db.enums import TaskStatus
 from danswer.db.models import TaskQueueState
-from danswer.db.tasks import check_task_is_live_and_not_timed_out
-from danswer.db.tasks import get_latest_task
-from danswer.db.tasks import get_latest_task_by_type
+from danswer.indexing.indexing_heartbeat import IndexingHeartbeatInterface
+from danswer.redis.redis_connector import RedisConnector
 from danswer.server.documents.models import DeletionAttemptSnapshot
 from danswer.utils.logger import setup_logger
 
+
 logger = setup_logger()
 
 
 def _get_deletion_status(
-    connector_id: int, credential_id: int, db_session: Session
+    connector_id: int,
+    credential_id: int,
+    db_session: Session,
+    tenant_id: str | None = None,
 ) -> TaskQueueState | None:
-    cleanup_task_name = name_cc_cleanup_task(
-        connector_id=connector_id, credential_id=credential_id
+    """We no longer store TaskQueueState in the DB for a deletion attempt.
+    This function populates TaskQueueState by just checking redis.
+    """
+    cc_pair = get_connector_credential_pair(
+        connector_id=connector_id, credential_id=credential_id, db_session=db_session
+    )
+    if not cc_pair:
+        return None
+
+    redis_connector = RedisConnector(tenant_id, cc_pair.id)
+    if not redis_connector.delete.fenced:
+        return None
+
+    return TaskQueueState(
+        task_id="",
+        task_name=redis_connector.delete.fence_key,
+        status=TaskStatus.STARTED,
     )
-    return get_latest_task(task_name=cleanup_task_name, db_session=db_session)
 
 
 def get_deletion_attempt_snapshot(
-    connector_id: int, credential_id: int, db_session: Session
+    connector_id: int,
+    credential_id: int,
+    db_session: Session,
+    tenant_id: str | None = None,
 ) -> DeletionAttemptSnapshot | None:
-    deletion_task = _get_deletion_status(connector_id, credential_id, db_session)
+    deletion_task = _get_deletion_status(
+        connector_id, credential_id, db_session, tenant_id
+    )
     if not deletion_task:
         return None
 
@@ -54,85 +70,31 @@ def get_deletion_attempt_snapshot(
     )
 
 
-def should_kick_off_deletion_of_cc_pair(
-    cc_pair: ConnectorCredentialPair, db_session: Session
-) -> bool:
-    if cc_pair.status != ConnectorCredentialPairStatus.DELETING:
-        return False
-
-    if check_deletion_attempt_is_allowed(cc_pair, db_session):
-        return False
-
-    deletion_task = _get_deletion_status(
-        connector_id=cc_pair.connector_id,
-        credential_id=cc_pair.credential_id,
-        db_session=db_session,
-    )
-    if deletion_task and check_task_is_live_and_not_timed_out(
-        deletion_task,
-        db_session,
-        # 1 hour timeout
-        timeout=60 * 60,
-    ):
-        return False
-
-    return True
-
-
-def should_prune_cc_pair(
-    connector: Connector, credential: Credential, db_session: Session
-) -> bool:
-    if not connector.prune_freq:
-        return False
-
-    pruning_task_name = name_cc_prune_task(
-        connector_id=connector.id, credential_id=credential.id
-    )
-    last_pruning_task = get_latest_task(pruning_task_name, db_session)
-    current_db_time = get_db_current_time(db_session)
-
-    if not last_pruning_task:
-        time_since_initialization = current_db_time - connector.time_created
-        if time_since_initialization.total_seconds() >= connector.prune_freq:
-            return True
-        return False
-
-    if not ALLOW_SIMULTANEOUS_PRUNING:
-        pruning_type_task_name = name_cc_prune_task()
-        last_pruning_type_task = get_latest_task_by_type(
-            pruning_type_task_name, db_session
-        )
-
-        if last_pruning_type_task and check_task_is_live_and_not_timed_out(
-            last_pruning_type_task, db_session
-        ):
-            return False
-
-    if check_task_is_live_and_not_timed_out(last_pruning_task, db_session):
-        return False
-
-    if not last_pruning_task.start_time:
-        return False
-
-    time_since_last_pruning = current_db_time - last_pruning_task.start_time
-    return time_since_last_pruning.total_seconds() >= connector.prune_freq
-
-
-def document_batch_to_ids(doc_batch: list[Document]) -> set[str]:
+def document_batch_to_ids(
+    doc_batch: list[Document],
+) -> set[str]:
     return {doc.id for doc in doc_batch}
 
 
-def extract_ids_from_runnable_connector(runnable_connector: BaseConnector) -> set[str]:
+def extract_ids_from_runnable_connector(
+    runnable_connector: BaseConnector,
+    callback: IndexingHeartbeatInterface | None = None,
+) -> set[str]:
     """
-    If the PruneConnector hasnt been implemented for the given connector, just pull
-    all docs using the load_from_state and grab out the IDs
+    If the SlimConnector hasnt been implemented for the given connector, just pull
+    all docs using the load_from_state and grab out the IDs.
+
+    Optionally, a callback can be passed to handle the length of each document batch.
     """
     all_connector_doc_ids: set[str] = set()
 
+    if isinstance(runnable_connector, SlimConnector):
+        for metadata_batch in runnable_connector.retrieve_all_slim_documents():
+            all_connector_doc_ids.update({doc.id for doc in metadata_batch})
+
     doc_batch_generator = None
-    if isinstance(runnable_connector, IdConnector):
-        all_connector_doc_ids = runnable_connector.retrieve_all_source_ids()
-    elif isinstance(runnable_connector, LoadConnector):
+
+    if isinstance(runnable_connector, LoadConnector):
         doc_batch_generator = runnable_connector.load_from_state()
     elif isinstance(runnable_connector, PollConnector):
         start = datetime(1970, 1, 1, tzinfo=timezone.utc).timestamp()
@@ -141,13 +103,46 @@ def extract_ids_from_runnable_connector(runnable_connector: BaseConnector) -> se
     else:
         raise RuntimeError("Pruning job could not find a valid runnable_connector.")
 
-    if doc_batch_generator:
-        doc_batch_processing_func = document_batch_to_ids
-        if MAX_PRUNING_DOCUMENT_RETRIEVAL_PER_MINUTE:
-            doc_batch_processing_func = rate_limit_builder(
-                max_calls=MAX_PRUNING_DOCUMENT_RETRIEVAL_PER_MINUTE, period=60
-            )(document_batch_to_ids)
-        for doc_batch in doc_batch_generator:
-            all_connector_doc_ids.update(doc_batch_processing_func(doc_batch))
+    doc_batch_processing_func = document_batch_to_ids
+    if MAX_PRUNING_DOCUMENT_RETRIEVAL_PER_MINUTE:
+        doc_batch_processing_func = rate_limit_builder(
+            max_calls=MAX_PRUNING_DOCUMENT_RETRIEVAL_PER_MINUTE, period=60
+        )(document_batch_to_ids)
+    for doc_batch in doc_batch_generator:
+        if callback:
+            if callback.should_stop():
+                raise RuntimeError(
+                    "extract_ids_from_runnable_connector: Stop signal detected"
+                )
+
+        all_connector_doc_ids.update(doc_batch_processing_func(doc_batch))
+
+        if callback:
+            callback.progress("extract_ids_from_runnable_connector", len(doc_batch))
 
     return all_connector_doc_ids
+
+
+def celery_is_listening_to_queue(worker: Any, name: str) -> bool:
+    """Checks to see if we're listening to the named queue"""
+
+    # how to get a list of queues this worker is listening to
+    # https://stackoverflow.com/questions/29790523/how-to-determine-which-queues-a-celery-worker-is-consuming-at-runtime
+    queue_names = list(worker.app.amqp.queues.consume_from.keys())
+    for queue_name in queue_names:
+        if queue_name == name:
+            return True
+
+    return False
+
+
+def celery_is_worker_primary(worker: Any) -> bool:
+    """There are multiple approaches that could be taken to determine if a celery worker
+    is 'primary', as defined by us. But the way we do it is to check the hostname set
+    for the celery worker, which can be done on the
+    command line with '--hostname'."""
+    hostname = worker.hostname
+    if hostname.startswith("primary"):
+        return True
+
+    return False

backend/danswer/background/celery/celeryconfig.py

@@ -1,35 +0,0 @@
-# docs: https://docs.celeryq.dev/en/stable/userguide/configuration.html
-from danswer.configs.app_configs import REDIS_DB_NUMBER_CELERY
-from danswer.configs.app_configs import REDIS_HOST
-from danswer.configs.app_configs import REDIS_PASSWORD
-from danswer.configs.app_configs import REDIS_PORT
-from danswer.configs.constants import DanswerCeleryPriority
-
-CELERY_SEPARATOR = ":"
-
-CELERY_PASSWORD_PART = ""
-if REDIS_PASSWORD:
-    CELERY_PASSWORD_PART = f":{REDIS_PASSWORD}@"
-
-# example celery_broker_url: "redis://:password@localhost:6379/15"
-broker_url = (
-    f"redis://{CELERY_PASSWORD_PART}{REDIS_HOST}:{REDIS_PORT}/{REDIS_DB_NUMBER_CELERY}"
-)
-
-result_backend = (
-    f"redis://{CELERY_PASSWORD_PART}{REDIS_HOST}:{REDIS_PORT}/{REDIS_DB_NUMBER_CELERY}"
-)
-
-# NOTE: prefetch 4 is significantly faster than prefetch 1
-# however, prefetching is bad when tasks are lengthy as those tasks
-# can stall other tasks.
-worker_prefetch_multiplier = 4
-
-broker_transport_options = {
-    "priority_steps": list(range(len(DanswerCeleryPriority))),
-    "sep": CELERY_SEPARATOR,
-    "queue_order_strategy": "priority",
-}
-
-task_default_priority = DanswerCeleryPriority.MEDIUM
-task_acks_late = True