Compare commits

..

4 Commits

Author SHA1 Message Date
SubashMohan
c54752e3ba fix non accepted files popup 2025-10-23 18:19:45 +05:30
SubashMohan
bd32795804 fix file item issues 2025-10-23 18:19:45 +05:30
SubashMohan
7aa6b01ac0 Update Truncated component styling for better text handling 2025-10-23 18:19:45 +05:30
SubashMohan
93084a3a39 fix project bugs 2025-10-23 18:19:45 +05:30
857 changed files with 26290 additions and 51104 deletions

View File

@@ -1,42 +0,0 @@
self-hosted-runner:
# Labels of self-hosted runner in array of strings.
labels:
- extras=ecr-cache
- extras=s3-cache
- hdd=256
- runs-on
- runner=1cpu-linux-arm64
- runner=1cpu-linux-x64
- runner=2cpu-linux-arm64
- runner=2cpu-linux-x64
- runner=4cpu-linux-arm64
- runner=4cpu-linux-x64
- runner=8cpu-linux-arm64
- runner=8cpu-linux-x64
- runner=16cpu-linux-arm64
- runner=16cpu-linux-x64
- ubuntu-slim # Currently in public preview
- volume=40gb
# Configuration variables in array of strings defined in your repository or
# organization. `null` means disabling configuration variables check.
# Empty array means no configuration variable is allowed.
config-variables: null
# Configuration for file paths. The keys are glob patterns to match to file
# paths relative to the repository root. The values are the configurations for
# the file paths. Note that the path separator is always '/'.
# The following configurations are available.
#
# "ignore" is an array of regular expression patterns. Matched error messages
# are ignored. This is similar to the "-ignore" command line option.
paths:
# Glob pattern relative to the repository root for matching files. The path separator is always '/'.
# This example configures any YAML file under the '.github/workflows/' directory.
.github/workflows/**/*.{yml,yaml}:
# TODO: These are real and should be fixed eventually.
ignore:
- 'shellcheck reported issue in this script: SC2038:.+'
- 'shellcheck reported issue in this script: SC2046:.+'
- 'shellcheck reported issue in this script: SC2086:.+'
- 'shellcheck reported issue in this script: SC2193:.+'

View File

@@ -59,7 +59,7 @@ runs:
steps:
- name: Build and push Docker image (Attempt 1 of 3)
id: buildx1
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
uses: docker/build-push-action@v6
continue-on-error: true
with:
context: ${{ inputs.context }}
@@ -86,7 +86,7 @@ runs:
- name: Build and push Docker image (Attempt 2 of 3)
id: buildx2
if: steps.buildx1.outcome != 'success'
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
uses: docker/build-push-action@v6
with:
context: ${{ inputs.context }}
file: ${{ inputs.file }}
@@ -112,7 +112,7 @@ runs:
- 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@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
uses: docker/build-push-action@v6
with:
context: ${{ inputs.context }}
file: ${{ inputs.file }}

View File

@@ -1,15 +1,25 @@
name: "Prepare Build (OpenAPI generation)"
description: "Sets up Python with uv, installs deps, generates OpenAPI schema and Python client, uploads artifact"
inputs:
docker-username:
required: true
docker-password:
required: true
runs:
using: "composite"
steps:
- name: Setup Python and Install Dependencies
uses: ./.github/actions/setup-python-and-install-dependencies
- name: Checkout code
uses: actions/checkout@v4
- name: Setup uv
uses: astral-sh/setup-uv@v3
- name: Setup Python
uses: actions/setup-python@v5
with:
python-version: "3.11"
- name: Install Python dependencies with uv
shell: bash
run: |
uv pip install --system \
-r backend/requirements/default.txt \
-r backend/requirements/dev.txt
- name: Generate OpenAPI schema
shell: bash
@@ -19,15 +29,6 @@ runs:
run: |
python scripts/onyx_openapi_schema.py --filename generated/openapi.json
# needed for pulling openapitools/openapi-generator-cli
# otherwise, we hit the "Unauthenticated users" limit
# https://docs.docker.com/docker-hub/usage/
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ inputs['docker-username'] }}
password: ${{ inputs['docker-password'] }}
- name: Generate OpenAPI Python client
shell: bash
run: |
@@ -40,3 +41,10 @@ runs:
--package-name onyx_openapi_client \
--skip-validate-spec \
--openapi-normalizer "SIMPLIFY_ONEOF_ANYOF=true,SET_OAS3_NULLABLE=true"
- name: Upload OpenAPI artifacts
uses: actions/upload-artifact@v4
with:
name: openapi-artifacts
path: backend/generated/

View File

@@ -1,17 +0,0 @@
name: "Setup Playwright"
description: "Sets up Playwright and system deps (assumes Python and Playwright are installed)"
runs:
using: "composite"
steps:
- name: Cache playwright cache
uses: runs-on/cache@50350ad4242587b6c8c2baa2e740b1bc11285ff4 # ratchet:runs-on/cache@v4
with:
path: ~/.cache/ms-playwright
key: ${{ runner.os }}-playwright-${{ hashFiles('backend/requirements/default.txt') }}
restore-keys: |
${{ runner.os }}-playwright-
- name: Install playwright
shell: bash
run: |
playwright install chromium --with-deps

View File

@@ -1,38 +0,0 @@
name: "Setup Python and Install Dependencies"
description: "Sets up Python with uv and installs deps"
runs:
using: "composite"
steps:
- name: Setup uv
uses: astral-sh/setup-uv@caf0cab7a618c569241d31dcd442f54681755d39 # ratchet:astral-sh/setup-uv@v3
# TODO: Enable caching once there is a uv.lock file checked in.
# with:
# enable-cache: true
- name: Cache uv cache directory
uses: runs-on/cache@50350ad4242587b6c8c2baa2e740b1bc11285ff4 # ratchet:runs-on/cache@v4
with:
path: ~/.cache/uv
key: ${{ runner.os }}-uv-${{ hashFiles('backend/requirements/*.txt', 'backend/pyproject.toml') }}
restore-keys: |
${{ runner.os }}-uv-
- name: Setup Python
uses: actions/setup-python@a26af69be951a213d495a4c3e4e4022e16d87065 # ratchet:actions/setup-python@v5
with:
python-version: "3.11"
- name: Create virtual environment
shell: bash
run: |
uv venv ${{ runner.temp }}/venv
echo "VENV_PATH=${{ runner.temp }}/venv" >> $GITHUB_ENV
echo "${{ runner.temp }}/venv/bin" >> $GITHUB_PATH
- name: Install Python dependencies with uv
shell: bash
run: |
uv pip install \
-r backend/requirements/default.txt \
-r backend/requirements/dev.txt \
-r backend/requirements/model_server.txt

View File

@@ -1,101 +0,0 @@
name: "Slack Notify on Failure"
description: "Sends a Slack notification when a workflow fails"
inputs:
webhook-url:
description: "Slack webhook URL (can also use SLACK_WEBHOOK_URL env var)"
required: false
failed-jobs:
description: "List of failed job names (newline-separated)"
required: false
title:
description: "Title for the notification"
required: false
default: "🚨 Workflow Failed"
ref-name:
description: "Git ref name (tag/branch)"
required: false
runs:
using: "composite"
steps:
- name: Send Slack notification
shell: bash
env:
SLACK_WEBHOOK_URL: ${{ inputs.webhook-url }}
run: |
if [ -z "$SLACK_WEBHOOK_URL" ]; then
echo "webhook-url input or SLACK_WEBHOOK_URL env var is not set, skipping notification"
exit 0
fi
# Get inputs with defaults
FAILED_JOBS="${{ inputs.failed-jobs }}"
TITLE="${{ inputs.title }}"
REF_NAME="${{ inputs.ref-name }}"
REPO="${{ github.repository }}"
WORKFLOW="${{ github.workflow }}"
RUN_NUMBER="${{ github.run_number }}"
RUN_ID="${{ github.run_id }}"
SERVER_URL="${{ github.server_url }}"
WORKFLOW_URL="${SERVER_URL}/${REPO}/actions/runs/${RUN_ID}"
# Use ref_name from input or fall back to github.ref_name
if [ -z "$REF_NAME" ]; then
REF_NAME="${{ github.ref_name }}"
fi
# Escape JSON special characters
escape_json() {
local input="$1"
# Escape backslashes first (but preserve \n sequences)
# Protect \n sequences temporarily
input=$(printf '%s' "$input" | sed 's/\\n/\x01NL\x01/g')
# Escape remaining backslashes
input=$(printf '%s' "$input" | sed 's/\\/\\\\/g')
# Restore \n sequences (single backslash, will be correct in JSON)
input=$(printf '%s' "$input" | sed 's/\x01NL\x01/\\n/g')
# Escape quotes
printf '%s' "$input" | sed 's/"/\\"/g'
}
REF_NAME_ESC=$(escape_json "$REF_NAME")
FAILED_JOBS_ESC=$(escape_json "$FAILED_JOBS")
WORKFLOW_URL_ESC=$(escape_json "$WORKFLOW_URL")
TITLE_ESC=$(escape_json "$TITLE")
# Build JSON payload piece by piece
# Note: FAILED_JOBS_ESC already contains \n sequences that should remain as \n in JSON
PAYLOAD="{"
PAYLOAD="${PAYLOAD}\"text\":\"${TITLE_ESC}\","
PAYLOAD="${PAYLOAD}\"blocks\":[{"
PAYLOAD="${PAYLOAD}\"type\":\"header\","
PAYLOAD="${PAYLOAD}\"text\":{\"type\":\"plain_text\",\"text\":\"${TITLE_ESC}\"}"
PAYLOAD="${PAYLOAD}},{"
PAYLOAD="${PAYLOAD}\"type\":\"section\","
PAYLOAD="${PAYLOAD}\"fields\":["
if [ -n "$REF_NAME" ]; then
PAYLOAD="${PAYLOAD}{\"type\":\"mrkdwn\",\"text\":\"*Ref:*\\n${REF_NAME_ESC}\"},"
fi
PAYLOAD="${PAYLOAD}{\"type\":\"mrkdwn\",\"text\":\"*Run ID:*\\n#${RUN_NUMBER}\"}"
PAYLOAD="${PAYLOAD}]"
PAYLOAD="${PAYLOAD}}"
if [ -n "$FAILED_JOBS" ]; then
PAYLOAD="${PAYLOAD},{"
PAYLOAD="${PAYLOAD}\"type\":\"section\","
PAYLOAD="${PAYLOAD}\"text\":{\"type\":\"mrkdwn\",\"text\":\"*Failed Jobs:*\\n${FAILED_JOBS_ESC}\"}"
PAYLOAD="${PAYLOAD}}"
fi
PAYLOAD="${PAYLOAD},{"
PAYLOAD="${PAYLOAD}\"type\":\"actions\","
PAYLOAD="${PAYLOAD}\"elements\":[{"
PAYLOAD="${PAYLOAD}\"type\":\"button\","
PAYLOAD="${PAYLOAD}\"text\":{\"type\":\"plain_text\",\"text\":\"View Workflow Run\"},"
PAYLOAD="${PAYLOAD}\"url\":\"${WORKFLOW_URL_ESC}\""
PAYLOAD="${PAYLOAD}}]"
PAYLOAD="${PAYLOAD}}"
PAYLOAD="${PAYLOAD}]"
PAYLOAD="${PAYLOAD}}"
curl -X POST -H 'Content-type: application/json' \
--data "$PAYLOAD" \
"$SLACK_WEBHOOK_URL"

View File

@@ -1,20 +0,0 @@
version: 2
updates:
- package-ecosystem: "github-actions"
directory: "/"
schedule:
interval: "weekly"
open-pull-requests-limit: 3
assignees:
- "jmelahman"
labels:
- "dependabot:actions"
- package-ecosystem: "pip"
directory: "/backend"
schedule:
interval: "weekly"
open-pull-requests-limit: 3
assignees:
- "jmelahman"
labels:
- "dependabot:python"

1
.github/runs-on.yml vendored
View File

@@ -1 +0,0 @@
_extend: .github-private

View File

@@ -1,7 +1,4 @@
name: Check Lazy Imports
concurrency:
group: Check-Lazy-Imports-${{ github.workflow }}-${{ github.head_ref || github.event.workflow_run.head_branch || github.run_id }}
cancel-in-progress: true
on:
merge_group:
@@ -16,12 +13,12 @@ jobs:
steps:
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Set up Python
uses: actions/setup-python@e797f83bcb11b83ae66e0230d6156d7c80228e7c # ratchet:actions/setup-python@v6
uses: actions/setup-python@v4
with:
python-version: '3.11'
- name: Check lazy imports
run: python3 backend/scripts/check_lazy_imports.py
run: python3 backend/scripts/check_lazy_imports.py

View File

@@ -1,934 +0,0 @@
name: Build and Push Docker Images on Tag
on:
push:
tags:
- "*"
workflow_dispatch:
env:
IS_DRY_RUN: ${{ github.event_name == 'workflow_dispatch' }}
EDGE_TAG: ${{ startsWith(github.ref_name, 'nightly-latest') }}
jobs:
# Determine which components to build based on the tag
determine-builds:
# NOTE: Github-hosted runners have about 20s faster queue times and are preferred here.
runs-on: ubuntu-slim
outputs:
build-web: ${{ steps.check.outputs.build-web }}
build-web-cloud: ${{ steps.check.outputs.build-web-cloud }}
build-backend: ${{ steps.check.outputs.build-backend }}
build-model-server: ${{ steps.check.outputs.build-model-server }}
is-cloud-tag: ${{ steps.check.outputs.is-cloud-tag }}
is-stable: ${{ steps.check.outputs.is-stable }}
is-beta: ${{ steps.check.outputs.is-beta }}
is-stable-standalone: ${{ steps.check.outputs.is-stable-standalone }}
is-beta-standalone: ${{ steps.check.outputs.is-beta-standalone }}
sanitized-tag: ${{ steps.check.outputs.sanitized-tag }}
steps:
- name: Check which components to build and version info
id: check
run: |
TAG="${{ github.ref_name }}"
# Sanitize tag name by replacing slashes with hyphens (for Docker tag compatibility)
SANITIZED_TAG=$(echo "$TAG" | tr '/' '-')
IS_CLOUD=false
BUILD_WEB=false
BUILD_WEB_CLOUD=false
BUILD_BACKEND=true
BUILD_MODEL_SERVER=true
IS_STABLE=false
IS_BETA=false
IS_STABLE_STANDALONE=false
IS_BETA_STANDALONE=false
if [[ "$TAG" == *cloud* ]]; then
IS_CLOUD=true
BUILD_WEB_CLOUD=true
else
BUILD_WEB=true
fi
# Version checks (for web - any stable version)
if [[ "$TAG" =~ ^v[0-9]+\.[0-9]+\.[0-9]+$ ]]; then
IS_STABLE=true
fi
if [[ "$TAG" =~ ^v[0-9]+\.[0-9]+\.[0-9]+-beta\.[0-9]+$ ]]; then
IS_BETA=true
fi
# Version checks (for backend/model-server - stable version excluding cloud tags)
if [[ "$TAG" =~ ^v[0-9]+\.[0-9]+\.[0-9]+$ ]] && [[ "$TAG" != *cloud* ]]; then
IS_STABLE_STANDALONE=true
fi
if [[ "$TAG" =~ ^v[0-9]+\.[0-9]+\.[0-9]+-beta\.[0-9]+$ ]] && [[ "$TAG" != *cloud* ]]; then
IS_BETA_STANDALONE=true
fi
{
echo "build-web=$BUILD_WEB"
echo "build-web-cloud=$BUILD_WEB_CLOUD"
echo "build-backend=$BUILD_BACKEND"
echo "build-model-server=$BUILD_MODEL_SERVER"
echo "is-cloud-tag=$IS_CLOUD"
echo "is-stable=$IS_STABLE"
echo "is-beta=$IS_BETA"
echo "is-stable-standalone=$IS_STABLE_STANDALONE"
echo "is-beta-standalone=$IS_BETA_STANDALONE"
echo "sanitized-tag=$SANITIZED_TAG"
} >> "$GITHUB_OUTPUT"
build-web-amd64:
needs: determine-builds
if: needs.determine-builds.outputs.build-web == 'true'
runs-on:
- runs-on
- runner=4cpu-linux-x64
- run-id=${{ github.run_id }}-web-amd64
- extras=ecr-cache
outputs:
digest: ${{ steps.build.outputs.digest }}
env:
REGISTRY_IMAGE: onyxdotapp/onyx-web-server
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
- name: Docker meta
id: meta
uses: docker/metadata-action@318604b99e75e41977312d83839a89be02ca4893 # ratchet:docker/metadata-action@v5
with:
images: ${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
flavor: |
latest=false
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Build and push AMD64
id: build
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
with:
context: ./web
file: ./web/Dockerfile
platforms: linux/amd64
labels: ${{ steps.meta.outputs.labels }}
build-args: |
ONYX_VERSION=${{ github.ref_name }}
NODE_OPTIONS=--max-old-space-size=8192
cache-from: |
type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:web-cache-amd64
cache-to: |
type=inline
type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:web-cache-amd64,mode=max
outputs: type=image,name=${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }},push-by-digest=true,name-canonical=true,push=true
build-web-arm64:
needs: determine-builds
if: needs.determine-builds.outputs.build-web == 'true'
runs-on:
- runs-on
- runner=4cpu-linux-arm64
- run-id=${{ github.run_id }}-web-arm64
- extras=ecr-cache
outputs:
digest: ${{ steps.build.outputs.digest }}
env:
REGISTRY_IMAGE: onyxdotapp/onyx-web-server
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
- name: Docker meta
id: meta
uses: docker/metadata-action@318604b99e75e41977312d83839a89be02ca4893 # ratchet:docker/metadata-action@v5
with:
images: ${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
flavor: |
latest=false
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Build and push ARM64
id: build
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
with:
context: ./web
file: ./web/Dockerfile
platforms: linux/arm64
labels: ${{ steps.meta.outputs.labels }}
build-args: |
ONYX_VERSION=${{ github.ref_name }}
NODE_OPTIONS=--max-old-space-size=8192
cache-from: |
type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:web-cache-arm64
cache-to: |
type=inline
type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:web-cache-arm64,mode=max
outputs: type=image,name=${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }},push-by-digest=true,name-canonical=true,push=true
merge-web:
needs:
- determine-builds
- build-web-amd64
- build-web-arm64
runs-on:
- runs-on
- runner=2cpu-linux-x64
- run-id=${{ github.run_id }}-merge-web
- extras=ecr-cache
env:
REGISTRY_IMAGE: onyxdotapp/onyx-web-server
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Docker meta
id: meta
uses: docker/metadata-action@318604b99e75e41977312d83839a89be02ca4893 # ratchet:docker/metadata-action@v5
with:
images: ${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
flavor: |
latest=false
tags: |
type=raw,value=${{ github.event_name == 'workflow_dispatch' && format('web-{0}', needs.determine-builds.outputs.sanitized-tag) || github.ref_name }}
type=raw,value=${{ github.event_name != 'workflow_dispatch' && needs.determine-builds.outputs.is-stable == 'true' && 'latest' || '' }}
type=raw,value=${{ github.event_name != 'workflow_dispatch' && env.EDGE_TAG == 'true' && 'edge' || '' }}
type=raw,value=${{ github.event_name != 'workflow_dispatch' && needs.determine-builds.outputs.is-beta == 'true' && 'beta' || '' }}
- name: Create and push manifest
run: |
IMAGES="${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}@${{ needs.build-web-amd64.outputs.digest }} ${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}@${{ needs.build-web-arm64.outputs.digest }}"
docker buildx imagetools create \
$(printf '%s\n' "${{ steps.meta.outputs.tags }}" | xargs -I {} echo -t {}) \
$IMAGES
build-web-cloud-amd64:
needs: determine-builds
if: needs.determine-builds.outputs.build-web-cloud == 'true'
runs-on:
- runs-on
- runner=4cpu-linux-x64
- run-id=${{ github.run_id }}-web-cloud-amd64
- extras=ecr-cache
outputs:
digest: ${{ steps.build.outputs.digest }}
env:
REGISTRY_IMAGE: onyxdotapp/onyx-web-server-cloud
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
- name: Docker meta
id: meta
uses: docker/metadata-action@318604b99e75e41977312d83839a89be02ca4893 # ratchet:docker/metadata-action@v5
with:
images: ${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
flavor: |
latest=false
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Build and push AMD64
id: build
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
with:
context: ./web
file: ./web/Dockerfile
platforms: linux/amd64
labels: ${{ steps.meta.outputs.labels }}
build-args: |
ONYX_VERSION=${{ 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=${{ secrets.SENTRY_DSN }}
NEXT_PUBLIC_STRIPE_PUBLISHABLE_KEY=${{ secrets.STRIPE_PUBLISHABLE_KEY }}
NEXT_PUBLIC_GTM_ENABLED=true
NEXT_PUBLIC_FORGOT_PASSWORD_ENABLED=true
NEXT_PUBLIC_INCLUDE_ERROR_POPUP_SUPPORT_LINK=true
NODE_OPTIONS=--max-old-space-size=8192
cache-from: |
type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:cloudweb-cache-amd64
cache-to: |
type=inline
type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:cloudweb-cache-amd64,mode=max
outputs: type=image,name=${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }},push-by-digest=true,name-canonical=true,push=true
build-web-cloud-arm64:
needs: determine-builds
if: needs.determine-builds.outputs.build-web-cloud == 'true'
runs-on:
- runs-on
- runner=4cpu-linux-arm64
- run-id=${{ github.run_id }}-web-cloud-arm64
- extras=ecr-cache
outputs:
digest: ${{ steps.build.outputs.digest }}
env:
REGISTRY_IMAGE: onyxdotapp/onyx-web-server-cloud
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
- name: Docker meta
id: meta
uses: docker/metadata-action@318604b99e75e41977312d83839a89be02ca4893 # ratchet:docker/metadata-action@v5
with:
images: ${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
flavor: |
latest=false
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Build and push ARM64
id: build
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
with:
context: ./web
file: ./web/Dockerfile
platforms: linux/arm64
labels: ${{ steps.meta.outputs.labels }}
build-args: |
ONYX_VERSION=${{ 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=${{ secrets.SENTRY_DSN }}
NEXT_PUBLIC_STRIPE_PUBLISHABLE_KEY=${{ secrets.STRIPE_PUBLISHABLE_KEY }}
NEXT_PUBLIC_GTM_ENABLED=true
NEXT_PUBLIC_FORGOT_PASSWORD_ENABLED=true
NEXT_PUBLIC_INCLUDE_ERROR_POPUP_SUPPORT_LINK=true
NODE_OPTIONS=--max-old-space-size=8192
cache-from: |
type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:cloudweb-cache-arm64
cache-to: |
type=inline
type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:cloudweb-cache-arm64,mode=max
outputs: type=image,name=${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }},push-by-digest=true,name-canonical=true,push=true
merge-web-cloud:
needs:
- determine-builds
- build-web-cloud-amd64
- build-web-cloud-arm64
runs-on:
- runs-on
- runner=2cpu-linux-x64
- run-id=${{ github.run_id }}-merge-web-cloud
- extras=ecr-cache
env:
REGISTRY_IMAGE: onyxdotapp/onyx-web-server-cloud
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Docker meta
id: meta
uses: docker/metadata-action@318604b99e75e41977312d83839a89be02ca4893 # ratchet:docker/metadata-action@v5
with:
images: ${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
flavor: |
latest=false
tags: |
type=raw,value=${{ github.event_name == 'workflow_dispatch' && format('web-cloud-{0}', needs.determine-builds.outputs.sanitized-tag) || github.ref_name }}
- name: Create and push manifest
run: |
IMAGES="${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}@${{ needs.build-web-cloud-amd64.outputs.digest }} ${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}@${{ needs.build-web-cloud-arm64.outputs.digest }}"
docker buildx imagetools create \
$(printf '%s\n' "${{ steps.meta.outputs.tags }}" | xargs -I {} echo -t {}) \
$IMAGES
build-backend-amd64:
needs: determine-builds
if: needs.determine-builds.outputs.build-backend == 'true'
runs-on:
- runs-on
- runner=2cpu-linux-x64
- run-id=${{ github.run_id }}-backend-amd64
- extras=ecr-cache
outputs:
digest: ${{ steps.build.outputs.digest }}
env:
REGISTRY_IMAGE: ${{ contains(github.ref_name, 'cloud') && 'onyxdotapp/onyx-backend-cloud' || 'onyxdotapp/onyx-backend' }}
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
- name: Docker meta
id: meta
uses: docker/metadata-action@318604b99e75e41977312d83839a89be02ca4893 # ratchet:docker/metadata-action@v5
with:
images: ${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
flavor: |
latest=false
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Build and push AMD64
id: build
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
with:
context: ./backend
file: ./backend/Dockerfile
platforms: linux/amd64
labels: ${{ steps.meta.outputs.labels }}
build-args: |
ONYX_VERSION=${{ github.ref_name }}
cache-from: |
type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:backend-cache-amd64
cache-to: |
type=inline
type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:backend-cache-amd64,mode=max
outputs: type=image,name=${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }},push-by-digest=true,name-canonical=true,push=true
build-backend-arm64:
needs: determine-builds
if: needs.determine-builds.outputs.build-backend == 'true'
runs-on:
- runs-on
- runner=2cpu-linux-arm64
- run-id=${{ github.run_id }}-backend-arm64
- extras=ecr-cache
outputs:
digest: ${{ steps.build.outputs.digest }}
env:
REGISTRY_IMAGE: ${{ contains(github.ref_name, 'cloud') && 'onyxdotapp/onyx-backend-cloud' || 'onyxdotapp/onyx-backend' }}
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
- name: Docker meta
id: meta
uses: docker/metadata-action@318604b99e75e41977312d83839a89be02ca4893 # ratchet:docker/metadata-action@v5
with:
images: ${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
flavor: |
latest=false
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Build and push ARM64
id: build
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
with:
context: ./backend
file: ./backend/Dockerfile
platforms: linux/arm64
labels: ${{ steps.meta.outputs.labels }}
build-args: |
ONYX_VERSION=${{ github.ref_name }}
cache-from: |
type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:backend-cache-arm64
cache-to: |
type=inline
type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:backend-cache-arm64,mode=max
outputs: type=image,name=${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }},push-by-digest=true,name-canonical=true,push=true
merge-backend:
needs:
- determine-builds
- build-backend-amd64
- build-backend-arm64
runs-on:
- runs-on
- runner=2cpu-linux-x64
- run-id=${{ github.run_id }}-merge-backend
- extras=ecr-cache
env:
REGISTRY_IMAGE: ${{ contains(github.ref_name, 'cloud') && 'onyxdotapp/onyx-backend-cloud' || 'onyxdotapp/onyx-backend' }}
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Docker meta
id: meta
uses: docker/metadata-action@318604b99e75e41977312d83839a89be02ca4893 # ratchet:docker/metadata-action@v5
with:
images: ${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
flavor: |
latest=false
tags: |
type=raw,value=${{ github.event_name == 'workflow_dispatch' && format('backend-{0}', needs.determine-builds.outputs.sanitized-tag) || github.ref_name }}
type=raw,value=${{ github.event_name != 'workflow_dispatch' && needs.determine-builds.outputs.is-stable-standalone == 'true' && 'latest' || '' }}
type=raw,value=${{ github.event_name != 'workflow_dispatch' && env.EDGE_TAG == 'true' && 'edge' || '' }}
type=raw,value=${{ github.event_name != 'workflow_dispatch' && needs.determine-builds.outputs.is-beta-standalone == 'true' && 'beta' || '' }}
- name: Create and push manifest
run: |
IMAGES="${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}@${{ needs.build-backend-amd64.outputs.digest }} ${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}@${{ needs.build-backend-arm64.outputs.digest }}"
docker buildx imagetools create \
$(printf '%s\n' "${{ steps.meta.outputs.tags }}" | xargs -I {} echo -t {}) \
$IMAGES
build-model-server-amd64:
needs: determine-builds
if: needs.determine-builds.outputs.build-model-server == 'true'
runs-on:
- runs-on
- runner=2cpu-linux-x64
- run-id=${{ github.run_id }}-model-server-amd64
- volume=40gb
- extras=ecr-cache
outputs:
digest: ${{ steps.build.outputs.digest }}
env:
REGISTRY_IMAGE: ${{ contains(github.ref_name, 'cloud') && 'onyxdotapp/onyx-model-server-cloud' || 'onyxdotapp/onyx-model-server' }}
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
- name: Docker meta
id: meta
uses: docker/metadata-action@318604b99e75e41977312d83839a89be02ca4893 # ratchet:docker/metadata-action@v5
with:
images: ${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
flavor: |
latest=false
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Build and push AMD64
id: build
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
with:
context: ./backend
file: ./backend/Dockerfile.model_server
platforms: linux/amd64
labels: ${{ steps.meta.outputs.labels }}
build-args: |
ONYX_VERSION=${{ github.ref_name }}
cache-from: |
type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:model-server-cache-amd64
cache-to: |
type=inline
type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:model-server-cache-amd64,mode=max
outputs: type=image,name=${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }},push-by-digest=true,name-canonical=true,push=true
build-model-server-arm64:
needs: determine-builds
if: needs.determine-builds.outputs.build-model-server == 'true'
runs-on:
- runs-on
- runner=2cpu-linux-arm64
- run-id=${{ github.run_id }}-model-server-arm64
- volume=40gb
- extras=ecr-cache
outputs:
digest: ${{ steps.build.outputs.digest }}
env:
REGISTRY_IMAGE: ${{ contains(github.ref_name, 'cloud') && 'onyxdotapp/onyx-model-server-cloud' || 'onyxdotapp/onyx-model-server' }}
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
- name: Docker meta
id: meta
uses: docker/metadata-action@318604b99e75e41977312d83839a89be02ca4893 # ratchet:docker/metadata-action@v5
with:
images: ${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
flavor: |
latest=false
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Build and push ARM64
id: build
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
with:
context: ./backend
file: ./backend/Dockerfile.model_server
platforms: linux/arm64
labels: ${{ steps.meta.outputs.labels }}
build-args: |
ONYX_VERSION=${{ github.ref_name }}
cache-from: |
type=registry,ref=${{ env.REGISTRY_IMAGE }}:latest
type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:model-server-cache-arm64
cache-to: |
type=inline
type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:model-server-cache-arm64,mode=max
outputs: type=image,name=${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }},push-by-digest=true,name-canonical=true,push=true
merge-model-server:
needs:
- determine-builds
- build-model-server-amd64
- build-model-server-arm64
runs-on:
- runs-on
- runner=2cpu-linux-x64
- run-id=${{ github.run_id }}-merge-model-server
- extras=ecr-cache
env:
REGISTRY_IMAGE: ${{ contains(github.ref_name, 'cloud') && 'onyxdotapp/onyx-model-server-cloud' || 'onyxdotapp/onyx-model-server' }}
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Docker meta
id: meta
uses: docker/metadata-action@318604b99e75e41977312d83839a89be02ca4893 # ratchet:docker/metadata-action@v5
with:
images: ${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
flavor: |
latest=false
tags: |
type=raw,value=${{ github.event_name == 'workflow_dispatch' && format('model-server-{0}', needs.determine-builds.outputs.sanitized-tag) || github.ref_name }}
type=raw,value=${{ github.event_name != 'workflow_dispatch' && needs.determine-builds.outputs.is-stable-standalone == 'true' && 'latest' || '' }}
type=raw,value=${{ github.event_name != 'workflow_dispatch' && env.EDGE_TAG == 'true' && 'edge' || '' }}
type=raw,value=${{ github.event_name != 'workflow_dispatch' && needs.determine-builds.outputs.is-beta-standalone == 'true' && 'beta' || '' }}
- name: Create and push manifest
run: |
IMAGES="${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}@${{ needs.build-model-server-amd64.outputs.digest }} ${{ github.event_name == 'workflow_dispatch' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}@${{ needs.build-model-server-arm64.outputs.digest }}"
docker buildx imagetools create \
$(printf '%s\n' "${{ steps.meta.outputs.tags }}" | xargs -I {} echo -t {}) \
$IMAGES
trivy-scan-web:
needs:
- determine-builds
- merge-web
if: needs.merge-web.result == 'success'
runs-on:
- runs-on
- runner=2cpu-linux-arm64
- run-id=${{ github.run_id }}-trivy-scan-web
- extras=ecr-cache
env:
REGISTRY_IMAGE: onyxdotapp/onyx-web-server
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Run Trivy vulnerability scanner
uses: nick-fields/retry@ce71cc2ab81d554ebbe88c79ab5975992d79ba08 # ratchet:nick-fields/retry@v3
with:
timeout_minutes: 30
max_attempts: 3
retry_wait_seconds: 10
command: |
if [ "${{ github.event_name }}" == "workflow_dispatch" ]; then
SCAN_IMAGE="${{ env.RUNS_ON_ECR_CACHE }}:web-${{ needs.determine-builds.outputs.sanitized-tag }}"
else
SCAN_IMAGE="docker.io/${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}"
fi
docker run --rm -v $HOME/.cache/trivy:/root/.cache/trivy \
-e TRIVY_DB_REPOSITORY="public.ecr.aws/aquasecurity/trivy-db:2" \
-e TRIVY_JAVA_DB_REPOSITORY="public.ecr.aws/aquasecurity/trivy-java-db:1" \
-e TRIVY_USERNAME="${{ secrets.DOCKER_USERNAME }}" \
-e TRIVY_PASSWORD="${{ secrets.DOCKER_TOKEN }}" \
aquasec/trivy@sha256:a22415a38938a56c379387a8163fcb0ce38b10ace73e593475d3658d578b2436 \
image \
--skip-version-check \
--timeout 20m \
--severity CRITICAL,HIGH \
${SCAN_IMAGE}
trivy-scan-web-cloud:
needs:
- determine-builds
- merge-web-cloud
if: needs.merge-web-cloud.result == 'success'
runs-on:
- runs-on
- runner=2cpu-linux-arm64
- run-id=${{ github.run_id }}-trivy-scan-web-cloud
- extras=ecr-cache
env:
REGISTRY_IMAGE: onyxdotapp/onyx-web-server-cloud
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Run Trivy vulnerability scanner
uses: nick-fields/retry@ce71cc2ab81d554ebbe88c79ab5975992d79ba08 # ratchet:nick-fields/retry@v3
with:
timeout_minutes: 30
max_attempts: 3
retry_wait_seconds: 10
command: |
if [ "${{ github.event_name }}" == "workflow_dispatch" ]; then
SCAN_IMAGE="${{ env.RUNS_ON_ECR_CACHE }}:web-cloud-${{ needs.determine-builds.outputs.sanitized-tag }}"
else
SCAN_IMAGE="docker.io/${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}"
fi
docker run --rm -v $HOME/.cache/trivy:/root/.cache/trivy \
-e TRIVY_DB_REPOSITORY="public.ecr.aws/aquasecurity/trivy-db:2" \
-e TRIVY_JAVA_DB_REPOSITORY="public.ecr.aws/aquasecurity/trivy-java-db:1" \
-e TRIVY_USERNAME="${{ secrets.DOCKER_USERNAME }}" \
-e TRIVY_PASSWORD="${{ secrets.DOCKER_TOKEN }}" \
aquasec/trivy@sha256:a22415a38938a56c379387a8163fcb0ce38b10ace73e593475d3658d578b2436 \
image \
--skip-version-check \
--timeout 20m \
--severity CRITICAL,HIGH \
${SCAN_IMAGE}
trivy-scan-backend:
needs:
- determine-builds
- merge-backend
if: needs.merge-backend.result == 'success'
runs-on:
- runs-on
- runner=2cpu-linux-arm64
- run-id=${{ github.run_id }}-trivy-scan-backend
- extras=ecr-cache
env:
REGISTRY_IMAGE: ${{ contains(github.ref_name, 'cloud') && 'onyxdotapp/onyx-backend-cloud' || 'onyxdotapp/onyx-backend' }}
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
- name: Run Trivy vulnerability scanner
uses: nick-fields/retry@ce71cc2ab81d554ebbe88c79ab5975992d79ba08 # ratchet:nick-fields/retry@v3
with:
timeout_minutes: 30
max_attempts: 3
retry_wait_seconds: 10
command: |
if [ "${{ github.event_name }}" == "workflow_dispatch" ]; then
SCAN_IMAGE="${{ env.RUNS_ON_ECR_CACHE }}:backend-${{ needs.determine-builds.outputs.sanitized-tag }}"
else
SCAN_IMAGE="docker.io/${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}"
fi
docker run --rm -v $HOME/.cache/trivy:/root/.cache/trivy \
-v ${{ github.workspace }}/backend/.trivyignore:/tmp/.trivyignore:ro \
-e TRIVY_DB_REPOSITORY="public.ecr.aws/aquasecurity/trivy-db:2" \
-e TRIVY_JAVA_DB_REPOSITORY="public.ecr.aws/aquasecurity/trivy-java-db:1" \
-e TRIVY_USERNAME="${{ secrets.DOCKER_USERNAME }}" \
-e TRIVY_PASSWORD="${{ secrets.DOCKER_TOKEN }}" \
aquasec/trivy@sha256:a22415a38938a56c379387a8163fcb0ce38b10ace73e593475d3658d578b2436 \
image \
--skip-version-check \
--timeout 20m \
--severity CRITICAL,HIGH \
--ignorefile /tmp/.trivyignore \
${SCAN_IMAGE}
trivy-scan-model-server:
needs:
- determine-builds
- merge-model-server
if: needs.merge-model-server.result == 'success'
runs-on:
- runs-on
- runner=2cpu-linux-arm64
- run-id=${{ github.run_id }}-trivy-scan-model-server
- extras=ecr-cache
env:
REGISTRY_IMAGE: ${{ contains(github.ref_name, 'cloud') && 'onyxdotapp/onyx-model-server-cloud' || 'onyxdotapp/onyx-model-server' }}
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Run Trivy vulnerability scanner
uses: nick-fields/retry@ce71cc2ab81d554ebbe88c79ab5975992d79ba08 # ratchet:nick-fields/retry@v3
with:
timeout_minutes: 30
max_attempts: 3
retry_wait_seconds: 10
command: |
if [ "${{ github.event_name }}" == "workflow_dispatch" ]; then
SCAN_IMAGE="${{ env.RUNS_ON_ECR_CACHE }}:model-server-${{ needs.determine-builds.outputs.sanitized-tag }}"
else
SCAN_IMAGE="docker.io/${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}"
fi
docker run --rm -v $HOME/.cache/trivy:/root/.cache/trivy \
-e TRIVY_DB_REPOSITORY="public.ecr.aws/aquasecurity/trivy-db:2" \
-e TRIVY_JAVA_DB_REPOSITORY="public.ecr.aws/aquasecurity/trivy-java-db:1" \
-e TRIVY_USERNAME="${{ secrets.DOCKER_USERNAME }}" \
-e TRIVY_PASSWORD="${{ secrets.DOCKER_TOKEN }}" \
aquasec/trivy@sha256:a22415a38938a56c379387a8163fcb0ce38b10ace73e593475d3658d578b2436 \
image \
--skip-version-check \
--timeout 20m \
--severity CRITICAL,HIGH \
${SCAN_IMAGE}
notify-slack-on-failure:
needs:
- build-web-amd64
- build-web-arm64
- merge-web
- build-web-cloud-amd64
- build-web-cloud-arm64
- merge-web-cloud
- build-backend-amd64
- build-backend-arm64
- merge-backend
- build-model-server-amd64
- build-model-server-arm64
- merge-model-server
if: always() && (needs.build-web-amd64.result == 'failure' || needs.build-web-arm64.result == 'failure' || needs.merge-web.result == 'failure' || needs.build-web-cloud-amd64.result == 'failure' || needs.build-web-cloud-arm64.result == 'failure' || needs.merge-web-cloud.result == 'failure' || needs.build-backend-amd64.result == 'failure' || needs.build-backend-arm64.result == 'failure' || needs.merge-backend.result == 'failure' || needs.build-model-server-amd64.result == 'failure' || needs.build-model-server-arm64.result == 'failure' || needs.merge-model-server.result == 'failure') && github.event_name != 'workflow_dispatch'
# NOTE: Github-hosted runners have about 20s faster queue times and are preferred here.
runs-on: ubuntu-slim
steps:
- name: Checkout
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
- name: Determine failed jobs
id: failed-jobs
shell: bash
run: |
FAILED_JOBS=""
if [ "${{ needs.build-web-amd64.result }}" == "failure" ]; then
FAILED_JOBS="${FAILED_JOBS}• build-web-amd64\\n"
fi
if [ "${{ needs.build-web-arm64.result }}" == "failure" ]; then
FAILED_JOBS="${FAILED_JOBS}• build-web-arm64\\n"
fi
if [ "${{ needs.merge-web.result }}" == "failure" ]; then
FAILED_JOBS="${FAILED_JOBS}• merge-web\\n"
fi
if [ "${{ needs.build-web-cloud-amd64.result }}" == "failure" ]; then
FAILED_JOBS="${FAILED_JOBS}• build-web-cloud-amd64\\n"
fi
if [ "${{ needs.build-web-cloud-arm64.result }}" == "failure" ]; then
FAILED_JOBS="${FAILED_JOBS}• build-web-cloud-arm64\\n"
fi
if [ "${{ needs.merge-web-cloud.result }}" == "failure" ]; then
FAILED_JOBS="${FAILED_JOBS}• merge-web-cloud\\n"
fi
if [ "${{ needs.build-backend-amd64.result }}" == "failure" ]; then
FAILED_JOBS="${FAILED_JOBS}• build-backend-amd64\\n"
fi
if [ "${{ needs.build-backend-arm64.result }}" == "failure" ]; then
FAILED_JOBS="${FAILED_JOBS}• build-backend-arm64\\n"
fi
if [ "${{ needs.merge-backend.result }}" == "failure" ]; then
FAILED_JOBS="${FAILED_JOBS}• merge-backend\\n"
fi
if [ "${{ needs.build-model-server-amd64.result }}" == "failure" ]; then
FAILED_JOBS="${FAILED_JOBS}• build-model-server-amd64\\n"
fi
if [ "${{ needs.build-model-server-arm64.result }}" == "failure" ]; then
FAILED_JOBS="${FAILED_JOBS}• build-model-server-arm64\\n"
fi
if [ "${{ needs.merge-model-server.result }}" == "failure" ]; then
FAILED_JOBS="${FAILED_JOBS}• merge-model-server\\n"
fi
# Remove trailing \n and set output
FAILED_JOBS=$(printf '%s' "$FAILED_JOBS" | sed 's/\\n$//')
echo "jobs=$FAILED_JOBS" >> "$GITHUB_OUTPUT"
- name: Send Slack notification
uses: ./.github/actions/slack-notify
with:
webhook-url: ${{ secrets.MONITOR_DEPLOYMENTS_WEBHOOK }}
failed-jobs: ${{ steps.failed-jobs.outputs.jobs }}
title: "🚨 Deployment Workflow Failed"
ref-name: ${{ github.ref_name }}

View File

@@ -0,0 +1,186 @@
name: Build and Push Backend Image on Tag
on:
push:
tags:
- "*"
env:
REGISTRY_IMAGE: ${{ contains(github.ref_name, 'cloud') && 'onyxdotapp/onyx-backend-cloud' || 'onyxdotapp/onyx-backend' }}
DEPLOYMENT: ${{ contains(github.ref_name, 'cloud') && 'cloud' || 'standalone' }}
# tag nightly builds with "edge"
EDGE_TAG: ${{ startsWith(github.ref_name, 'nightly-latest') }}
jobs:
build-and-push:
# TODO: investigate a matrix build like the web container
# See https://runs-on.com/runners/linux/
runs-on:
- runs-on
- runner=${{ 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//\//-}" >> $GITHUB_ENV
- name: Check if stable release version
id: check_version
run: |
if [[ "${{ github.ref_name }}" =~ ^v[0-9]+\.[0-9]+\.[0-9]+$ ]] && [[ "${{ github.ref_name }}" != *"cloud"* ]]; then
echo "is_stable=true" >> $GITHUB_OUTPUT
else
echo "is_stable=false" >> $GITHUB_OUTPUT
fi
- name: Checkout code
uses: actions/checkout@v4
- name: Docker meta
id: meta
uses: docker/metadata-action@v5
with:
images: ${{ env.REGISTRY_IMAGE }}
flavor: |
latest=false
tags: |
type=raw,value=${{ github.ref_name }}
type=raw,value=${{ steps.check_version.outputs.is_stable == 'true' && 'latest' || '' }}
type=raw,value=${{ env.EDGE_TAG == 'true' && 'edge' || '' }}
- 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: Install build-essential
run: |
sudo apt-get update
sudo apt-get install -y build-essential
- name: Backend Image Docker Build and Push
id: build
uses: docker/build-push-action@v6
with:
context: ./backend
file: ./backend/Dockerfile
platforms: ${{ matrix.platform }}
push: true
build-args: |
ONYX_VERSION=${{ github.ref_name }}
labels: ${{ steps.meta.outputs.labels }}
outputs: type=image,name=${{ env.REGISTRY_IMAGE }},push-by-digest=true,name-canonical=true,push=true
cache-from: type=s3,prefix=cache/${{ github.repository }}/${{ env.DEPLOYMENT }}/backend-${{ env.PLATFORM_PAIR }}/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }}
cache-to: type=s3,prefix=cache/${{ github.repository }}/${{ env.DEPLOYMENT }}/backend-${{ env.PLATFORM_PAIR }}/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }},mode=max
- 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: backend-digests-${{ env.PLATFORM_PAIR }}-${{ github.run_id }}
path: /tmp/digests/*
if-no-files-found: error
retention-days: 1
merge:
runs-on: ubuntu-latest
needs:
- build-and-push
steps:
# Needed for trivyignore
- name: Checkout
uses: actions/checkout@v4
- name: Check if stable release version
id: check_version
run: |
if [[ "${{ github.ref_name }}" =~ ^v[0-9]+\.[0-9]+\.[0-9]+$ ]] && [[ "${{ github.ref_name }}" != *"cloud"* ]]; then
echo "is_stable=true" >> $GITHUB_OUTPUT
else
echo "is_stable=false" >> $GITHUB_OUTPUT
fi
- name: Download digests
uses: actions/download-artifact@v4
with:
path: /tmp/digests
pattern: backend-digests-*-${{ github.run_id }}
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: ${{ env.REGISTRY_IMAGE }}
flavor: |
latest=false
tags: |
type=raw,value=${{ github.ref_name }}
type=raw,value=${{ steps.check_version.outputs.is_stable == 'true' && 'latest' || '' }}
type=raw,value=${{ env.EDGE_TAG == 'true' && 'edge' || '' }}
- 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
# Security: Using pinned digest (0.65.0@sha256:a22415a38938a56c379387a8163fcb0ce38b10ace73e593475d3658d578b2436)
# Security: No Docker socket mount needed for remote registry scanning
- name: Run Trivy vulnerability scanner
uses: nick-fields/retry@v3
with:
timeout_minutes: 30
max_attempts: 3
retry_wait_seconds: 10
command: |
docker run --rm -v $HOME/.cache/trivy:/root/.cache/trivy \
-v ${{ github.workspace }}/backend/.trivyignore:/tmp/.trivyignore:ro \
-e TRIVY_DB_REPOSITORY="public.ecr.aws/aquasecurity/trivy-db:2" \
-e TRIVY_JAVA_DB_REPOSITORY="public.ecr.aws/aquasecurity/trivy-java-db:1" \
-e TRIVY_USERNAME="${{ secrets.DOCKER_USERNAME }}" \
-e TRIVY_PASSWORD="${{ secrets.DOCKER_TOKEN }}" \
aquasec/trivy@sha256:a22415a38938a56c379387a8163fcb0ce38b10ace73e593475d3658d578b2436 \
image \
--skip-version-check \
--timeout 20m \
--severity CRITICAL,HIGH \
--ignorefile /tmp/.trivyignore \
docker.io/${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}

View File

@@ -0,0 +1,158 @@
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:
- "*cloud*"
env:
REGISTRY_IMAGE: onyxdotapp/onyx-web-server-cloud
DEPLOYMENT: cloud
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: ${{ env.REGISTRY_IMAGE }}
flavor: |
latest=false
tags: |
type=raw,value=${{ github.ref_name }}
- 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@v6
with:
context: ./web
file: ./web/Dockerfile
platforms: ${{ matrix.platform }}
push: true
build-args: |
ONYX_VERSION=${{ 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=${{ secrets.SENTRY_DSN }}
NEXT_PUBLIC_STRIPE_PUBLISHABLE_KEY=${{ secrets.STRIPE_PUBLISHABLE_KEY }}
NEXT_PUBLIC_GTM_ENABLED=true
NEXT_PUBLIC_FORGOT_PASSWORD_ENABLED=true
NEXT_PUBLIC_INCLUDE_ERROR_POPUP_SUPPORT_LINK=true
NODE_OPTIONS=--max-old-space-size=8192
labels: ${{ steps.meta.outputs.labels }}
outputs: type=image,name=${{ env.REGISTRY_IMAGE }},push-by-digest=true,name-canonical=true,push=true
cache-from: type=s3,prefix=cache/${{ github.repository }}/${{ env.DEPLOYMENT }}/cloudweb-${{ env.PLATFORM_PAIR }}/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }}
cache-to: type=s3,prefix=cache/${{ github.repository }}/${{ env.DEPLOYMENT }}/cloudweb-${{ env.PLATFORM_PAIR }}/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }},mode=max
# no-cache needed due to weird interactions with the builds for different platforms
# NOTE(rkuo): this may not be true any more with the proper cache prefixing by architecture - currently testing with it off
- 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: cloudweb-digests-${{ env.PLATFORM_PAIR }}-${{ github.run_id }}
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: cloudweb-digests-*-${{ github.run_id }}
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: ${{ env.REGISTRY_IMAGE }}
flavor: |
latest=false
tags: |
type=raw,value=${{ github.ref_name }}
- 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: nick-fields/retry@v3
with:
timeout_minutes: 30
max_attempts: 3
retry_wait_seconds: 10
command: |
docker run --rm -v $HOME/.cache/trivy:/root/.cache/trivy \
-e TRIVY_DB_REPOSITORY="public.ecr.aws/aquasecurity/trivy-db:2" \
-e TRIVY_JAVA_DB_REPOSITORY="public.ecr.aws/aquasecurity/trivy-java-db:1" \
-e TRIVY_USERNAME="${{ secrets.DOCKER_USERNAME }}" \
-e TRIVY_PASSWORD="${{ secrets.DOCKER_TOKEN }}" \
aquasec/trivy@sha256:a22415a38938a56c379387a8163fcb0ce38b10ace73e593475d3658d578b2436 \
image \
--skip-version-check \
--timeout 20m \
--severity CRITICAL,HIGH \
docker.io/${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}

View File

@@ -0,0 +1,197 @@
name: Build and Push Model Server Image on Tag
on:
push:
tags:
- "*"
env:
REGISTRY_IMAGE: ${{ contains(github.ref_name, 'cloud') && 'onyxdotapp/onyx-model-server-cloud' || 'onyxdotapp/onyx-model-server' }}
DOCKER_BUILDKIT: 1
BUILDKIT_PROGRESS: plain
DEPLOYMENT: ${{ contains(github.ref_name, 'cloud') && 'cloud' || 'standalone' }}
# tag nightly builds with "edge"
EDGE_TAG: ${{ startsWith(github.ref_name, 'nightly-latest') }}
jobs:
# Bypassing this for now as the idea of not building is glitching
# releases and builds that depends on everything being tagged in docker
# 1) Preliminary job to check if the changed files are relevant
# check_model_server_changes:
# runs-on: ubuntu-latest
# outputs:
# changed: ${{ steps.check.outputs.changed }}
# steps:
# - name: Checkout code
# uses: actions/checkout@v4
#
# - name: Check if relevant files changed
# id: check
# run: |
# # Default to "false"
# echo "changed=false" >> $GITHUB_OUTPUT
#
# # Compare the previous commit (github.event.before) to the current one (github.sha)
# # If any file in backend/model_server/** or backend/Dockerfile.model_server is changed,
# # set changed=true
# if git diff --name-only ${{ github.event.before }} ${{ github.sha }} \
# | grep -E '^backend/model_server/|^backend/Dockerfile.model_server'; then
# echo "changed=true" >> $GITHUB_OUTPUT
# fi
check_model_server_changes:
runs-on: ubuntu-latest
outputs:
changed: "true"
steps:
- name: Bypass check and set output
run: echo "changed=true" >> $GITHUB_OUTPUT
build-amd64:
needs: [check_model_server_changes]
if: needs.check_model_server_changes.outputs.changed == 'true'
runs-on:
[runs-on, runner=8cpu-linux-x64, "run-id=${{ github.run_id }}-amd64"]
env:
PLATFORM_PAIR: linux-amd64
steps:
- name: Checkout code
uses: actions/checkout@v4
- name: System Info
run: |
df -h
free -h
docker system prune -af --volumes
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@v3
with:
driver-opts: |
image=moby/buildkit:latest
network=host
- name: Login to Docker Hub
uses: docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Build and Push AMD64
uses: docker/build-push-action@v6
with:
context: ./backend
file: ./backend/Dockerfile.model_server
platforms: linux/amd64
push: true
tags: ${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}-amd64
build-args: |
ONYX_VERSION=${{ github.ref_name }}
outputs: type=registry
provenance: false
cache-from: type=s3,prefix=cache/${{ github.repository }}/${{ env.DEPLOYMENT }}/model-server-${{ env.PLATFORM_PAIR }}/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }}
cache-to: type=s3,prefix=cache/${{ github.repository }}/${{ env.DEPLOYMENT }}/model-server-${{ env.PLATFORM_PAIR }}/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }},mode=max
# no-cache: true
build-arm64:
needs: [check_model_server_changes]
if: needs.check_model_server_changes.outputs.changed == 'true'
runs-on:
[runs-on, runner=8cpu-linux-arm64, "run-id=${{ github.run_id }}-arm64"]
env:
PLATFORM_PAIR: linux-arm64
steps:
- name: Checkout code
uses: actions/checkout@v4
- name: System Info
run: |
df -h
free -h
docker system prune -af --volumes
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@v3
with:
driver-opts: |
image=moby/buildkit:latest
network=host
- name: Login to Docker Hub
uses: docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Build and Push ARM64
uses: docker/build-push-action@v6
with:
context: ./backend
file: ./backend/Dockerfile.model_server
platforms: linux/arm64
push: true
tags: ${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}-arm64
build-args: |
ONYX_VERSION=${{ github.ref_name }}
outputs: type=registry
provenance: false
cache-from: type=s3,prefix=cache/${{ github.repository }}/${{ env.DEPLOYMENT }}/model-server-${{ env.PLATFORM_PAIR }}/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }}
cache-to: type=s3,prefix=cache/${{ github.repository }}/${{ env.DEPLOYMENT }}/model-server-${{ env.PLATFORM_PAIR }}/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }},mode=max
merge-and-scan:
needs: [build-amd64, build-arm64, check_model_server_changes]
if: needs.check_model_server_changes.outputs.changed == 'true'
runs-on: ubuntu-latest
steps:
- name: Check if stable release version
id: check_version
run: |
if [[ "${{ github.ref_name }}" =~ ^v[0-9]+\.[0-9]+\.[0-9]+$ ]] && [[ "${{ github.ref_name }}" != *"cloud"* ]]; then
echo "is_stable=true" >> $GITHUB_OUTPUT
else
echo "is_stable=false" >> $GITHUB_OUTPUT
fi
- name: Login to Docker Hub
uses: docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Create and Push Multi-arch Manifest
run: |
docker buildx create --use
docker buildx imagetools create -t ${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }} \
${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}-amd64 \
${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}-arm64
if [[ "${{ steps.check_version.outputs.is_stable }}" == "true" ]]; then
docker buildx imagetools create -t ${{ env.REGISTRY_IMAGE }}:latest \
${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}-amd64 \
${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}-arm64
fi
if [[ "${{ env.EDGE_TAG }}" == "true" ]]; then
docker buildx imagetools create -t ${{ env.REGISTRY_IMAGE }}:edge \
${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}-amd64 \
${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}-arm64
fi
- name: Run Trivy vulnerability scanner
uses: nick-fields/retry@v3
with:
timeout_minutes: 30
max_attempts: 3
retry_wait_seconds: 10
command: |
docker run --rm -v $HOME/.cache/trivy:/root/.cache/trivy \
-e TRIVY_DB_REPOSITORY="public.ecr.aws/aquasecurity/trivy-db:2" \
-e TRIVY_JAVA_DB_REPOSITORY="public.ecr.aws/aquasecurity/trivy-java-db:1" \
-e TRIVY_USERNAME="${{ secrets.DOCKER_USERNAME }}" \
-e TRIVY_PASSWORD="${{ secrets.DOCKER_TOKEN }}" \
aquasec/trivy@sha256:a22415a38938a56c379387a8163fcb0ce38b10ace73e593475d3658d578b2436 \
image \
--skip-version-check \
--timeout 20m \
--severity CRITICAL,HIGH \
docker.io/${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}

View File

@@ -0,0 +1,192 @@
name: Build and Push Web Image on Tag
on:
push:
tags:
- "*"
env:
REGISTRY_IMAGE: onyxdotapp/onyx-web-server
# tag nightly builds with "edge"
EDGE_TAG: ${{ startsWith(github.ref_name, 'nightly-latest') }}
DEPLOYMENT: standalone
jobs:
precheck:
runs-on: [runs-on, runner=2cpu-linux-x64, "run-id=${{ github.run_id }}"]
outputs:
should-run: ${{ steps.set-output.outputs.should-run }}
steps:
- name: Check if tag contains "cloud"
id: set-output
run: |
if [[ "${{ github.ref_name }}" == *cloud* ]]; then
echo "should-run=false" >> "$GITHUB_OUTPUT"
else
echo "should-run=true" >> "$GITHUB_OUTPUT"
fi
build:
needs: precheck
if: needs.precheck.outputs.should-run == 'true'
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//\//-}" >> $GITHUB_ENV
- name: Check if stable release version
id: check_version
run: |
if [[ "${{ github.ref_name }}" =~ ^v[0-9]+\.[0-9]+\.[0-9]+$ ]]; then
echo "is_stable=true" >> $GITHUB_OUTPUT
else
echo "is_stable=false" >> $GITHUB_OUTPUT
fi
- name: Checkout
uses: actions/checkout@v4
- name: Docker meta
id: meta
uses: docker/metadata-action@v5
with:
images: ${{ env.REGISTRY_IMAGE }}
flavor: |
latest=false
tags: |
type=raw,value=${{ github.ref_name }}
type=raw,value=${{ steps.check_version.outputs.is_stable == 'true' && 'latest' || '' }}
type=raw,value=${{ env.EDGE_TAG == 'true' && 'edge' || '' }}
- 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@v6
with:
context: ./web
file: ./web/Dockerfile
platforms: ${{ matrix.platform }}
push: true
build-args: |
ONYX_VERSION=${{ github.ref_name }}
NODE_OPTIONS=--max-old-space-size=8192
labels: ${{ steps.meta.outputs.labels }}
outputs: type=image,name=${{ env.REGISTRY_IMAGE }},push-by-digest=true,name-canonical=true,push=true
cache-from: type=s3,prefix=cache/${{ github.repository }}/${{ env.DEPLOYMENT }}/web-${{ env.PLATFORM_PAIR }}/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }}
cache-to: type=s3,prefix=cache/${{ github.repository }}/${{ env.DEPLOYMENT }}/web-${{ env.PLATFORM_PAIR }}/,region=${{ env.RUNS_ON_AWS_REGION }},bucket=${{ env.RUNS_ON_S3_BUCKET_CACHE }},mode=max
# no-cache needed due to weird interactions with the builds for different platforms
# NOTE(rkuo): this may not be true any more with the proper cache prefixing by architecture - currently testing with it off
- 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: web-digests-${{ env.PLATFORM_PAIR }}-${{ github.run_id }}
path: /tmp/digests/*
if-no-files-found: error
retention-days: 1
merge:
needs:
- build
if: needs.precheck.outputs.should-run == 'true'
runs-on: ubuntu-latest
steps:
- name: Check if stable release version
id: check_version
run: |
if [[ "${{ github.ref_name }}" =~ ^v[0-9]+\.[0-9]+\.[0-9]+$ ]] && [[ "${{ github.ref_name }}" != *"cloud"* ]]; then
echo "is_stable=true" >> $GITHUB_OUTPUT
else
echo "is_stable=false" >> $GITHUB_OUTPUT
fi
- name: Download digests
uses: actions/download-artifact@v4
with:
path: /tmp/digests
pattern: web-digests-*-${{ github.run_id }}
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: ${{ env.REGISTRY_IMAGE }}
flavor: |
latest=false
tags: |
type=raw,value=${{ github.ref_name }}
type=raw,value=${{ steps.check_version.outputs.is_stable == 'true' && 'latest' || '' }}
type=raw,value=${{ env.EDGE_TAG == 'true' && 'edge' || '' }}
- 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: nick-fields/retry@v3
with:
timeout_minutes: 30
max_attempts: 3
retry_wait_seconds: 10
command: |
docker run --rm -v $HOME/.cache/trivy:/root/.cache/trivy \
-e TRIVY_DB_REPOSITORY="public.ecr.aws/aquasecurity/trivy-db:2" \
-e TRIVY_JAVA_DB_REPOSITORY="public.ecr.aws/aquasecurity/trivy-java-db:1" \
-e TRIVY_USERNAME="${{ secrets.DOCKER_USERNAME }}" \
-e TRIVY_PASSWORD="${{ secrets.DOCKER_TOKEN }}" \
aquasec/trivy@sha256:a22415a38938a56c379387a8163fcb0ce38b10ace73e593475d3658d578b2436 \
image \
--skip-version-check \
--timeout 20m \
--severity CRITICAL,HIGH \
docker.io/${{ env.REGISTRY_IMAGE }}:${{ github.ref_name }}

View File

@@ -1,41 +0,0 @@
# 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 "beta".
name: Tag Beta Version
on:
workflow_dispatch:
inputs:
version:
description: "The version (ie v1.0.0-beta.0) to tag as beta"
required: true
jobs:
tag:
# 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 }}-tag"]
steps:
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Enable Docker CLI experimental features
run: echo "DOCKER_CLI_EXPERIMENTAL=enabled" >> $GITHUB_ENV
- name: Pull, Tag and Push Web Server Image
run: |
docker buildx imagetools create -t onyxdotapp/onyx-web-server:beta onyxdotapp/onyx-web-server:${{ github.event.inputs.version }}
- name: Pull, Tag and Push API Server Image
run: |
docker buildx imagetools create -t onyxdotapp/onyx-backend:beta onyxdotapp/onyx-backend:${{ github.event.inputs.version }}
- name: Pull, Tag and Push Model Server Image
run: |
docker buildx imagetools create -t onyxdotapp/onyx-model-server:beta onyxdotapp/onyx-model-server:${{ github.event.inputs.version }}

View File

@@ -14,13 +14,13 @@ jobs:
tag:
# 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=${{ github.run_id }}-tag"]
runs-on: [runs-on, runner=2cpu-linux-x64, "run-id=${{ github.run_id }}"]
steps:
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
uses: docker/setup-buildx-action@v1
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
uses: docker/login-action@v1
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}

View File

@@ -14,12 +14,12 @@ jobs:
runs-on: ubuntu-latest
steps:
- name: Checkout
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
with:
fetch-depth: 0
- name: Install Helm CLI
uses: azure/setup-helm@1a275c3b69536ee54be43f2070a358922e12c8d4 # ratchet:azure/setup-helm@v4
uses: azure/setup-helm@v4
with:
version: v3.12.1
@@ -43,7 +43,7 @@ jobs:
done
- name: Publish Helm charts to gh-pages
uses: stefanprodan/helm-gh-pages@0ad2bb377311d61ac04ad9eb6f252fb68e207260 # ratchet:stefanprodan/helm-gh-pages@v1.7.0
uses: stefanprodan/helm-gh-pages@v1.7.0
with:
token: ${{ secrets.GITHUB_TOKEN }}
charts_dir: deployment/helm/charts

View File

@@ -7,12 +7,12 @@ permissions:
# contents: write # only for delete-branch option
issues: write
pull-requests: write
jobs:
stale:
runs-on: ubuntu-latest
steps:
- uses: actions/stale@5bef64f19d7facfb25b37b414482c7164d639639 # ratchet:actions/stale@v9
- 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.'
@@ -20,3 +20,4 @@ jobs:
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

View File

@@ -16,18 +16,18 @@ 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=${{ github.run_id }}-scan-licenses"]
runs-on: [runs-on,runner=2cpu-linux-x64,"run-id=${{ github.run_id }}"]
steps:
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Set up Python
uses: actions/setup-python@e797f83bcb11b83ae66e0230d6156d7c80228e7c # ratchet:actions/setup-python@v6
uses: actions/setup-python@v5
with:
python-version: '3.11'
cache: 'pip'
@@ -35,7 +35,7 @@ jobs:
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
@@ -43,28 +43,28 @@ jobs:
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@e909b0226ff49d3235c99c4585bc617f49fff16a # ratchet:pilosus/action-pip-license-checker@v3
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 "${{ steps.license_check_report.outputs.report }}"
- name: Install npm dependencies
working-directory: ./web
run: npm ci
# be careful enabling the sarif and upload as it may spam the security tab
# with a huge amount of items. Work out the issues before enabling upload.
# with a huge amount of items. Work out the issues before enabling upload.
# - name: Run Trivy vulnerability scanner in repo mode
# if: always()
# uses: aquasecurity/trivy-action@b6643a29fecd7f34b3597bc6acb0a98b03d33ff8 # ratchet:aquasecurity/trivy-action@0.33.1
# uses: aquasecurity/trivy-action@0.29.0
# with:
# scan-type: fs
# scan-ref: .
@@ -73,7 +73,7 @@ jobs:
# severity: HIGH,CRITICAL
# # format: sarif
# # output: trivy-results.sarif
#
#
# # - name: Upload Trivy scan results to GitHub Security tab
# # uses: github/codeql-action/upload-sarif@v3
# # with:
@@ -81,14 +81,14 @@ jobs:
scan-trivy:
# See https://runs-on.com/runners/linux/
runs-on: [runs-on,runner=2cpu-linux-x64,"run-id=${{ github.run_id }}-scan-trivy"]
runs-on: [runs-on,runner=2cpu-linux-x64,"run-id=${{ github.run_id }}"]
steps:
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
uses: docker/setup-buildx-action@v3
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
uses: docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
@@ -98,7 +98,7 @@ jobs:
run: docker pull onyxdotapp/onyx-backend:latest
- name: Run Trivy vulnerability scanner on backend
uses: aquasecurity/trivy-action@b6643a29fecd7f34b3597bc6acb0a98b03d33ff8 # ratchet:aquasecurity/trivy-action@0.33.1
uses: aquasecurity/trivy-action@0.29.0
env:
TRIVY_DB_REPOSITORY: 'public.ecr.aws/aquasecurity/trivy-db:2'
TRIVY_JAVA_DB_REPOSITORY: 'public.ecr.aws/aquasecurity/trivy-java-db:1'
@@ -112,9 +112,9 @@ jobs:
# Web server
- name: Pull web server docker image
run: docker pull onyxdotapp/onyx-web-server:latest
- name: Run Trivy vulnerability scanner on web server
uses: aquasecurity/trivy-action@b6643a29fecd7f34b3597bc6acb0a98b03d33ff8 # ratchet:aquasecurity/trivy-action@0.33.1
uses: aquasecurity/trivy-action@0.29.0
env:
TRIVY_DB_REPOSITORY: 'public.ecr.aws/aquasecurity/trivy-db:2'
TRIVY_JAVA_DB_REPOSITORY: 'public.ecr.aws/aquasecurity/trivy-java-db:1'
@@ -130,7 +130,7 @@ jobs:
run: docker pull onyxdotapp/onyx-model-server:latest
- name: Run Trivy vulnerability scanner
uses: aquasecurity/trivy-action@b6643a29fecd7f34b3597bc6acb0a98b03d33ff8 # ratchet:aquasecurity/trivy-action@0.33.1
uses: aquasecurity/trivy-action@0.29.0
env:
TRIVY_DB_REPOSITORY: 'public.ecr.aws/aquasecurity/trivy-db:2'
TRIVY_JAVA_DB_REPOSITORY: 'public.ecr.aws/aquasecurity/trivy-java-db:1'
@@ -139,4 +139,4 @@ jobs:
scanners: license
severity: HIGH,CRITICAL
vuln-type: library
exit-code: 0
exit-code: 0

View File

@@ -1,7 +1,4 @@
name: External Dependency Unit Tests
concurrency:
group: External-Dependency-Unit-Tests-${{ github.workflow }}-${{ github.head_ref || github.event.workflow_run.head_branch || github.run_id }}
cancel-in-progress: true
on:
merge_group:
@@ -30,14 +27,13 @@ env:
jobs:
discover-test-dirs:
# NOTE: Github-hosted runners have about 20s faster queue times and are preferred here.
runs-on: ubuntu-slim
runs-on: ubuntu-latest
outputs:
test-dirs: ${{ steps.set-matrix.outputs.test-dirs }}
steps:
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Discover test directories
id: set-matrix
run: |
@@ -48,11 +44,8 @@ jobs:
external-dependency-unit-tests:
needs: discover-test-dirs
# Use larger runner with more resources for Vespa
runs-on:
- runs-on
- runner=2cpu-linux-arm64
- ${{ format('run-id={0}-external-dependency-unit-tests-job-{1}', github.run_id, strategy['job-index']) }}
- extras=s3-cache
runs-on: [runs-on, runner=16cpu-linux-x64, "run-id=${{ github.run_id }}"]
strategy:
fail-fast: false
matrix:
@@ -63,31 +56,42 @@ jobs:
MODEL_SERVER_HOST: "disabled"
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Setup Python and Install Dependencies
uses: ./.github/actions/setup-python-and-install-dependencies
- name: Setup Playwright
uses: ./.github/actions/setup-playwright
# needed for pulling Vespa, Redis, Postgres, and Minio images
# otherwise, we hit the "Unauthenticated users" limit
# https://docs.docker.com/docker-hub/usage/
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
- name: Set up Python
uses: actions/setup-python@v5
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
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
playwright install chromium
playwright install-deps chromium
- name: Set up Standard Dependencies
run: |
cd deployment/docker_compose
docker compose -f docker-compose.yml -f docker-compose.dev.yml up -d minio relational_db cache index
- name: Wait for services
run: |
echo "Waiting for services to be ready..."
sleep 30
# Wait for Vespa specifically
echo "Waiting for Vespa to be ready..."
timeout 300 bash -c 'until curl -f -s http://localhost:8081/ApplicationStatus > /dev/null 2>&1; do echo "Vespa not ready, waiting..."; sleep 10; done' || echo "Vespa timeout - continuing anyway"
echo "Services should be ready now"
- name: Run migrations
run: |
cd backend

View File

@@ -1,33 +1,30 @@
name: Helm - Lint and Test Charts
concurrency:
group: Helm-Lint-and-Test-Charts-${{ github.workflow }}-${{ github.head_ref || github.event.workflow_run.head_branch || github.run_id }}
cancel-in-progress: true
on:
merge_group:
pull_request:
branches: [ main ]
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=${{ github.run_id }}-helm-chart-check"]
runs-on: [runs-on,runner=8cpu-linux-x64,hdd=256,"run-id=${{ github.run_id }}"]
# fetch-depth 0 is required for helm/chart-testing-action
steps:
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
with:
fetch-depth: 0
- name: Set up Helm
uses: azure/setup-helm@1a275c3b69536ee54be43f2070a358922e12c8d4 # ratchet:azure/setup-helm@v4.3.1
uses: azure/setup-helm@v4.2.0
with:
version: v3.19.0
version: v3.17.0
- name: Set up chart-testing
uses: helm/chart-testing-action@6ec842c01de15ebb84c8627d2744a0c2f2755c9f # ratchet:helm/chart-testing-action@v2.8.0
uses: helm/chart-testing-action@v2.7.0
# 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)
@@ -44,7 +41,7 @@ jobs:
# - name: Force run chart-testing (list-changed)
# id: list-changed
# run: echo "changed=true" >> $GITHUB_OUTPUT
# lint all charts if any changes were detected
- name: Run chart-testing (lint)
if: steps.list-changed.outputs.changed == 'true'
@@ -54,7 +51,7 @@ jobs:
- name: Create kind cluster
if: steps.list-changed.outputs.changed == 'true'
uses: helm/kind-action@92086f6be054225fa813e0a4b13787fc9088faab # ratchet:helm/kind-action@v1.13.0
uses: helm/kind-action@v1.12.0
- name: Pre-install cluster status check
if: steps.list-changed.outputs.changed == 'true'
@@ -121,7 +118,7 @@ jobs:
if: steps.list-changed.outputs.changed == 'true'
run: |
echo "=== Starting chart installation with monitoring ==="
# Function to monitor cluster state
monitor_cluster() {
while true; do
@@ -143,11 +140,11 @@ jobs:
sleep 60
done
}
# Start monitoring in background
monitor_cluster &
MONITOR_PID=$!
# Set up cleanup
cleanup() {
echo "=== Cleaning up monitoring process ==="
@@ -156,10 +153,10 @@ jobs:
kubectl get pods --all-namespaces
kubectl get events --all-namespaces --sort-by=.lastTimestamp | tail -20
}
# Trap cleanup on exit
trap cleanup EXIT
# Run the actual installation with detailed logging
echo "=== Starting ct install ==="
set +e
@@ -217,15 +214,15 @@ jobs:
echo "=== Final cluster state ==="
kubectl get pods --all-namespaces
kubectl get events --all-namespaces --sort-by=.lastTimestamp | tail -10
echo "=== Pod descriptions for debugging ==="
kubectl describe pods --all-namespaces | grep -A 10 -B 3 "Failed\|Error\|Warning\|Pending" || echo "No problematic pods found"
echo "=== Recent logs for debugging ==="
kubectl logs --all-namespaces --tail=50 | grep -i "error\|timeout\|failed\|pull" || echo "No error logs found"
echo "=== Helm releases ==="
helm list --all-namespaces
# the following would install only changed charts, but we only have one chart so
# 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 }}

View File

@@ -11,6 +11,11 @@ on:
- "release/**"
env:
# Private Registry Configuration
PRIVATE_REGISTRY: experimental-registry.blacksmith.sh:5000
PRIVATE_REGISTRY_USERNAME: ${{ secrets.PRIVATE_REGISTRY_USERNAME }}
PRIVATE_REGISTRY_PASSWORD: ${{ secrets.PRIVATE_REGISTRY_PASSWORD }}
# Test Environment Variables
OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
@@ -26,17 +31,15 @@ env:
PERM_SYNC_SHAREPOINT_PRIVATE_KEY: ${{ secrets.PERM_SYNC_SHAREPOINT_PRIVATE_KEY }}
PERM_SYNC_SHAREPOINT_CERTIFICATE_PASSWORD: ${{ secrets.PERM_SYNC_SHAREPOINT_CERTIFICATE_PASSWORD }}
PERM_SYNC_SHAREPOINT_DIRECTORY_ID: ${{ secrets.PERM_SYNC_SHAREPOINT_DIRECTORY_ID }}
EXA_API_KEY: ${{ secrets.EXA_API_KEY }}
jobs:
discover-test-dirs:
# NOTE: Github-hosted runners have about 20s faster queue times and are preferred here.
runs-on: ubuntu-slim
runs-on: blacksmith-2vcpu-ubuntu-2404-arm
outputs:
test-dirs: ${{ steps.set-matrix.outputs.test-dirs }}
steps:
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Discover test directories
id: set-matrix
@@ -58,92 +61,105 @@ jobs:
all_dirs="[${all_dirs%,}]"
echo "test-dirs=$all_dirs" >> $GITHUB_OUTPUT
prepare-build:
runs-on: blacksmith-2vcpu-ubuntu-2404-arm
steps:
- name: Checkout code
uses: actions/checkout@v4
- name: Prepare build
uses: ./.github/actions/prepare-build
build-backend-image:
runs-on: [runs-on, runner=1cpu-linux-arm64, "run-id=${{ github.run_id }}-build-backend-image", "extras=ecr-cache"]
runs-on: blacksmith-16vcpu-ubuntu-2404-arm
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Login to Private Registry
uses: docker/login-action@v3
with:
registry: ${{ env.PRIVATE_REGISTRY }}
username: ${{ env.PRIVATE_REGISTRY_USERNAME }}
password: ${{ env.PRIVATE_REGISTRY_PASSWORD }}
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
# needed for pulling Vespa, Redis, Postgres, and Minio images
# otherwise, we hit the "Unauthenticated users" limit
# https://docs.docker.com/docker-hub/usage/
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
uses: useblacksmith/setup-docker-builder@v1
- name: Build and push Backend Docker image
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
uses: useblacksmith/build-push-action@v2
with:
context: ./backend
file: ./backend/Dockerfile
platforms: linux/arm64
tags: ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-backend:test-${{ github.run_id }}
push: true
tags: ${{ env.RUNS_ON_ECR_CACHE }}:integration-test-backend-test-${{ github.run_id }}
cache-from: type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:integration-test-backend-cache
cache-to: type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:integration-test-backend-cache,mode=max
outputs: type=registry
no-cache: ${{ vars.DOCKER_NO_CACHE == 'true' }}
build-model-server-image:
runs-on: [runs-on, runner=1cpu-linux-arm64, "run-id=${{ github.run_id }}-build-model-server-image", "extras=ecr-cache"]
runs-on: blacksmith-16vcpu-ubuntu-2404-arm
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Login to Private Registry
uses: docker/login-action@v3
with:
registry: ${{ env.PRIVATE_REGISTRY }}
username: ${{ env.PRIVATE_REGISTRY_USERNAME }}
password: ${{ env.PRIVATE_REGISTRY_PASSWORD }}
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
# needed for pulling Vespa, Redis, Postgres, and Minio images
# otherwise, we hit the "Unauthenticated users" limit
# https://docs.docker.com/docker-hub/usage/
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
uses: useblacksmith/setup-docker-builder@v1
- name: Build and push Model Server Docker image
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
uses: useblacksmith/build-push-action@v2
with:
context: ./backend
file: ./backend/Dockerfile.model_server
platforms: linux/arm64
tags: ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-model-server:test-${{ github.run_id }}
push: true
tags: ${{ env.RUNS_ON_ECR_CACHE }}:integration-test-model-server-test-${{ github.run_id }}
cache-from: type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:integration-test-model-server-cache
cache-to: type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:integration-test-model-server-cache,mode=max
outputs: type=registry
provenance: false
no-cache: ${{ vars.DOCKER_NO_CACHE == 'true' }}
build-integration-image:
runs-on: [runs-on, runner=2cpu-linux-arm64, "run-id=${{ github.run_id }}-build-integration-image", "extras=ecr-cache"]
needs: prepare-build
runs-on: blacksmith-16vcpu-ubuntu-2404-arm
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Login to Private Registry
uses: docker/login-action@v3
with:
registry: ${{ env.PRIVATE_REGISTRY }}
username: ${{ env.PRIVATE_REGISTRY_USERNAME }}
password: ${{ env.PRIVATE_REGISTRY_PASSWORD }}
- name: Download OpenAPI artifacts
uses: actions/download-artifact@v4
with:
name: openapi-artifacts
path: backend/generated/
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
uses: useblacksmith/setup-docker-builder@v1
# needed for pulling openapitools/openapi-generator-cli
# otherwise, we hit the "Unauthenticated users" limit
# https://docs.docker.com/docker-hub/usage/
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
- name: Build and push integration test Docker image
uses: useblacksmith/build-push-action@v2
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Build and push integration test image with Docker Bake
env:
REPOSITORY: ${{ env.RUNS_ON_ECR_CACHE }}
TAG: integration-test-${{ github.run_id }}
run: cd backend && docker buildx bake --push integration
context: ./backend
file: ./backend/tests/integration/Dockerfile
platforms: linux/arm64
tags: ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-integration:test-${{ github.run_id }}
push: true
outputs: type=registry
no-cache: ${{ vars.DOCKER_NO_CACHE == 'true' }}
integration-tests:
needs:
@@ -153,11 +169,7 @@ jobs:
build-model-server-image,
build-integration-image,
]
runs-on:
- runs-on
- runner=4cpu-linux-arm64
- ${{ format('run-id={0}-integration-tests-job-{1}', github.run_id, strategy['job-index']) }}
- extras=ecr-cache
runs-on: blacksmith-8vcpu-ubuntu-2404-arm
strategy:
fail-fast: false
@@ -165,19 +177,43 @@ jobs:
test-dir: ${{ fromJson(needs.discover-test-dirs.outputs.test-dirs) }}
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Login to Private Registry
uses: docker/login-action@v3
with:
registry: ${{ env.PRIVATE_REGISTRY }}
username: ${{ env.PRIVATE_REGISTRY_USERNAME }}
password: ${{ env.PRIVATE_REGISTRY_PASSWORD }}
# needed for pulling Vespa, Redis, Postgres, and Minio images
# otherwise, we hit the "Unauthenticated users" limit
# https://docs.docker.com/docker-hub/usage/
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
uses: docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Pull Docker images
run: |
# Pull all images from registry in parallel
echo "Pulling Docker images in parallel..."
# Pull images from private registry
(docker pull --platform linux/arm64 ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-backend:test-${{ github.run_id }}) &
(docker pull --platform linux/arm64 ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-model-server:test-${{ github.run_id }}) &
(docker pull --platform linux/arm64 ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-integration:test-${{ github.run_id }}) &
# Wait for all background jobs to complete
wait
echo "All Docker images pulled successfully"
# Re-tag to remove registry prefix for docker-compose
docker tag ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-backend:test-${{ github.run_id }} onyxdotapp/onyx-backend:test
docker tag ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-model-server:test-${{ github.run_id }} onyxdotapp/onyx-model-server:test
docker tag ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-integration:test-${{ github.run_id }} onyxdotapp/onyx-integration:test
# NOTE: Use pre-ping/null pool to reduce flakiness due to dropped connections
# NOTE: don't need web server for integration tests
- name: Start Docker containers
@@ -189,8 +225,7 @@ jobs:
POSTGRES_USE_NULL_POOL=true \
REQUIRE_EMAIL_VERIFICATION=false \
DISABLE_TELEMETRY=true \
ONYX_BACKEND_IMAGE=${{ env.RUNS_ON_ECR_CACHE }}:integration-test-backend-test-${{ github.run_id }} \
ONYX_MODEL_SERVER_IMAGE=${{ env.RUNS_ON_ECR_CACHE }}:integration-test-model-server-test-${{ github.run_id }} \
IMAGE_TAG=test \
INTEGRATION_TESTS_MODE=true \
CHECK_TTL_MANAGEMENT_TASK_FREQUENCY_IN_HOURS=0.001 \
docker compose -f docker-compose.yml -f docker-compose.dev.yml up \
@@ -217,15 +252,15 @@ jobs:
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
@@ -234,7 +269,7 @@ jobs:
else
echo "Service not ready yet (HTTP status $response). Retrying in 5 seconds..."
fi
sleep 5
done
echo "Finished waiting for service."
@@ -246,7 +281,7 @@ jobs:
-p mock-it-services-stack up -d
- name: Run Integration Tests for ${{ matrix.test-dir.name }}
uses: nick-fields/retry@ce71cc2ab81d554ebbe88c79ab5975992d79ba08 # ratchet:nick-fields/retry@v3
uses: nick-fields/retry@v3
with:
timeout_minutes: 20
max_attempts: 3
@@ -283,7 +318,7 @@ jobs:
-e TEST_WEB_HOSTNAME=test-runner \
-e MOCK_CONNECTOR_SERVER_HOST=mock_connector_server \
-e MOCK_CONNECTOR_SERVER_PORT=8001 \
${{ env.RUNS_ON_ECR_CACHE }}:integration-test-${{ github.run_id }} \
onyxdotapp/onyx-integration:test \
/app/tests/integration/${{ matrix.test-dir.path }}
# ------------------------------------------------------------
@@ -302,12 +337,18 @@ jobs:
- name: Upload logs
if: always()
uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # ratchet:actions/upload-artifact@v4
uses: actions/upload-artifact@v4
with:
name: docker-all-logs-${{ matrix.test-dir.name }}
path: ${{ github.workspace }}/docker-compose.log
# ------------------------------------------------------------
- name: Stop Docker containers
if: always()
run: |
cd deployment/docker_compose
docker compose down -v
multitenant-tests:
needs:
@@ -316,19 +357,35 @@ jobs:
build-model-server-image,
build-integration-image,
]
runs-on: [runs-on, runner=8cpu-linux-arm64, "run-id=${{ github.run_id }}-multitenant-tests", "extras=ecr-cache"]
runs-on: blacksmith-8vcpu-ubuntu-2404-arm
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Login to Private Registry
uses: docker/login-action@v3
with:
registry: ${{ env.PRIVATE_REGISTRY }}
username: ${{ env.PRIVATE_REGISTRY_USERNAME }}
password: ${{ env.PRIVATE_REGISTRY_PASSWORD }}
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
uses: docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Pull Docker images
run: |
(docker pull --platform linux/arm64 ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-backend:test-${{ github.run_id }}) &
(docker pull --platform linux/arm64 ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-model-server:test-${{ github.run_id }}) &
(docker pull --platform linux/arm64 ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-integration:test-${{ github.run_id }}) &
wait
docker tag ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-backend:test-${{ github.run_id }} onyxdotapp/onyx-backend:test
docker tag ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-model-server:test-${{ github.run_id }} onyxdotapp/onyx-model-server:test
docker tag ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-integration:test-${{ github.run_id }} onyxdotapp/onyx-integration:test
- name: Start Docker containers for multi-tenant tests
run: |
cd deployment/docker_compose
@@ -337,8 +394,7 @@ jobs:
AUTH_TYPE=cloud \
REQUIRE_EMAIL_VERIFICATION=false \
DISABLE_TELEMETRY=true \
ONYX_BACKEND_IMAGE=${{ env.RUNS_ON_ECR_CACHE }}:integration-test-backend-test-${{ github.run_id }} \
ONYX_MODEL_SERVER_IMAGE=${{ env.RUNS_ON_ECR_CACHE }}:integration-test-model-server-test-${{ github.run_id }} \
IMAGE_TAG=test \
DEV_MODE=true \
docker compose -f docker-compose.multitenant-dev.yml up \
relational_db \
@@ -401,8 +457,9 @@ jobs:
-e SKIP_RESET=true \
-e REQUIRE_EMAIL_VERIFICATION=false \
-e DISABLE_TELEMETRY=true \
-e IMAGE_TAG=test \
-e DEV_MODE=true \
${{ env.RUNS_ON_ECR_CACHE }}:integration-test-${{ github.run_id }} \
onyxdotapp/onyx-integration:test \
/app/tests/integration/multitenant_tests
- name: Dump API server logs (multi-tenant)
@@ -419,7 +476,7 @@ jobs:
- name: Upload logs (multi-tenant)
if: always()
uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # ratchet:actions/upload-artifact@v4
uses: actions/upload-artifact@v4
with:
name: docker-all-logs-multitenant
path: ${{ github.workspace }}/docker-compose-multitenant.log
@@ -430,13 +487,12 @@ jobs:
cd deployment/docker_compose
docker compose -f docker-compose.multitenant-dev.yml down -v
required:
# NOTE: Github-hosted runners have about 20s faster queue times and are preferred here.
runs-on: ubuntu-slim
required:
runs-on: blacksmith-2vcpu-ubuntu-2404-arm
needs: [integration-tests, multitenant-tests]
if: ${{ always() }}
steps:
- uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # ratchet:actions/github-script@v8
- uses: actions/github-script@v7
with:
script: |
const needs = ${{ toJSON(needs) }};

View File

@@ -11,14 +11,12 @@ jobs:
runs-on: ubuntu-latest
steps:
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Setup node
uses: actions/setup-node@2028fbc5c25fe9cf00d9f06a71cc4710d4507903 # ratchet:actions/setup-node@v4
uses: actions/setup-node@v4
with:
node-version: 22
cache: 'npm'
cache-dependency-path: ./web/package-lock.json
- name: Install node dependencies
working-directory: ./web
@@ -30,7 +28,7 @@ jobs:
- name: Upload coverage reports
if: always()
uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # ratchet:actions/upload-artifact@v4
uses: actions/upload-artifact@v4
with:
name: jest-coverage-${{ github.run_id }}
path: ./web/coverage

View File

@@ -27,7 +27,7 @@ jobs:
echo "::error::❌ Your PR title does not follow the Conventional Commits format.
This check ensures that all pull requests use clear, consistent titles that help automate changelogs and improve project history.
Please update your PR title to follow the Conventional Commits style.
Please update your PR title to follow the Conventional Commits style.
Here is a link to a blog explaining the reason why we've included the Conventional Commits style into our PR titles: https://xfuture-blog.com/working-with-conventional-commits
**Here are some examples of valid PR titles:**

View File

@@ -1,7 +1,4 @@
name: Ensure PR references Linear
concurrency:
group: Ensure-PR-references-Linear-${{ github.workflow }}-${{ github.head_ref || github.event.workflow_run.head_branch || github.run_id }}
cancel-in-progress: true
on:
pull_request:

View File

@@ -8,6 +8,11 @@ on:
types: [checks_requested]
env:
# Private Registry Configuration
PRIVATE_REGISTRY: experimental-registry.blacksmith.sh:5000
PRIVATE_REGISTRY_USERNAME: ${{ secrets.PRIVATE_REGISTRY_USERNAME }}
PRIVATE_REGISTRY_PASSWORD: ${{ secrets.PRIVATE_REGISTRY_PASSWORD }}
# Test Environment Variables
OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
@@ -26,13 +31,12 @@ env:
jobs:
discover-test-dirs:
# NOTE: Github-hosted runners have about 20s faster queue times and are preferred here.
runs-on: ubuntu-slim
runs-on: blacksmith-2vcpu-ubuntu-2404-arm
outputs:
test-dirs: ${{ steps.set-matrix.outputs.test-dirs }}
steps:
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Discover test directories
id: set-matrix
@@ -54,89 +58,106 @@ jobs:
all_dirs="[${all_dirs%,}]"
echo "test-dirs=$all_dirs" >> $GITHUB_OUTPUT
build-backend-image:
runs-on: [runs-on, runner=1cpu-linux-arm64, "run-id=${{ github.run_id }}-build-backend-image", "extras=ecr-cache"]
prepare-build:
runs-on: blacksmith-2vcpu-ubuntu-2404-arm
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Prepare build
uses: ./.github/actions/prepare-build
build-backend-image:
runs-on: blacksmith-16vcpu-ubuntu-2404-arm
steps:
- name: Checkout code
uses: actions/checkout@v4
- name: Login to Private Registry
uses: docker/login-action@v3
with:
registry: ${{ env.PRIVATE_REGISTRY }}
username: ${{ env.PRIVATE_REGISTRY_USERNAME }}
password: ${{ env.PRIVATE_REGISTRY_PASSWORD }}
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
# needed for pulling Vespa, Redis, Postgres, and Minio images
# otherwise, we hit the "Unauthenticated users" limit
# https://docs.docker.com/docker-hub/usage/
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
uses: useblacksmith/setup-docker-builder@v1
- name: Build and push Backend Docker image
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
uses: useblacksmith/build-push-action@v2
with:
context: ./backend
file: ./backend/Dockerfile
platforms: linux/arm64
tags: ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-backend:test-${{ github.run_id }}
push: true
tags: ${{ env.RUNS_ON_ECR_CACHE }}:integration-test-backend-test-${{ github.run_id }}
cache-from: type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:integration-test-backend-cache
cache-to: type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:integration-test-backend-cache,mode=max
outputs: type=registry
no-cache: ${{ vars.DOCKER_NO_CACHE == 'true' }}
build-model-server-image:
runs-on: [runs-on, runner=1cpu-linux-arm64, "run-id=${{ github.run_id }}-build-model-server-image", "extras=ecr-cache"]
runs-on: blacksmith-16vcpu-ubuntu-2404-arm
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Login to Private Registry
uses: docker/login-action@v3
with:
registry: ${{ env.PRIVATE_REGISTRY }}
username: ${{ env.PRIVATE_REGISTRY_USERNAME }}
password: ${{ env.PRIVATE_REGISTRY_PASSWORD }}
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
# needed for pulling Vespa, Redis, Postgres, and Minio images
# otherwise, we hit the "Unauthenticated users" limit
# https://docs.docker.com/docker-hub/usage/
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
uses: useblacksmith/setup-docker-builder@v1
- name: Build and push Model Server Docker image
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
uses: useblacksmith/build-push-action@v2
with:
context: ./backend
file: ./backend/Dockerfile.model_server
platforms: linux/arm64
tags: ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-model-server:test-${{ github.run_id }}
push: true
tags: ${{ env.RUNS_ON_ECR_CACHE }}:integration-test-model-server-test-${{ github.run_id }}
cache-from: type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:integration-test-model-server-cache
cache-to: type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:integration-test-model-server-cache,mode=max
outputs: type=registry
provenance: false
no-cache: ${{ vars.DOCKER_NO_CACHE == 'true' }}
build-integration-image:
runs-on: [runs-on, runner=2cpu-linux-arm64, "run-id=${{ github.run_id }}-build-integration-image", "extras=ecr-cache"]
needs: prepare-build
runs-on: blacksmith-16vcpu-ubuntu-2404-arm
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Login to Private Registry
uses: docker/login-action@v3
with:
registry: ${{ env.PRIVATE_REGISTRY }}
username: ${{ env.PRIVATE_REGISTRY_USERNAME }}
password: ${{ env.PRIVATE_REGISTRY_PASSWORD }}
- name: Download OpenAPI artifacts
uses: actions/download-artifact@v4
with:
name: openapi-artifacts
path: backend/generated/
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
uses: useblacksmith/setup-docker-builder@v1
# needed for pulling openapitools/openapi-generator-cli
# otherwise, we hit the "Unauthenticated users" limit
# https://docs.docker.com/docker-hub/usage/
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
- name: Build and push integration test Docker image
uses: useblacksmith/build-push-action@v2
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
context: ./backend
file: ./backend/tests/integration/Dockerfile
platforms: linux/arm64
tags: ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-integration:test-${{ github.run_id }}
push: true
outputs: type=registry
no-cache: ${{ vars.DOCKER_NO_CACHE == 'true' }}
- name: Build and push integration test image with Docker Bake
env:
REPOSITORY: ${{ env.RUNS_ON_ECR_CACHE }}
TAG: integration-test-${{ github.run_id }}
run: cd backend && docker buildx bake --push integration
integration-tests-mit:
needs:
@@ -146,11 +167,8 @@ jobs:
build-model-server-image,
build-integration-image,
]
runs-on:
- runs-on
- runner=4cpu-linux-arm64
- ${{ format('run-id={0}-integration-tests-mit-job-{1}', github.run_id, strategy['job-index']) }}
- extras=ecr-cache
# See https://docs.blacksmith.sh/blacksmith-runners/overview
runs-on: blacksmith-8vcpu-ubuntu-2404-arm
strategy:
fail-fast: false
@@ -158,19 +176,43 @@ jobs:
test-dir: ${{ fromJson(needs.discover-test-dirs.outputs.test-dirs) }}
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Login to Private Registry
uses: docker/login-action@v3
with:
registry: ${{ env.PRIVATE_REGISTRY }}
username: ${{ env.PRIVATE_REGISTRY_USERNAME }}
password: ${{ env.PRIVATE_REGISTRY_PASSWORD }}
# needed for pulling Vespa, Redis, Postgres, and Minio images
# otherwise, we hit the "Unauthenticated users" limit
# https://docs.docker.com/docker-hub/usage/
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
uses: docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Pull Docker images
run: |
# Pull all images from registry in parallel
echo "Pulling Docker images in parallel..."
# Pull images from private registry
(docker pull --platform linux/arm64 ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-backend:test-${{ github.run_id }}) &
(docker pull --platform linux/arm64 ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-model-server:test-${{ github.run_id }}) &
(docker pull --platform linux/arm64 ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-integration:test-${{ github.run_id }}) &
# Wait for all background jobs to complete
wait
echo "All Docker images pulled successfully"
# Re-tag to remove registry prefix for docker-compose
docker tag ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-backend:test-${{ github.run_id }} onyxdotapp/onyx-backend:test
docker tag ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-model-server:test-${{ github.run_id }} onyxdotapp/onyx-model-server:test
docker tag ${{ env.PRIVATE_REGISTRY }}/integration-test-onyx-integration:test-${{ github.run_id }} onyxdotapp/onyx-integration:test
# NOTE: Use pre-ping/null pool to reduce flakiness due to dropped connections
# NOTE: don't need web server for integration tests
- name: Start Docker containers
@@ -181,8 +223,7 @@ jobs:
POSTGRES_USE_NULL_POOL=true \
REQUIRE_EMAIL_VERIFICATION=false \
DISABLE_TELEMETRY=true \
ONYX_BACKEND_IMAGE=${{ env.RUNS_ON_ECR_CACHE }}:integration-test-backend-test-${{ github.run_id }} \
ONYX_MODEL_SERVER_IMAGE=${{ env.RUNS_ON_ECR_CACHE }}:integration-test-model-server-test-${{ github.run_id }} \
IMAGE_TAG=test \
INTEGRATION_TESTS_MODE=true \
docker compose -f docker-compose.yml -f docker-compose.dev.yml up \
relational_db \
@@ -208,15 +249,15 @@ jobs:
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
@@ -225,7 +266,7 @@ jobs:
else
echo "Service not ready yet (HTTP status $response). Retrying in 5 seconds..."
fi
sleep 5
done
echo "Finished waiting for service."
@@ -238,7 +279,7 @@ jobs:
# NOTE: Use pre-ping/null to reduce flakiness due to dropped connections
- name: Run Integration Tests for ${{ matrix.test-dir.name }}
uses: nick-fields/retry@ce71cc2ab81d554ebbe88c79ab5975992d79ba08 # ratchet:nick-fields/retry@v3
uses: nick-fields/retry@v3
with:
timeout_minutes: 20
max_attempts: 3
@@ -275,7 +316,7 @@ jobs:
-e TEST_WEB_HOSTNAME=test-runner \
-e MOCK_CONNECTOR_SERVER_HOST=mock_connector_server \
-e MOCK_CONNECTOR_SERVER_PORT=8001 \
${{ env.RUNS_ON_ECR_CACHE }}:integration-test-${{ github.run_id }} \
onyxdotapp/onyx-integration:test \
/app/tests/integration/${{ matrix.test-dir.path }}
# ------------------------------------------------------------
@@ -294,20 +335,25 @@ jobs:
- name: Upload logs
if: always()
uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # ratchet:actions/upload-artifact@v4
uses: actions/upload-artifact@v4
with:
name: docker-all-logs-${{ matrix.test-dir.name }}
path: ${{ github.workspace }}/docker-compose.log
# ------------------------------------------------------------
- name: Stop Docker containers
if: always()
run: |
cd deployment/docker_compose
docker compose down -v
required:
# NOTE: Github-hosted runners have about 20s faster queue times and are preferred here.
runs-on: ubuntu-slim
required:
runs-on: blacksmith-2vcpu-ubuntu-2404-arm
needs: [integration-tests-mit]
if: ${{ always() }}
steps:
- uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # ratchet:actions/github-script@v8
- uses: actions/github-script@v7
with:
script: |
const needs = ${{ toJSON(needs) }};

View File

@@ -6,6 +6,13 @@ concurrency:
on: push
env:
# AWS ECR Configuration
AWS_REGION: ${{ secrets.AWS_REGION || 'us-west-2' }}
ECR_REGISTRY: ${{ secrets.ECR_REGISTRY }}
AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID_ECR }}
AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY_ECR }}
BUILDX_NO_DEFAULT_ATTESTATIONS: 1
# Test Environment Variables
OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
@@ -16,153 +23,164 @@ env:
SLACK_CLIENT_ID: ${{ secrets.SLACK_CLIENT_ID }}
SLACK_CLIENT_SECRET: ${{ secrets.SLACK_CLIENT_SECRET }}
# for MCP Oauth tests
MCP_OAUTH_CLIENT_ID: ${{ secrets.MCP_OAUTH_CLIENT_ID }}
MCP_OAUTH_CLIENT_SECRET: ${{ secrets.MCP_OAUTH_CLIENT_SECRET }}
MCP_OAUTH_ISSUER: ${{ secrets.MCP_OAUTH_ISSUER }}
MCP_OAUTH_JWKS_URI: ${{ secrets.MCP_OAUTH_JWKS_URI }}
MCP_OAUTH_USERNAME: ${{ vars.MCP_OAUTH_USERNAME }}
MCP_OAUTH_PASSWORD: ${{ secrets.MCP_OAUTH_PASSWORD }}
MOCK_LLM_RESPONSE: true
MCP_TEST_SERVER_PORT: 8004
MCP_TEST_SERVER_URL: http://host.docker.internal:8004/mcp
MCP_TEST_SERVER_PUBLIC_URL: http://host.docker.internal:8004/mcp
MCP_TEST_SERVER_BIND_HOST: 0.0.0.0
MCP_TEST_SERVER_PUBLIC_HOST: host.docker.internal
MCP_SERVER_HOST: 0.0.0.0
MCP_SERVER_PUBLIC_HOST: host.docker.internal
MCP_SERVER_PUBLIC_URL: http://host.docker.internal:8004/mcp
jobs:
build-web-image:
runs-on: [runs-on, runner=4cpu-linux-arm64, "run-id=${{ github.run_id }}-build-web-image", "extras=ecr-cache"]
runs-on: blacksmith-8vcpu-ubuntu-2404-arm
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Configure AWS credentials
uses: aws-actions/configure-aws-credentials@v4
with:
aws-access-key-id: ${{ env.AWS_ACCESS_KEY_ID }}
aws-secret-access-key: ${{ env.AWS_SECRET_ACCESS_KEY }}
aws-region: ${{ env.AWS_REGION }}
- name: Login to Amazon ECR
id: login-ecr
uses: aws-actions/amazon-ecr-login@v2
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
# needed for pulling external images otherwise, we hit the "Unauthenticated users" limit
# https://docs.docker.com/docker-hub/usage/
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
uses: useblacksmith/setup-docker-builder@v1
- name: Build and push Web Docker image
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
uses: useblacksmith/build-push-action@v2
with:
context: ./web
file: ./web/Dockerfile
platforms: linux/arm64
tags: ${{ env.RUNS_ON_ECR_CACHE }}:playwright-test-web-${{ github.run_id }}
tags: ${{ env.ECR_REGISTRY }}/integration-test-onyx-web-server:playwright-test-${{ github.run_id }}
provenance: false
sbom: false
push: true
cache-from: type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:playwright-test-web-cache
cache-to: type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:playwright-test-web-cache,mode=max
outputs: type=registry
no-cache: ${{ vars.DOCKER_NO_CACHE == 'true' }}
build-backend-image:
runs-on: [runs-on, runner=1cpu-linux-arm64, "run-id=${{ github.run_id }}-build-backend-image", "extras=ecr-cache"]
runs-on: blacksmith-8vcpu-ubuntu-2404-arm
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Configure AWS credentials
uses: aws-actions/configure-aws-credentials@v4
with:
aws-access-key-id: ${{ env.AWS_ACCESS_KEY_ID }}
aws-secret-access-key: ${{ env.AWS_SECRET_ACCESS_KEY }}
aws-region: ${{ env.AWS_REGION }}
- name: Login to Amazon ECR
id: login-ecr
uses: aws-actions/amazon-ecr-login@v2
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
# needed for pulling external images otherwise, we hit the "Unauthenticated users" limit
# https://docs.docker.com/docker-hub/usage/
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
uses: useblacksmith/setup-docker-builder@v1
- name: Build and push Backend Docker image
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
uses: useblacksmith/build-push-action@v2
with:
context: ./backend
file: ./backend/Dockerfile
platforms: linux/arm64
tags: ${{ env.RUNS_ON_ECR_CACHE }}:playwright-test-backend-${{ github.run_id }}
tags: ${{ env.ECR_REGISTRY }}/integration-test-onyx-backend:playwright-test-${{ github.run_id }}
provenance: false
sbom: false
push: true
cache-from: type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:playwright-test-backend-cache
cache-to: type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:playwright-test-backend-cache,mode=max
outputs: type=registry
no-cache: ${{ vars.DOCKER_NO_CACHE == 'true' }}
build-model-server-image:
runs-on: [runs-on, runner=1cpu-linux-arm64, "run-id=${{ github.run_id }}-build-model-server-image", "extras=ecr-cache"]
runs-on: blacksmith-8vcpu-ubuntu-2404-arm
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Configure AWS credentials
uses: aws-actions/configure-aws-credentials@v4
with:
aws-access-key-id: ${{ env.AWS_ACCESS_KEY_ID }}
aws-secret-access-key: ${{ env.AWS_SECRET_ACCESS_KEY }}
aws-region: ${{ env.AWS_REGION }}
- name: Login to Amazon ECR
id: login-ecr
uses: aws-actions/amazon-ecr-login@v2
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # ratchet:docker/setup-buildx-action@v3
# needed for pulling external images otherwise, we hit the "Unauthenticated users" limit
# https://docs.docker.com/docker-hub/usage/
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
uses: useblacksmith/setup-docker-builder@v1
- name: Build and push Model Server Docker image
uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
uses: useblacksmith/build-push-action@v2
with:
context: ./backend
file: ./backend/Dockerfile.model_server
platforms: linux/arm64
tags: ${{ env.RUNS_ON_ECR_CACHE }}:playwright-test-model-server-${{ github.run_id }}
tags: ${{ env.ECR_REGISTRY }}/integration-test-onyx-model-server:playwright-test-${{ github.run_id }}
provenance: false
sbom: false
push: true
cache-from: type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:playwright-test-model-server-cache
cache-to: type=registry,ref=${{ env.RUNS_ON_ECR_CACHE }}:playwright-test-model-server-cache,mode=max
outputs: type=registry
no-cache: ${{ vars.DOCKER_NO_CACHE == 'true' }}
playwright-tests:
needs: [build-web-image, build-backend-image, build-model-server-image]
name: Playwright Tests (${{ matrix.project }})
runs-on: [runs-on, runner=8cpu-linux-arm64, "run-id=${{ github.run_id }}-playwright-tests-${{ matrix.project }}", "extras=ecr-cache"]
strategy:
fail-fast: false
matrix:
project: [admin, no-auth, exclusive]
name: Playwright Tests
runs-on: blacksmith-8vcpu-ubuntu-2404-arm
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
with:
fetch-depth: 0
- name: Configure AWS credentials
uses: aws-actions/configure-aws-credentials@v4
with:
aws-access-key-id: ${{ env.AWS_ACCESS_KEY_ID }}
aws-secret-access-key: ${{ env.AWS_SECRET_ACCESS_KEY }}
aws-region: ${{ env.AWS_REGION }}
- name: Login to Amazon ECR
id: login-ecr
uses: aws-actions/amazon-ecr-login@v2
# needed for pulling Vespa, Redis, Postgres, and Minio images
# otherwise, we hit the "Unauthenticated users" limit
# https://docs.docker.com/docker-hub/usage/
- name: Login to Docker Hub
uses: docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Pull Docker images
run: |
# Pull all images from ECR in parallel
echo "Pulling Docker images in parallel..."
(docker pull ${{ env.ECR_REGISTRY }}/integration-test-onyx-web-server:playwright-test-${{ github.run_id }}) &
(docker pull ${{ env.ECR_REGISTRY }}/integration-test-onyx-backend:playwright-test-${{ github.run_id }}) &
(docker pull ${{ env.ECR_REGISTRY }}/integration-test-onyx-model-server:playwright-test-${{ github.run_id }}) &
# Wait for all background jobs to complete
wait
echo "All Docker images pulled successfully"
# Re-tag with expected names for docker-compose
docker tag ${{ env.ECR_REGISTRY }}/integration-test-onyx-web-server:playwright-test-${{ github.run_id }} onyxdotapp/onyx-web-server:test
docker tag ${{ env.ECR_REGISTRY }}/integration-test-onyx-backend:playwright-test-${{ github.run_id }} onyxdotapp/onyx-backend:test
docker tag ${{ env.ECR_REGISTRY }}/integration-test-onyx-model-server:playwright-test-${{ github.run_id }} onyxdotapp/onyx-model-server:test
- name: Setup node
uses: actions/setup-node@2028fbc5c25fe9cf00d9f06a71cc4710d4507903 # ratchet:actions/setup-node@v4
uses: actions/setup-node@v4
with:
node-version: 22
cache: 'npm'
cache-dependency-path: ./web/package-lock.json
- name: Install node dependencies
working-directory: ./web
run: npm ci
- name: Cache playwright cache
uses: runs-on/cache@50350ad4242587b6c8c2baa2e740b1bc11285ff4 # ratchet:runs-on/cache@v4
with:
path: ~/.cache/ms-playwright
key: ${{ runner.os }}-playwright-npm-${{ hashFiles('web/package-lock.json') }}
restore-keys: |
${{ runner.os }}-playwright-npm-
- name: Install playwright browsers
working-directory: ./web
run: npx playwright install --with-deps
@@ -176,24 +194,13 @@ jobs:
EXA_API_KEY=${{ env.EXA_API_KEY }}
REQUIRE_EMAIL_VERIFICATION=false
DISABLE_TELEMETRY=true
ONYX_BACKEND_IMAGE=${{ env.RUNS_ON_ECR_CACHE }}:playwright-test-backend-${{ github.run_id }}
ONYX_MODEL_SERVER_IMAGE=${{ env.RUNS_ON_ECR_CACHE }}:playwright-test-model-server-${{ github.run_id }}
ONYX_WEB_SERVER_IMAGE=${{ env.RUNS_ON_ECR_CACHE }}:playwright-test-web-${{ github.run_id }}
IMAGE_TAG=test
EOF
# needed for pulling Vespa, Redis, Postgres, and Minio images
# otherwise, we hit the "Unauthenticated users" limit
# https://docs.docker.com/docker-hub/usage/
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- name: Start Docker containers
run: |
cd deployment/docker_compose
docker compose -f docker-compose.yml -f docker-compose.dev.yml -f docker-compose.mcp-oauth-test.yml up -d
docker compose -f docker-compose.yml -f docker-compose.dev.yml up -d
id: start_docker
- name: Wait for service to be ready
@@ -208,15 +215,15 @@ jobs:
while true; do
current_time=$(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
@@ -225,47 +232,24 @@ jobs:
else
echo "Service not ready yet (HTTP status $response). Retrying in 5 seconds..."
fi
sleep 5
done
echo "Finished waiting for service."
- name: Wait for MCP OAuth mock server
run: |
echo "Waiting for MCP OAuth mock server on port ${MCP_TEST_SERVER_PORT:-8004}..."
start_time=$(date +%s)
timeout=120
while true; do
current_time=$(date +%s)
elapsed_time=$((current_time - start_time))
if [ $elapsed_time -ge $timeout ]; then
echo "Timeout reached. MCP OAuth mock server did not become ready in ${timeout}s."
exit 1
fi
if curl -sf "http://localhost:${MCP_TEST_SERVER_PORT:-8004}/healthz" > /dev/null; then
echo "MCP OAuth mock server is ready!"
break
fi
sleep 3
done
- name: Run Playwright tests
working-directory: ./web
run: |
# Create test-results directory to ensure it exists for artifact upload
mkdir -p test-results
npx playwright test --project ${{ matrix.project }}
npx playwright test
- uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # ratchet:actions/upload-artifact@v4
- uses: actions/upload-artifact@v4
if: always()
with:
# Includes test results and trace.zip files
name: playwright-test-results-${{ matrix.project }}-${{ github.run_id }}
path: ./web/test-results/
# Includes test results and debug screenshots
name: playwright-test-results-${{ github.run_id }}
path: ./web/test-results
retention-days: 30
# save before stopping the containers so the logs can be captured
@@ -278,11 +262,15 @@ jobs:
- name: Upload logs
if: success() || failure()
uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # ratchet:actions/upload-artifact@v4
uses: actions/upload-artifact@v4
with:
name: docker-logs-${{ matrix.project }}-${{ github.run_id }}
name: docker-logs
path: ${{ github.workspace }}/docker-compose.log
- name: Stop Docker containers
run: |
cd deployment/docker_compose
docker compose down -v
# NOTE: Chromatic UI diff testing is currently disabled.
# We are using Playwright for local and CI testing without visual regression checks.
@@ -301,12 +289,12 @@ jobs:
# ]
# steps:
# - name: Checkout code
# uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
# uses: actions/checkout@v4
# with:
# fetch-depth: 0
# - name: Setup node
# uses: actions/setup-node@2028fbc5c25fe9cf00d9f06a71cc4710d4507903 # ratchet:actions/setup-node@v4
# uses: actions/setup-node@v4
# with:
# node-version: 22
@@ -315,7 +303,7 @@ jobs:
# run: npm ci
# - name: Download Playwright test results
# uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # ratchet:actions/download-artifact@v4
# uses: actions/download-artifact@v4
# with:
# name: test-results
# path: ./web/test-results

View File

@@ -1,7 +1,4 @@
name: Python Checks
concurrency:
group: Python-Checks-${{ github.workflow }}-${{ github.head_ref || github.event.workflow_run.head_branch || github.run_id }}
cancel-in-progress: true
on:
merge_group:
@@ -13,51 +10,58 @@ on:
jobs:
mypy-check:
# See https://runs-on.com/runners/linux/
# Note: Mypy seems quite optimized for x64 compared to arm64.
# Similarly, mypy is single-threaded and incremental, so 2cpu is sufficient.
runs-on: [runs-on, runner=2cpu-linux-x64, "run-id=${{ github.run_id }}-mypy-check", "extras=s3-cache"]
runs-on: [runs-on,runner=8cpu-linux-x64,"run-id=${{ github.run_id }}"]
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
- name: Checkout code
uses: actions/checkout@v4
# needed for pulling openapitools/openapi-generator-cli
# otherwise, we hit the "Unauthenticated users" limit
# https://docs.docker.com/docker-hub/usage/
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
- 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: Prepare build
uses: ./.github/actions/prepare-build
with:
docker-username: ${{ secrets.DOCKER_USERNAME }}
docker-password: ${{ secrets.DOCKER_TOKEN }}
- name: Generate OpenAPI schema
working-directory: ./backend
env:
PYTHONPATH: "."
run: |
python scripts/onyx_openapi_schema.py --filename generated/openapi.json
- name: Cache mypy cache
if: ${{ vars.DISABLE_MYPY_CACHE != 'true' }}
uses: runs-on/cache@50350ad4242587b6c8c2baa2e740b1bc11285ff4 # ratchet:runs-on/cache@v4
with:
path: backend/.mypy_cache
key: mypy-${{ runner.os }}-${{ hashFiles('**/*.py', '**/*.pyi', 'backend/pyproject.toml') }}
restore-keys: |
mypy-${{ runner.os }}-
- name: Generate OpenAPI Python client
working-directory: ./backend
run: |
docker run --rm \
-v "${{ github.workspace }}/backend/generated:/local" \
openapitools/openapi-generator-cli generate \
-i /local/openapi.json \
-g python \
-o /local/onyx_openapi_client \
--package-name onyx_openapi_client \
--skip-validate-spec \
--openapi-normalizer "SIMPLIFY_ONEOF_ANYOF=true,SET_OAS3_NULLABLE=true"
- name: Run MyPy
run: |
cd backend
mypy .
- name: Run MyPy
working-directory: ./backend
env:
MYPY_FORCE_COLOR: 1
TERM: xterm-256color
run: mypy .
- name: Check import order with reorder-python-imports
run: |
cd backend
find ./onyx -name "*.py" | xargs reorder-python-imports --py311-plus
- name: Check import order with reorder-python-imports
working-directory: ./backend
run: |
find ./onyx -name "*.py" | xargs reorder-python-imports --py311-plus
- name: Check code formatting with Black
working-directory: ./backend
run: black --check .
- name: Check code formatting with Black
run: |
cd backend
black --check .

View File

@@ -1,7 +1,4 @@
name: Connector Tests
concurrency:
group: Connector-Tests-${{ github.workflow }}-${{ github.head_ref || github.event.workflow_run.head_branch || github.run_id }}
cancel-in-progress: true
on:
merge_group:
@@ -122,26 +119,35 @@ env:
jobs:
connectors-check:
# See https://runs-on.com/runners/linux/
runs-on: [runs-on, runner=8cpu-linux-x64, "run-id=${{ github.run_id }}-connectors-check", "extras=s3-cache"]
runs-on: [runs-on, runner=8cpu-linux-x64, "run-id=${{ github.run_id }}"]
env:
PYTHONPATH: ./backend
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Setup Python and Install Dependencies
uses: ./.github/actions/setup-python-and-install-dependencies
- 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: Setup Playwright
uses: ./.github/actions/setup-playwright
- 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
playwright install chromium
playwright install-deps chromium
- name: Detect Connector changes
id: changes
uses: dorny/paths-filter@de90cc6fb38fc0963ad72b210f1f284cd68cea36 # ratchet:dorny/paths-filter@v3
uses: dorny/paths-filter@v3
with:
filters: |
hubspot:

View File

@@ -10,7 +10,7 @@ on:
description: 'Branch to run the workflow on'
required: false
default: 'main'
env:
# Bedrock
AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }}
@@ -28,17 +28,17 @@ env:
jobs:
model-check:
# See https://runs-on.com/runners/linux/
runs-on: [runs-on,runner=8cpu-linux-x64,"run-id=${{ github.run_id }}-model-check"]
runs-on: [runs-on,runner=8cpu-linux-x64,"run-id=${{ github.run_id }}"]
env:
PYTHONPATH: ./backend
steps:
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Login to Docker Hub
uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # ratchet:docker/login-action@v3
uses: docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USERNAME }}
password: ${{ secrets.DOCKER_TOKEN }}
@@ -53,9 +53,9 @@ jobs:
run: |
docker pull onyxdotapp/onyx-model-server:latest
docker tag onyxdotapp/onyx-model-server:latest onyxdotapp/onyx-model-server:test
- name: Set up Python
uses: actions/setup-python@e797f83bcb11b83ae66e0230d6156d7c80228e7c # ratchet:actions/setup-python@v6
uses: actions/setup-python@v5
with:
python-version: "3.11"
cache: "pip"
@@ -90,15 +90,15 @@ jobs:
while true; do
current_time=$(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:9000/api/health || echo "curl_error")
if [ "$response" = "200" ]; then
echo "Service is ready!"
break
@@ -107,11 +107,11 @@ jobs:
else
echo "Service not ready yet (HTTP status $response). Retrying in 5 seconds..."
fi
sleep 5
done
echo "Finished waiting for service."
- name: Run Tests
shell: script -q -e -c "bash --noprofile --norc -eo pipefail {0}"
run: |
@@ -127,7 +127,7 @@ jobs:
-H 'Content-type: application/json' \
--data '{"text":"Scheduled Model Tests failed! Check the run at: https://github.com/${{ github.repository }}/actions/runs/${{ github.run_id }}"}' \
$SLACK_WEBHOOK
- name: Dump all-container logs (optional)
if: always()
run: |
@@ -136,7 +136,14 @@ jobs:
- name: Upload logs
if: always()
uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # ratchet:actions/upload-artifact@v4
uses: actions/upload-artifact@v4
with:
name: docker-all-logs
path: ${{ github.workspace }}/docker-compose.log
- name: Stop Docker containers
if: always()
run: |
cd deployment/docker_compose
docker compose -f docker-compose.model-server-test.yml down -v

View File

@@ -1,7 +1,4 @@
name: Python Unit Tests
concurrency:
group: Python-Unit-Tests-${{ github.workflow }}-${{ github.head_ref || github.event.workflow_run.head_branch || github.run_id }}
cancel-in-progress: true
on:
merge_group:
@@ -13,8 +10,7 @@ on:
jobs:
backend-check:
# See https://runs-on.com/runners/linux/
runs-on: [runs-on, runner=2cpu-linux-arm64, "run-id=${{ github.run_id }}-backend-check"]
runs-on: [runs-on,runner=8cpu-linux-x64,"run-id=${{ github.run_id }}"]
env:
PYTHONPATH: ./backend
@@ -22,15 +18,27 @@ jobs:
SF_USERNAME: ${{ secrets.SF_USERNAME }}
SF_PASSWORD: ${{ secrets.SF_PASSWORD }}
SF_SECURITY_TOKEN: ${{ secrets.SF_SECURITY_TOKEN }}
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
- name: Setup Python and Install Dependencies
uses: ./.github/actions/setup-python-and-install-dependencies
- 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: 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
pip install --retries 5 --timeout 30 -r backend/requirements/model_server.txt
- name: Run Tests
shell: script -q -e -c "bash --noprofile --norc -eo pipefail {0}"

View File

@@ -10,17 +10,14 @@ on:
jobs:
quality-checks:
# See https://runs-on.com/runners/linux/
runs-on: [runs-on, runner=1cpu-linux-arm64, "run-id=${{ github.run_id }}-quality-checks"]
runs-on: [runs-on,runner=8cpu-linux-x64,"run-id=${{ github.run_id }}"]
steps:
- uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
- uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
- uses: actions/checkout@v4
with:
fetch-depth: 0
- uses: actions/setup-python@e797f83bcb11b83ae66e0230d6156d7c80228e7c # ratchet:actions/setup-python@v6
- uses: actions/setup-python@v5
with:
python-version: "3.11"
- name: Setup Terraform
uses: hashicorp/setup-terraform@b9cd54a3c349d3f38e8881555d616ced269862dd # ratchet:hashicorp/setup-terraform@v3
- uses: pre-commit/action@2c7b3805fd2a0fd8c1884dcaebf91fc102a13ecd # ratchet:pre-commit/action@v3.0.1
- 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) || '' }}

View File

@@ -1,47 +0,0 @@
name: Sync FOSS Repo
on:
schedule:
# Run daily at 3am PT (11am UTC during PST)
- cron: '0 11 * * *'
workflow_dispatch:
jobs:
sync-foss:
runs-on: ubuntu-latest
permissions:
contents: read
steps:
- name: Checkout main Onyx repo
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
with:
fetch-depth: 0
- name: Install git-filter-repo
run: |
sudo apt-get update && sudo apt-get install -y git-filter-repo
- name: Configure SSH for deploy key
env:
FOSS_REPO_DEPLOY_KEY: ${{ secrets.FOSS_REPO_DEPLOY_KEY }}
run: |
mkdir -p ~/.ssh
echo "$FOSS_REPO_DEPLOY_KEY" > ~/.ssh/id_ed25519
chmod 600 ~/.ssh/id_ed25519
ssh-keyscan github.com >> ~/.ssh/known_hosts
- name: Set Git config
run: |
git config --global user.name "onyx-bot"
git config --global user.email "bot@onyx.app"
- name: Build FOSS version
run: bash backend/scripts/make_foss_repo.sh
- name: Push to FOSS repo
env:
FOSS_REPO_URL: git@github.com:onyx-dot-app/onyx-foss.git
run: |
cd /tmp/foss_repo
git remote add public "$FOSS_REPO_URL"
git push --force public main

View File

@@ -9,7 +9,7 @@ permissions:
jobs:
create-and-push-tag:
runs-on: [runs-on, runner=2cpu-linux-x64, "run-id=${{ github.run_id }}-create-and-push-tag"]
runs-on: [runs-on, runner=2cpu-linux-x64, "run-id=${{ github.run_id }}"]
steps:
# actions using GITHUB_TOKEN cannot trigger another workflow, but we do want this to trigger docker pushes
@@ -19,7 +19,7 @@ jobs:
# Additional NOTE: even though this is named "rkuo", the actual key is tied to the onyx repo
# and not rkuo's personal account. It is fine to leave this key as is!
- name: Checkout code
uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # ratchet:actions/checkout@v4
uses: actions/checkout@v4
with:
ssh-key: "${{ secrets.RKUO_DEPLOY_KEY }}"

8
.gitignore vendored
View File

@@ -18,7 +18,6 @@ backend/tests/regression/search_quality/eval-*
backend/tests/regression/search_quality/search_eval_config.yaml
backend/tests/regression/search_quality/*.json
backend/onyx/evals/data/
backend/onyx/evals/one_off/*.json
*.log
# secret files
@@ -32,10 +31,6 @@ settings.json
/backend/tests/regression/answer_quality/search_test_config.yaml
*.egg-info
# Claude
AGENTS.md
CLAUDE.md
# Local .terraform directories
**/.terraform/*
@@ -45,6 +40,3 @@ CLAUDE.md
# Local .terraform.lock.hcl file
.terraform.lock.hcl
# MCP configs
.playwright-mcp

View File

@@ -1,15 +1,4 @@
repos:
- repo: https://github.com/pre-commit/pre-commit-hooks
rev: v4.6.0
hooks:
- id: check-yaml
files: ^.github/
- repo: https://github.com/rhysd/actionlint
rev: v1.7.8
hooks:
- id: actionlint
- repo: https://github.com/psf/black
rev: 25.1.0
hooks:
@@ -40,7 +29,6 @@ repos:
rev: v0.11.4
hooks:
- id: ruff
- repo: https://github.com/pre-commit/mirrors-prettier
rev: v3.1.0
hooks:
@@ -59,18 +47,12 @@ repos:
- repo: local
hooks:
- id: terraform-fmt
name: terraform fmt
entry: terraform fmt -recursive
language: system
pass_filenames: false
files: \.tf$
- id: check-lazy-imports
name: Check lazy imports
name: Check lazy imports are not directly imported
entry: python3 backend/scripts/check_lazy_imports.py
language: system
files: ^backend/(?!\.venv/).*\.py$
files: ^backend/.*\.py$
pass_filenames: false
# We would like to have a mypy pre-commit hook, but due to the fact that
# pre-commit runs in it's own isolated environment, we would need to install

View File

@@ -1,6 +1,6 @@
# Copy this file to .env in the .vscode folder
# Fill in the <REPLACE THIS> values as needed, it is recommended to set the GEN_AI_API_KEY value to avoid having to set up an LLM in the UI
# Also check out onyx/backend/scripts/restart_containers.sh for a script to restart the containers which Onyx relies on outside of VSCode/Cursor processes
# Also check out danswer/backend/scripts/restart_containers.sh for a script to restart the containers which Danswer relies on outside of VSCode/Cursor processes
# For local dev, often user Authentication is not needed
AUTH_TYPE=disabled
@@ -37,8 +37,8 @@ OPENAI_API_KEY=<REPLACE THIS>
GEN_AI_MODEL_VERSION=gpt-4o
FAST_GEN_AI_MODEL_VERSION=gpt-4o
# For Onyx Slack Bot, overrides the UI values so no need to set this up via UI every time
# Only needed if using OnyxBot
# For Danswer Slack Bot, overrides the UI values so no need to set this up via UI every time
# Only needed if using DanswerBot
#ONYX_BOT_SLACK_APP_TOKEN=<REPLACE THIS>
#ONYX_BOT_SLACK_BOT_TOKEN=<REPLACE THIS>
@@ -75,9 +75,4 @@ SHOW_EXTRA_CONNECTORS=True
LANGSMITH_TRACING="true"
LANGSMITH_ENDPOINT="https://api.smith.langchain.com"
LANGSMITH_API_KEY=<REPLACE_THIS>
LANGSMITH_PROJECT=<REPLACE_THIS>
# Local Confluence OAuth testing
# OAUTH_CONFLUENCE_CLOUD_CLIENT_ID=<REPLACE_THIS>
# OAUTH_CONFLUENCE_CLOUD_CLIENT_SECRET=<REPLACE_THIS>
# NEXT_PUBLIC_TEST_ENV=True
LANGSMITH_PROJECT=<REPLACE_THIS>

View File

@@ -194,15 +194,13 @@ alembic -n schema_private upgrade head
### Creating Migrations
```bash
# Create migration
alembic revision -m "description"
# Auto-generate migration
alembic revision --autogenerate -m "description"
# Multi-tenant migration
alembic -n schema_private revision -m "description"
alembic -n schema_private revision --autogenerate -m "description"
```
Write the migration manually and place it in the file that alembic creates when running the above command.
## Testing Strategy
There are 4 main types of tests within Onyx:

View File

@@ -197,19 +197,15 @@ alembic -n schema_private upgrade head
### Creating Migrations
```bash
# Create migration
alembic revision -m "description"
# Auto-generate migration
alembic revision --autogenerate -m "description"
# Multi-tenant migration
alembic -n schema_private revision -m "description"
alembic -n schema_private revision --autogenerate -m "description"
```
Write the migration manually and place it in the file that alembic creates when running the above command.
## Testing Strategy
First, you must activate the virtual environment with `source .venv/bin/activate`.
There are 4 main types of tests within Onyx:
### Unit Tests
@@ -220,7 +216,7 @@ write these for complex, isolated modules e.g. `citation_processing.py`.
To run them:
```bash
pytest -xv backend/tests/unit
python -m dotenv -f .vscode/.env run -- pytest -xv backend/tests/unit
```
### External Dependency Unit Tests

View File

@@ -94,12 +94,6 @@ If using PowerShell, the command slightly differs:
Install the required python dependencies:
```bash
pip install -r backend/requirements/combined.txt
```
or
```bash
pip install -r backend/requirements/default.txt
pip install -r backend/requirements/dev.txt
@@ -128,7 +122,7 @@ Onyx uses Node v22.20.0. We highly recommend you use [Node Version Manager (nvm)
to manage your Node installations. Once installed, you can run
```bash
nvm install 22 && nvm use 22
nvm install 22 && nvm use 22`
node -v # verify your active version
```

View File

@@ -1,34 +1,29 @@
<a name="readme-top"></a>
<h2 align="center">
<a href="https://www.onyx.app/?utm_source=onyx_repo&utm_medium=github&utm_campaign=readme"> <img width="50%" src="https://github.com/onyx-dot-app/onyx/blob/logo/OnyxLogoCropped.jpg?raw=true" /></a>
<a href="https://www.onyx.app/"> <img width="50%" src="https://github.com/onyx-dot-app/onyx/blob/logo/OnyxLogoCropped.jpg?raw=true)" /></a>
</h2>
<p align="center">Open Source AI Platform</p>
<p align="center">
<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" />
<img src="https://img.shields.io/badge/discord-join-blue.svg?logo=discord&logoColor=white" alt="Discord">
</a>
<a href="https://docs.onyx.app/?utm_source=onyx_repo&utm_medium=github&utm_campaign=readme" target="_blank">
<img src="https://img.shields.io/badge/docs-view-blue" alt="Documentation" />
<a href="https://docs.onyx.app/" target="_blank">
<img src="https://img.shields.io/badge/docs-view-blue" alt="Documentation">
</a>
<a href="https://www.onyx.app/?utm_source=onyx_repo&utm_medium=github&utm_campaign=readme" target="_blank">
<img src="https://img.shields.io/website?url=https://www.onyx.app&up_message=visit&up_color=blue" alt="Documentation" />
<a href="https://docs.onyx.app/" target="_blank">
<img src="https://img.shields.io/website?url=https://www.onyx.app&up_message=visit&up_color=blue" alt="Documentation">
</a>
<a href="https://github.com/onyx-dot-app/onyx/blob/main/LICENSE" target="_blank">
<img src="https://img.shields.io/static/v1?label=license&message=MIT&color=blue" alt="License" />
<img src="https://img.shields.io/static/v1?label=license&message=MIT&color=blue" alt="License">
</a>
</p>
<p align="center">
<a href="https://trendshift.io/repositories/12516" target="_blank">
<img src="https://trendshift.io/api/badge/repositories/12516" alt="onyx-dot-app/onyx | Trendshift" style="width: 250px; height: 55px;" />
</a>
</p>
**[Onyx](https://www.onyx.app/?utm_source=onyx_repo&utm_medium=github&utm_campaign=readme)** is a feature-rich, self-hostable Chat UI that works with any LLM. It is easy to deploy and can run in a completely airgapped environment.
**[Onyx](https://www.onyx.app/)** is a feature-rich, self-hostable Chat UI that works with any LLM. It is easy to deploy and can run in a completely airgapped environment.
Onyx comes loaded with advanced features like Agents, Web Search, RAG, MCP, Deep Research, Connectors to 40+ knowledge sources, and more.
@@ -57,7 +52,7 @@ Onyx comes loaded with advanced features like Agents, Web Search, RAG, MCP, Deep
Onyx works with all LLMs (like OpenAI, Anthropic, Gemini, etc.) and self-hosted LLMs (like Ollama, vLLM, etc.)
To learn more about the features, check out our [documentation](https://docs.onyx.app/welcome?utm_source=onyx_repo&utm_medium=github&utm_campaign=readme)!
To learn more about the features, check out our [documentation](https://docs.onyx.app/welcome)!
@@ -65,13 +60,13 @@ To learn more about the features, check out our [documentation](https://docs.ony
Onyx supports deployments in Docker, Kubernetes, Terraform, along with guides for major cloud providers.
See guides below:
- [Docker](https://docs.onyx.app/deployment/local/docker?utm_source=onyx_repo&utm_medium=github&utm_campaign=readme) or [Quickstart](https://docs.onyx.app/deployment/getting_started/quickstart?utm_source=onyx_repo&utm_medium=github&utm_campaign=readme) (best for most users)
- [Kubernetes](https://docs.onyx.app/deployment/local/kubernetes?utm_source=onyx_repo&utm_medium=github&utm_campaign=readme) (best for large teams)
- [Terraform](https://docs.onyx.app/deployment/local/terraform?utm_source=onyx_repo&utm_medium=github&utm_campaign=readme) (best for teams already using Terraform)
- Cloud specific guides (best if specifically using [AWS EKS](https://docs.onyx.app/deployment/cloud/aws/eks?utm_source=onyx_repo&utm_medium=github&utm_campaign=readme), [Azure VMs](https://docs.onyx.app/deployment/cloud/azure?utm_source=onyx_repo&utm_medium=github&utm_campaign=readme), etc.)
- [Docker](https://docs.onyx.app/deployment/local/docker) or [Quickstart](https://docs.onyx.app/deployment/getting_started/quickstart) (best for most users)
- [Kubernetes](https://docs.onyx.app/deployment/local/kubernetes) (best for large teams)
- [Terraform](https://docs.onyx.app/deployment/local/terraform) (best for teams already using Terraform)
- Cloud specific guides (best if specifically using [AWS EKS](https://docs.onyx.app/deployment/cloud/aws/eks), [Azure VMs](https://docs.onyx.app/deployment/cloud/azure), etc.)
> [!TIP]
> **To try Onyx for free without deploying, check out [Onyx Cloud](https://cloud.onyx.app/signup?utm_source=onyx_repo&utm_medium=github&utm_campaign=readme)**.
> **To try Onyx for free without deploying, check out [Onyx Cloud](https://cloud.onyx.app/signup)**.
@@ -95,7 +90,7 @@ There are two editions of Onyx:
- Onyx Community Edition (CE) is available freely under the MIT license.
- Onyx Enterprise Edition (EE) includes extra features that are primarily useful for larger organizations.
For feature details, check out [our website](https://www.onyx.app/pricing?utm_source=onyx_repo&utm_medium=github&utm_campaign=readme).
For feature details, check out [our website](https://www.onyx.app/pricing).

View File

@@ -7,12 +7,15 @@ have a contract or agreement with DanswerAI, you are not permitted to use the En
Edition features outside of personal development or testing purposes. Please reach out to \
founders@onyx.app for more information. Please visit https://github.com/onyx-dot-app/onyx"
# Default ONYX_VERSION, typically overriden during builds by GitHub Actions.
ARG ONYX_VERSION=0.0.0-dev
# DO_NOT_TRACK is used to disable telemetry for Unstructured
ENV DANSWER_RUNNING_IN_DOCKER="true" \
ENV ONYX_VERSION=${ONYX_VERSION} \
DANSWER_RUNNING_IN_DOCKER="true" \
DO_NOT_TRACK="true" \
PLAYWRIGHT_BROWSERS_PATH="/app/.cache/ms-playwright"
COPY --from=ghcr.io/astral-sh/uv:0.9.9 /uv /uvx /bin/
COPY --from=ghcr.io/astral-sh/uv:latest /uv /uvx /bin/
# Install system dependencies
# cmake needed for psycopg (postgres)
@@ -87,10 +90,6 @@ nltk.download('stopwords', quiet=True); \
nltk.download('punkt_tab', quiet=True);"
# nltk.download('wordnet', quiet=True); introduce this back if lemmatization is needed
# Pre-downloading tiktoken for setups with limited egress
RUN python -c "import tiktoken; \
tiktoken.get_encoding('cl100k_base')"
# Set up application files
WORKDIR /app
@@ -125,10 +124,6 @@ COPY --chown=onyx:onyx ./assets /app/assets
ENV PYTHONPATH=/app
# Default ONYX_VERSION, typically overriden during builds by GitHub Actions.
ARG ONYX_VERSION=0.0.0-dev
ENV ONYX_VERSION=${ONYX_VERSION}
# Default command which does nothing
# This container is used by api server and background which specify their own CMD
CMD ["tail", "-f", "/dev/null"]

View File

@@ -6,10 +6,13 @@ AI models for Onyx. This container and all the code is MIT Licensed and free for
You can find it at https://hub.docker.com/r/onyx/onyx-model-server. For more details, \
visit https://github.com/onyx-dot-app/onyx."
ENV DANSWER_RUNNING_IN_DOCKER="true" \
# Default ONYX_VERSION, typically overriden during builds by GitHub Actions.
ARG ONYX_VERSION=0.0.0-dev
ENV ONYX_VERSION=${ONYX_VERSION} \
DANSWER_RUNNING_IN_DOCKER="true" \
HF_HOME=/app/.cache/huggingface
COPY --from=ghcr.io/astral-sh/uv:0.9.9 /uv /uvx /bin/
COPY --from=ghcr.io/astral-sh/uv:latest /uv /uvx /bin/
# Create non-root user for security best practices
RUN mkdir -p /app && \
@@ -20,6 +23,24 @@ RUN mkdir -p /app && \
chmod 755 /var/log/onyx && \
chown onyx:onyx /var/log/onyx
# --- add toolchain needed for Rust/Python builds (fastuuid) ---
ENV RUSTUP_HOME=/usr/local/rustup \
CARGO_HOME=/usr/local/cargo \
PATH=/usr/local/cargo/bin:$PATH
RUN set -eux; \
apt-get update && apt-get install -y --no-install-recommends \
build-essential \
pkg-config \
curl \
ca-certificates \
# Install latest stable Rust (supports Cargo.lock v4)
&& curl -sSf https://sh.rustup.rs | sh -s -- -y --profile minimal --default-toolchain stable \
&& rustc --version && cargo --version \
&& apt-get remove -y --allow-remove-essential perl-base \
&& apt-get autoremove -y \
&& rm -rf /var/lib/apt/lists/*
COPY ./requirements/model_server.txt /tmp/requirements.txt
RUN uv pip install --system --no-cache-dir --upgrade \
-r /tmp/requirements.txt && \
@@ -62,8 +83,4 @@ COPY ./model_server /app/model_server
ENV PYTHONPATH=/app
# Default ONYX_VERSION, typically overriden during builds by GitHub Actions.
ARG ONYX_VERSION=0.0.0-dev
ENV ONYX_VERSION=${ONYX_VERSION}
CMD ["uvicorn", "model_server.main:app", "--host", "0.0.0.0", "--port", "9000"]

View File

@@ -1,33 +0,0 @@
"""add theme_preference to user
Revision ID: 09995b8811eb
Revises: 3d1cca026fe8
Create Date: 2025-10-24 08:58:50.246949
"""
from alembic import op
import sqlalchemy as sa
from onyx.db.enums import ThemePreference
# revision identifiers, used by Alembic.
revision = "09995b8811eb"
down_revision = "3d1cca026fe8"
branch_labels = None
depends_on = None
def upgrade() -> None:
op.add_column(
"user",
sa.Column(
"theme_preference",
sa.Enum(ThemePreference, native_enum=False),
nullable=True,
),
)
def downgrade() -> None:
op.drop_column("user", "theme_preference")

View File

@@ -1,72 +0,0 @@
"""add switchover_type field and remove background_reindex_enabled
Revision ID: 2acdef638fc2
Revises: a4f23d6b71c8
Create Date: 2025-01-XX XX:XX:XX.XXXXXX
"""
from alembic import op
import sqlalchemy as sa
from onyx.db.enums import SwitchoverType
# revision identifiers, used by Alembic.
revision = "2acdef638fc2"
down_revision = "a4f23d6b71c8"
branch_labels = None
depends_on = None
def upgrade() -> None:
# Add switchover_type column with default value of REINDEX
op.add_column(
"search_settings",
sa.Column(
"switchover_type",
sa.Enum(SwitchoverType, native_enum=False),
nullable=False,
server_default=SwitchoverType.REINDEX.value,
),
)
# Migrate existing data: set switchover_type based on background_reindex_enabled
# REINDEX where background_reindex_enabled=True, INSTANT where False
op.execute(
"""
UPDATE search_settings
SET switchover_type = CASE
WHEN background_reindex_enabled = true THEN 'REINDEX'
ELSE 'INSTANT'
END
"""
)
# Remove the background_reindex_enabled column (replaced by switchover_type)
op.drop_column("search_settings", "background_reindex_enabled")
def downgrade() -> None:
# Re-add the background_reindex_enabled column with default value of True
op.add_column(
"search_settings",
sa.Column(
"background_reindex_enabled",
sa.Boolean(),
nullable=False,
server_default="true",
),
)
# Set background_reindex_enabled based on switchover_type
op.execute(
"""
UPDATE search_settings
SET background_reindex_enabled = CASE
WHEN switchover_type = 'INSTANT' THEN false
ELSE true
END
"""
)
# Remove the switchover_type column
op.drop_column("search_settings", "switchover_type")

View File

@@ -12,7 +12,6 @@ from alembic import op
import sqlalchemy as sa
from sqlalchemy import text
import logging
import fastapi_users_db_sqlalchemy
logger = logging.getLogger("alembic.runtime.migration")
@@ -59,9 +58,6 @@ def upgrade() -> None:
logger.info("Dropping chat_session.folder_id...")
# Drop foreign key constraint first
op.execute(
"ALTER TABLE chat_session DROP CONSTRAINT IF EXISTS chat_session_chat_folder_fk"
)
op.execute(
"ALTER TABLE chat_session DROP CONSTRAINT IF EXISTS chat_session_folder_fk"
)
@@ -176,6 +172,20 @@ def downgrade() -> None:
"user_file", sa.Column("folder_id", sa.Integer(), nullable=True)
)
# Recreate chat_folder table
if "chat_folder" not in inspector.get_table_names():
op.create_table(
"chat_folder",
sa.Column("id", sa.Integer(), nullable=False),
sa.Column("user_id", sa.UUID(), nullable=False),
sa.Column("name", sa.String(), nullable=False),
sa.Column("created_at", sa.DateTime(timezone=True), nullable=False),
sa.PrimaryKeyConstraint("id"),
sa.ForeignKeyConstraint(
["user_id"], ["user.id"], name="chat_folder_user_fk"
),
)
# Recreate persona__user_folder table
if "persona__user_folder" not in inspector.get_table_names():
op.create_table(
@@ -187,26 +197,6 @@ def downgrade() -> None:
sa.ForeignKeyConstraint(["user_folder_id"], ["user_project.id"]),
)
# Recreate chat_folder table and related structures
if "chat_folder" not in inspector.get_table_names():
op.create_table(
"chat_folder",
sa.Column("id", sa.Integer(), nullable=False),
sa.Column(
"user_id",
fastapi_users_db_sqlalchemy.generics.GUID(),
nullable=True,
),
sa.Column("name", sa.String(), nullable=True),
sa.Column("display_priority", sa.Integer(), nullable=False),
sa.ForeignKeyConstraint(
["user_id"],
["user.id"],
name="chat_folder_user_id_fkey",
),
sa.PrimaryKeyConstraint("id"),
)
# Add folder_id back to chat_session
if "chat_session" in inspector.get_table_names():
columns = [col["name"] for col in inspector.get_columns("chat_session")]
@@ -218,7 +208,7 @@ def downgrade() -> None:
# Add foreign key if chat_folder exists
if "chat_folder" in inspector.get_table_names():
op.create_foreign_key(
"chat_session_chat_folder_fk",
"chat_session_folder_fk",
"chat_session",
"chat_folder",
["folder_id"],

View File

@@ -292,7 +292,7 @@ def downgrade() -> None:
logger.error("CRITICAL: Downgrading data cleanup cannot restore deleted data!")
logger.error("Data restoration requires backup files or database backup.")
# raise NotImplementedError(
# "Downgrade of legacy data cleanup is not supported. "
# "Deleted data must be restored from backups."
# )
raise NotImplementedError(
"Downgrade of legacy data cleanup is not supported. "
"Deleted data must be restored from backups."
)

View File

@@ -1,121 +0,0 @@
"""add_oauth_config_and_user_tokens
Revision ID: 3d1cca026fe8
Revises: c8a93a2af083
Create Date: 2025-10-21 13:27:34.274721
"""
from alembic import op
import fastapi_users_db_sqlalchemy
import sqlalchemy as sa
from sqlalchemy.dialects import postgresql
# revision identifiers, used by Alembic.
revision = "3d1cca026fe8"
down_revision = "c8a93a2af083"
branch_labels = None
depends_on = None
def upgrade() -> None:
# Create oauth_config table
op.create_table(
"oauth_config",
sa.Column("id", sa.Integer(), nullable=False),
sa.Column("name", sa.String(), nullable=False),
sa.Column("authorization_url", sa.Text(), nullable=False),
sa.Column("token_url", sa.Text(), nullable=False),
sa.Column("client_id", sa.LargeBinary(), nullable=False),
sa.Column("client_secret", sa.LargeBinary(), nullable=False),
sa.Column("scopes", postgresql.JSONB(astext_type=sa.Text()), nullable=True),
sa.Column(
"additional_params",
postgresql.JSONB(astext_type=sa.Text()),
nullable=True,
),
sa.Column(
"created_at",
sa.DateTime(timezone=True),
server_default=sa.text("now()"),
nullable=False,
),
sa.Column(
"updated_at",
sa.DateTime(timezone=True),
server_default=sa.text("now()"),
nullable=False,
),
sa.PrimaryKeyConstraint("id"),
sa.UniqueConstraint("name"),
)
# Create oauth_user_token table
op.create_table(
"oauth_user_token",
sa.Column("id", sa.Integer(), nullable=False),
sa.Column("oauth_config_id", sa.Integer(), nullable=False),
sa.Column(
"user_id",
fastapi_users_db_sqlalchemy.generics.GUID(),
nullable=False,
),
sa.Column("token_data", sa.LargeBinary(), nullable=False),
sa.Column(
"created_at",
sa.DateTime(timezone=True),
server_default=sa.text("now()"),
nullable=False,
),
sa.Column(
"updated_at",
sa.DateTime(timezone=True),
server_default=sa.text("now()"),
nullable=False,
),
sa.ForeignKeyConstraint(
["oauth_config_id"], ["oauth_config.id"], ondelete="CASCADE"
),
sa.ForeignKeyConstraint(["user_id"], ["user.id"], ondelete="CASCADE"),
sa.PrimaryKeyConstraint("id"),
sa.UniqueConstraint("oauth_config_id", "user_id", name="uq_oauth_user_token"),
)
# Create index on user_id for efficient user-based token lookups
# Note: unique constraint on (oauth_config_id, user_id) already creates
# an index for config-based lookups
op.create_index(
"ix_oauth_user_token_user_id",
"oauth_user_token",
["user_id"],
)
# Add oauth_config_id column to tool table
op.add_column("tool", sa.Column("oauth_config_id", sa.Integer(), nullable=True))
# Create foreign key from tool to oauth_config
op.create_foreign_key(
"tool_oauth_config_fk",
"tool",
"oauth_config",
["oauth_config_id"],
["id"],
ondelete="SET NULL",
)
def downgrade() -> None:
# Drop foreign key from tool to oauth_config
op.drop_constraint("tool_oauth_config_fk", "tool", type_="foreignkey")
# Drop oauth_config_id column from tool table
op.drop_column("tool", "oauth_config_id")
# Drop index on user_id
op.drop_index("ix_oauth_user_token_user_id", table_name="oauth_user_token")
# Drop oauth_user_token table (will cascade delete tokens)
op.drop_table("oauth_user_token")
# Drop oauth_config table
op.drop_table("oauth_config")

View File

@@ -1,88 +0,0 @@
"""add_personal_access_token_table
Revision ID: 5e1c073d48a3
Revises: 09995b8811eb
Create Date: 2025-10-30 17:30:24.308521
"""
from alembic import op
import sqlalchemy as sa
from sqlalchemy.dialects import postgresql
# revision identifiers, used by Alembic.
revision = "5e1c073d48a3"
down_revision = "09995b8811eb"
branch_labels = None
depends_on = None
def upgrade() -> None:
# Create personal_access_token table
op.create_table(
"personal_access_token",
sa.Column("id", sa.Integer(), nullable=False),
sa.Column("name", sa.String(), nullable=False),
sa.Column("hashed_token", sa.String(length=64), nullable=False),
sa.Column("token_display", sa.String(), nullable=False),
sa.Column(
"user_id",
postgresql.UUID(as_uuid=True),
nullable=False,
),
sa.Column(
"expires_at",
sa.DateTime(timezone=True),
nullable=True,
),
sa.Column(
"created_at",
sa.DateTime(timezone=True),
server_default=sa.text("now()"),
nullable=False,
),
sa.Column(
"last_used_at",
sa.DateTime(timezone=True),
nullable=True,
),
sa.Column(
"is_revoked",
sa.Boolean(),
server_default=sa.text("false"),
nullable=False,
),
sa.ForeignKeyConstraint(
["user_id"],
["user.id"],
ondelete="CASCADE",
),
sa.PrimaryKeyConstraint("id"),
sa.UniqueConstraint("hashed_token"),
)
# Create indexes
op.create_index(
"ix_personal_access_token_expires_at",
"personal_access_token",
["expires_at"],
unique=False,
)
op.create_index(
"ix_pat_user_created",
"personal_access_token",
["user_id", sa.text("created_at DESC")],
unique=False,
)
def downgrade() -> None:
# Drop indexes first
op.drop_index("ix_pat_user_created", table_name="personal_access_token")
op.drop_index(
"ix_personal_access_token_expires_at", table_name="personal_access_token"
)
# Drop table
op.drop_table("personal_access_token")

View File

@@ -45,23 +45,8 @@ def upgrade() -> None:
def downgrade() -> None:
bind = op.get_bind()
inspector = sa.inspect(bind)
if "chat_session" in inspector.get_table_names():
chat_session_fks = {
fk.get("name") for fk in inspector.get_foreign_keys("chat_session")
}
if "chat_session_chat_folder_fk" in chat_session_fks:
op.drop_constraint(
"chat_session_chat_folder_fk", "chat_session", type_="foreignkey"
)
chat_session_columns = {
col["name"] for col in inspector.get_columns("chat_session")
}
if "folder_id" in chat_session_columns:
op.drop_column("chat_session", "folder_id")
if "chat_folder" in inspector.get_table_names():
op.drop_table("chat_folder")
op.drop_constraint(
"chat_session_chat_folder_fk", "chat_session", type_="foreignkey"
)
op.drop_column("chat_session", "folder_id")
op.drop_table("chat_folder")

View File

@@ -180,162 +180,14 @@ def downgrade() -> None:
)
logger.error("Only proceed if absolutely necessary and have backups.")
bind = op.get_bind()
inspector = sa.inspect(bind)
# The downgrade would need to:
# 1. Add back integer columns
# 2. Generate new sequential IDs
# 3. Update all foreign key references
# 4. Swap primary keys back
# This is complex and risky, so we raise an error instead
# Capture existing primary key definitions so we can restore them after swaps
persona_pk = inspector.get_pk_constraint("persona__user_file") or {}
persona_pk_name = persona_pk.get("name")
persona_pk_cols = persona_pk.get("constrained_columns") or []
project_pk = inspector.get_pk_constraint("project__user_file") or {}
project_pk_name = project_pk.get("name")
project_pk_cols = project_pk.get("constrained_columns") or []
# Drop foreign keys that reference the UUID primary key
op.drop_constraint(
"persona__user_file_user_file_id_fkey",
"persona__user_file",
type_="foreignkey",
)
op.drop_constraint(
"fk_project__user_file_user_file_id",
"project__user_file",
type_="foreignkey",
)
# Drop primary keys that rely on the UUID column so we can replace it
if persona_pk_name:
op.drop_constraint(persona_pk_name, "persona__user_file", type_="primary")
if project_pk_name:
op.drop_constraint(project_pk_name, "project__user_file", type_="primary")
# Rebuild integer IDs on user_file using a sequence-backed column
op.execute("CREATE SEQUENCE IF NOT EXISTS user_file_id_seq")
op.add_column(
"user_file",
sa.Column(
"id_int",
sa.Integer(),
server_default=sa.text("nextval('user_file_id_seq')"),
nullable=False,
),
)
op.execute("ALTER SEQUENCE user_file_id_seq OWNED BY user_file.id_int")
# Prepare integer foreign key columns on referencing tables
op.add_column(
"persona__user_file",
sa.Column("user_file_id_int", sa.Integer(), nullable=True),
)
op.add_column(
"project__user_file",
sa.Column("user_file_id_int", sa.Integer(), nullable=True),
)
# Populate the new integer foreign key columns by mapping from the UUID IDs
op.execute(
"""
UPDATE persona__user_file AS p
SET user_file_id_int = uf.id_int
FROM user_file AS uf
WHERE p.user_file_id = uf.id
"""
)
op.execute(
"""
UPDATE project__user_file AS p
SET user_file_id_int = uf.id_int
FROM user_file AS uf
WHERE p.user_file_id = uf.id
"""
)
op.alter_column(
"persona__user_file",
"user_file_id_int",
existing_type=sa.Integer(),
nullable=False,
)
op.alter_column(
"project__user_file",
"user_file_id_int",
existing_type=sa.Integer(),
nullable=False,
)
# Remove the UUID foreign key columns and rename the integer replacements
op.drop_column("persona__user_file", "user_file_id")
op.alter_column(
"persona__user_file",
"user_file_id_int",
new_column_name="user_file_id",
existing_type=sa.Integer(),
nullable=False,
)
op.drop_column("project__user_file", "user_file_id")
op.alter_column(
"project__user_file",
"user_file_id_int",
new_column_name="user_file_id",
existing_type=sa.Integer(),
nullable=False,
)
# Swap the user_file primary key back to the integer column
op.drop_constraint("user_file_pkey", "user_file", type_="primary")
op.drop_column("user_file", "id")
op.alter_column(
"user_file",
"id_int",
new_column_name="id",
existing_type=sa.Integer(),
)
op.alter_column(
"user_file",
"id",
existing_type=sa.Integer(),
nullable=False,
server_default=sa.text("nextval('user_file_id_seq')"),
)
op.execute("ALTER SEQUENCE user_file_id_seq OWNED BY user_file.id")
op.execute(
"""
SELECT setval(
'user_file_id_seq',
GREATEST(COALESCE(MAX(id), 1), 1),
MAX(id) IS NOT NULL
)
FROM user_file
"""
)
op.create_primary_key("user_file_pkey", "user_file", ["id"])
# Restore primary keys on referencing tables
if persona_pk_cols:
op.create_primary_key(
"persona__user_file_pkey", "persona__user_file", persona_pk_cols
)
if project_pk_cols:
op.create_primary_key(
"project__user_file_pkey",
"project__user_file",
project_pk_cols,
)
# Recreate foreign keys pointing at the integer primary key
op.create_foreign_key(
"persona__user_file_user_file_id_fkey",
"persona__user_file",
"user_file",
["user_file_id"],
["id"],
)
op.create_foreign_key(
"fk_project__user_file_user_file_id",
"project__user_file",
"user_file",
["user_file_id"],
["id"],
raise NotImplementedError(
"Downgrade of UUID primary key swap is not supported due to data loss risk. "
"Manual intervention with data backup/restore is required."
)

View File

@@ -181,21 +181,12 @@ def upgrade() -> None:
sa.Column("user_file_id", psql.UUID(as_uuid=True), nullable=False),
sa.PrimaryKeyConstraint("project_id", "user_file_id"),
)
logger.info("Created project__user_file table")
# Only create the index if it doesn't exist
existing_indexes = [
ix["name"] for ix in inspector.get_indexes("project__user_file")
]
if "idx_project__user_file_user_file_id" not in existing_indexes:
op.create_index(
"idx_project__user_file_user_file_id",
"project__user_file",
["user_file_id"],
)
logger.info(
"Created index idx_project__user_file_user_file_id on project__user_file"
)
logger.info("Created project__user_file table")
logger.info("Migration 1 (schema additions) completed successfully")
@@ -210,7 +201,7 @@ def downgrade() -> None:
# Drop project__user_file table
if "project__user_file" in inspector.get_table_names():
# op.drop_index("idx_project__user_file_user_file_id", "project__user_file")
op.drop_index("idx_project__user_file_user_file_id", "project__user_file")
op.drop_table("project__user_file")
logger.info("Dropped project__user_file table")

View File

@@ -1,97 +0,0 @@
"""add config to federated_connector
Revision ID: 9drpiiw74ljy
Revises: 2acdef638fc2
Create Date: 2025-11-03 12:00:00.000000
"""
from alembic import op
import sqlalchemy as sa
from sqlalchemy.dialects import postgresql
# revision identifiers, used by Alembic.
revision = "9drpiiw74ljy"
down_revision = "2acdef638fc2"
branch_labels = None
depends_on = None
def upgrade() -> None:
connection = op.get_bind()
# Check if column already exists in current schema
result = connection.execute(
sa.text(
"""
SELECT column_name
FROM information_schema.columns
WHERE table_schema = current_schema()
AND table_name = 'federated_connector'
AND column_name = 'config'
"""
)
)
column_exists = result.fetchone() is not None
# Add config column with default empty object (only if it doesn't exist)
if not column_exists:
op.add_column(
"federated_connector",
sa.Column(
"config", postgresql.JSONB(), nullable=False, server_default="{}"
),
)
# Data migration: Single bulk update for all Slack connectors
connection.execute(
sa.text(
"""
WITH connector_configs AS (
SELECT
fc.id as connector_id,
CASE
WHEN fcds.entities->'channels' IS NOT NULL
AND jsonb_typeof(fcds.entities->'channels') = 'array'
AND jsonb_array_length(fcds.entities->'channels') > 0
THEN
jsonb_build_object(
'channels', fcds.entities->'channels',
'search_all_channels', false
) ||
CASE
WHEN fcds.entities->'include_dm' IS NOT NULL
THEN jsonb_build_object('include_dm', fcds.entities->'include_dm')
ELSE '{}'::jsonb
END
ELSE
jsonb_build_object('search_all_channels', true) ||
CASE
WHEN fcds.entities->'include_dm' IS NOT NULL
THEN jsonb_build_object('include_dm', fcds.entities->'include_dm')
ELSE '{}'::jsonb
END
END as config
FROM federated_connector fc
LEFT JOIN LATERAL (
SELECT entities
FROM federated_connector__document_set
WHERE federated_connector_id = fc.id
AND entities IS NOT NULL
ORDER BY id
LIMIT 1
) fcds ON true
WHERE fc.source = 'FEDERATED_SLACK'
AND fcds.entities IS NOT NULL
)
UPDATE federated_connector fc
SET config = cc.config
FROM connector_configs cc
WHERE fc.id = cc.connector_id
"""
)
)
def downgrade() -> None:
op.drop_column("federated_connector", "config")

View File

@@ -1,61 +0,0 @@
"""add llm provider persona restrictions
Revision ID: a4f23d6b71c8
Revises: 5e1c073d48a3
Create Date: 2025-10-21 00:00:00.000000
"""
from alembic import op
import sqlalchemy as sa
# revision identifiers, used by Alembic.
revision = "a4f23d6b71c8"
down_revision = "5e1c073d48a3"
branch_labels = None
depends_on = None
def upgrade() -> None:
op.create_table(
"llm_provider__persona",
sa.Column("llm_provider_id", sa.Integer(), nullable=False),
sa.Column("persona_id", sa.Integer(), nullable=False),
sa.ForeignKeyConstraint(
["llm_provider_id"], ["llm_provider.id"], ondelete="CASCADE"
),
sa.ForeignKeyConstraint(["persona_id"], ["persona.id"], ondelete="CASCADE"),
sa.PrimaryKeyConstraint("llm_provider_id", "persona_id"),
)
op.create_index(
"ix_llm_provider__persona_llm_provider_id",
"llm_provider__persona",
["llm_provider_id"],
)
op.create_index(
"ix_llm_provider__persona_persona_id",
"llm_provider__persona",
["persona_id"],
)
op.create_index(
"ix_llm_provider__persona_composite",
"llm_provider__persona",
["persona_id", "llm_provider_id"],
)
def downgrade() -> None:
op.drop_index(
"ix_llm_provider__persona_composite",
table_name="llm_provider__persona",
)
op.drop_index(
"ix_llm_provider__persona_persona_id",
table_name="llm_provider__persona",
)
op.drop_index(
"ix_llm_provider__persona_llm_provider_id",
table_name="llm_provider__persona",
)
op.drop_table("llm_provider__persona")

View File

@@ -1,27 +0,0 @@
variable "REPOSITORY" {
default = "onyxdotapp/onyx-integration"
}
variable "TAG" {
default = "latest"
}
target "backend" {
context = "."
dockerfile = "Dockerfile"
}
target "integration" {
context = "."
dockerfile = "tests/integration/Dockerfile"
// Provide the base image via build context from the backend target
contexts = {
base = "target:backend"
}
cache-from = ["type=registry,ref=${REPOSITORY}:integration-test-backend-cache"]
cache-to = ["type=registry,ref=${REPOSITORY}:integration-test-backend-cache,mode=max"]
tags = ["${REPOSITORY}:${TAG}"]
}

View File

@@ -18,7 +18,7 @@
<!-- <document type="danswer_chunk" mode="index" /> -->
{{ document_elements }}
</documents>
<nodes count="50">
<nodes count="60">
<resources vcpu="8.0" memory="128.0Gb" architecture="arm64" storage-type="local"
disk="475.0Gb" />
</nodes>

View File

@@ -139,13 +139,19 @@ def get_all_space_permissions(
) -> dict[str, ExternalAccess]:
logger.debug("Getting space permissions")
# Gets all the spaces in the Confluence instance
all_space_keys = [
key
for space in confluence_client.retrieve_confluence_spaces(
limit=REQUEST_PAGINATION_LIMIT,
all_space_keys = []
start = 0
while True:
spaces_batch = confluence_client.get_all_spaces(
start=start, limit=REQUEST_PAGINATION_LIMIT
)
if (key := space.get("key"))
]
for space in spaces_batch.get("results", []):
all_space_keys.append(space.get("key"))
if len(spaces_batch.get("results", [])) < REQUEST_PAGINATION_LIMIT:
break
start += len(spaces_batch.get("results", []))
# Gets the permissions for each space
logger.debug(f"Got {len(all_space_keys)} spaces from confluence")

View File

@@ -8,7 +8,7 @@ from fastapi import Request
from fastapi import Response
from ee.onyx.auth.users import decode_anonymous_user_jwt_token
from onyx.auth.utils import extract_tenant_from_auth_header
from onyx.auth.api_key import extract_tenant_from_api_key_header
from onyx.configs.constants import ANONYMOUS_USER_COOKIE_NAME
from onyx.configs.constants import TENANT_ID_COOKIE_NAME
from onyx.db.engine.sql_engine import is_valid_schema_name
@@ -49,13 +49,13 @@ async def _get_tenant_id_from_request(
) -> str:
"""
Attempt to extract tenant_id from:
1) The API key or PAT (Personal Access Token) header
1) The API key header
2) The Redis-based token (stored in Cookie: fastapiusersauth)
3) The anonymous user cookie
Fallback: POSTGRES_DEFAULT_SCHEMA
"""
# Check for API key or PAT in Authorization header
tenant_id = extract_tenant_from_auth_header(request)
# Check for API key
tenant_id = extract_tenant_from_api_key_header(request)
if tenant_id is not None:
return tenant_id

View File

@@ -76,7 +76,6 @@ class ConfluenceCloudOAuth:
"read:confluence-content.permission%20"
"read:confluence-user%20"
"read:confluence-groups%20"
"read:space:confluence%20"
"readonly:content.attachment:confluence%20"
"search:confluence%20"
# granular scope

View File

@@ -161,7 +161,7 @@ def handle_send_message_simple_with_history(
persona_id=req.persona_id,
)
llm, _ = get_llms_for_persona(persona=chat_session.persona, user=user)
llm, _ = get_llms_for_persona(persona=chat_session.persona)
llm_tokenizer = get_tokenizer(
model_name=llm.config.model_name,

View File

@@ -24,7 +24,6 @@ from onyx.chat.models import PersonaOverrideConfig
from onyx.chat.models import QADocsResponse
from onyx.chat.process_message import gather_stream
from onyx.chat.process_message import stream_chat_message_objects
from onyx.configs.chat_configs import NUM_RETURNED_HITS
from onyx.configs.onyxbot_configs import MAX_THREAD_CONTEXT_PERCENTAGE
from onyx.context.search.models import SavedSearchDocWithContent
from onyx.context.search.models import SearchRequest
@@ -49,42 +48,9 @@ logger = setup_logger()
basic_router = APIRouter(prefix="/query")
class DocumentSearchPagination(BaseModel):
offset: int
limit: int
returned_count: int
has_more: bool
next_offset: int | None = None
class DocumentSearchResponse(BaseModel):
top_documents: list[SavedSearchDocWithContent]
llm_indices: list[int]
pagination: DocumentSearchPagination
def _normalize_pagination(limit: int | None, offset: int | None) -> tuple[int, int]:
if limit is None:
resolved_limit = NUM_RETURNED_HITS
else:
resolved_limit = limit
if resolved_limit <= 0:
raise HTTPException(
status_code=400, detail="retrieval_options.limit must be positive"
)
if offset is None:
resolved_offset = 0
else:
resolved_offset = offset
if resolved_offset < 0:
raise HTTPException(
status_code=400, detail="retrieval_options.offset cannot be negative"
)
return resolved_limit, resolved_offset
@basic_router.post("/document-search")
@@ -98,10 +64,6 @@ def handle_search_request(
logger.notice(f"Received document search query: {query}")
llm, fast_llm = get_default_llms()
pagination_limit, pagination_offset = _normalize_pagination(
limit=search_request.retrieval_options.limit,
offset=search_request.retrieval_options.offset,
)
search_pipeline = SearchPipeline(
search_request=SearchRequest(
@@ -110,8 +72,8 @@ def handle_search_request(
human_selected_filters=search_request.retrieval_options.filters,
enable_auto_detect_filters=search_request.retrieval_options.enable_auto_detect_filters,
persona=None, # For simplicity, default settings should be good for this search
offset=pagination_offset,
limit=pagination_limit + 1,
offset=search_request.retrieval_options.offset,
limit=search_request.retrieval_options.limit,
rerank_settings=search_request.rerank_settings,
evaluation_type=search_request.evaluation_type,
chunks_above=search_request.chunks_above,
@@ -154,9 +116,6 @@ def handle_search_request(
for section in top_sections
]
# Track whether the underlying retrieval produced more items than requested
has_more_results = len(top_docs) > pagination_limit
# Deduping happens at the last step to avoid harming quality by dropping content early on
deduped_docs = top_docs
dropped_inds = None
@@ -175,22 +134,7 @@ def handle_search_request(
dropped_indices=dropped_inds,
)
paginated_docs = deduped_docs[:pagination_limit]
llm_indices = [index for index in llm_indices if index < len(paginated_docs)]
has_more = has_more_results
pagination = DocumentSearchPagination(
offset=pagination_offset,
limit=pagination_limit,
returned_count=len(paginated_docs),
has_more=has_more,
next_offset=(pagination_offset + pagination_limit) if has_more else None,
)
return DocumentSearchResponse(
top_documents=paginated_docs,
llm_indices=llm_indices,
pagination=pagination,
)
return DocumentSearchResponse(top_documents=deduped_docs, llm_indices=llm_indices)
def get_answer_stream(
@@ -218,7 +162,7 @@ def get_answer_stream(
is_for_edit=False,
)
llm = get_main_llm_from_tuple(get_llms_for_persona(persona=persona_info, user=user))
llm = get_main_llm_from_tuple(get_llms_for_persona(persona_info))
llm_tokenizer = get_tokenizer(
model_name=llm.config.model_name,

View File

@@ -517,7 +517,7 @@ def run_analysis(intent_req: IntentRequest) -> tuple[bool, list[str]]:
try:
keywords = map_keywords(model_input.input_ids[0], tokenizer, keyword_preds)
except Exception as e:
logger.warning(
logger.error(
f"Failed to extract keywords for query: {intent_req.query} due to {e}"
)
# Fallback to keeping all words

View File

@@ -1,47 +0,0 @@
from typing import Any
from typing import Literal
from typing import TypeAlias
from pydantic import BaseModel
from onyx.llm.model_response import ModelResponseStream
class ToolCallStreamItem(BaseModel):
call_id: str | None = None
id: str | None = None
name: str | None = None
arguments: str | None = None
type: Literal["function_call"] = "function_call"
index: int | None = None
class ToolCallOutputStreamItem(BaseModel):
call_id: str | None = None
output: Any
type: Literal["function_call_output"] = "function_call_output"
RunItemStreamEventDetails: TypeAlias = ToolCallStreamItem | ToolCallOutputStreamItem
class RunItemStreamEvent(BaseModel):
type: Literal[
"message_start",
"message_done",
"reasoning_start",
"reasoning_done",
"tool_call",
"tool_call_output",
]
details: RunItemStreamEventDetails | None = None
StreamEvent: TypeAlias = ModelResponseStream | RunItemStreamEvent

View File

@@ -1,215 +0,0 @@
import json
from collections.abc import Iterator
from collections.abc import Sequence
from dataclasses import dataclass
from typing import Any
from onyx.agents.agent_framework.models import RunItemStreamEvent
from onyx.agents.agent_framework.models import StreamEvent
from onyx.agents.agent_framework.models import ToolCallOutputStreamItem
from onyx.agents.agent_framework.models import ToolCallStreamItem
from onyx.llm.interfaces import LanguageModelInput
from onyx.llm.interfaces import LLM
from onyx.llm.interfaces import ToolChoiceOptions
from onyx.llm.message_types import ChatCompletionMessage
from onyx.llm.message_types import ToolCall
from onyx.llm.model_response import ModelResponseStream
from onyx.tools.tool import RunContextWrapper
from onyx.tools.tool import Tool
@dataclass
class QueryResult:
stream: Iterator[StreamEvent]
new_messages_stateful: list[ChatCompletionMessage]
def _serialize_tool_output(output: Any) -> str:
if isinstance(output, str):
return output
try:
return json.dumps(output)
except TypeError:
return str(output)
def _update_tool_call_with_delta(
tool_calls_in_progress: dict[int, dict[str, Any]],
tool_call_delta: Any,
) -> None:
index = tool_call_delta.index
if index not in tool_calls_in_progress:
tool_calls_in_progress[index] = {
"id": None,
"name": None,
"arguments": "",
}
if tool_call_delta.id:
tool_calls_in_progress[index]["id"] = tool_call_delta.id
if tool_call_delta.function:
if tool_call_delta.function.name:
tool_calls_in_progress[index]["name"] = tool_call_delta.function.name
if tool_call_delta.function.arguments:
tool_calls_in_progress[index][
"arguments"
] += tool_call_delta.function.arguments
def query(
llm_with_default_settings: LLM,
messages: LanguageModelInput,
tools: Sequence[Tool],
context: Any,
tool_choice: ToolChoiceOptions | None = None,
) -> QueryResult:
tool_definitions = [tool.tool_definition() for tool in tools]
tools_by_name = {tool.name: tool for tool in tools}
new_messages_stateful: list[ChatCompletionMessage] = []
def stream_generator() -> Iterator[StreamEvent]:
reasoning_started = False
message_started = False
tool_calls_in_progress: dict[int, dict[str, Any]] = {}
content_parts: list[str] = []
reasoning_parts: list[str] = []
for chunk in llm_with_default_settings.stream(
prompt=messages,
tools=tool_definitions,
tool_choice=tool_choice,
):
assert isinstance(chunk, ModelResponseStream)
delta = chunk.choice.delta
finish_reason = chunk.choice.finish_reason
if delta.reasoning_content:
reasoning_parts.append(delta.reasoning_content)
if not reasoning_started:
yield RunItemStreamEvent(type="reasoning_start")
reasoning_started = True
if delta.content:
content_parts.append(delta.content)
if reasoning_started:
yield RunItemStreamEvent(type="reasoning_done")
reasoning_started = False
if not message_started:
yield RunItemStreamEvent(type="message_start")
message_started = True
if delta.tool_calls:
if reasoning_started and not message_started:
yield RunItemStreamEvent(type="reasoning_done")
reasoning_started = False
if message_started:
yield RunItemStreamEvent(type="message_done")
message_started = False
for tool_call_delta in delta.tool_calls:
_update_tool_call_with_delta(
tool_calls_in_progress, tool_call_delta
)
yield chunk
if not finish_reason:
continue
if message_started:
yield RunItemStreamEvent(type="message_done")
message_started = False
if finish_reason == "tool_calls" and tool_calls_in_progress:
sorted_tool_calls = sorted(tool_calls_in_progress.items())
# Build tool calls for the message and execute tools
assistant_tool_calls: list[ToolCall] = []
tool_outputs: dict[str, str] = {}
for _, tool_call_data in sorted_tool_calls:
call_id = tool_call_data["id"]
name = tool_call_data["name"]
arguments_str = tool_call_data["arguments"]
if call_id is None or name is None:
continue
assistant_tool_calls.append(
{
"id": call_id,
"type": "function",
"function": {
"name": name,
"arguments": arguments_str,
},
}
)
yield RunItemStreamEvent(
type="tool_call",
details=ToolCallStreamItem(
call_id=call_id,
name=name,
arguments=arguments_str,
),
)
if name in tools_by_name:
tool = tools_by_name[name]
arguments = json.loads(arguments_str)
run_context = RunContextWrapper(context=context)
# TODO: Instead of executing sequentially, execute in parallel
# In practice, it's not a must right now since we don't use parallel
# tool calls, so kicking the can down the road for now.
output = tool.run_v2(run_context, **arguments)
tool_outputs[call_id] = _serialize_tool_output(output)
yield RunItemStreamEvent(
type="tool_call_output",
details=ToolCallOutputStreamItem(
call_id=call_id,
output=output,
),
)
new_messages_stateful.append(
{
"role": "assistant",
"content": None,
"tool_calls": assistant_tool_calls,
}
)
for _, tool_call_data in sorted_tool_calls:
call_id = tool_call_data["id"]
if call_id in tool_outputs:
new_messages_stateful.append(
{
"role": "tool",
"content": tool_outputs[call_id],
"tool_call_id": call_id,
}
)
elif finish_reason == "stop" and content_parts:
new_messages_stateful.append(
{
"role": "assistant",
"content": "".join(content_parts),
}
)
return QueryResult(
stream=stream_generator(),
new_messages_stateful=new_messages_stateful,
)

View File

@@ -2,27 +2,15 @@ from collections.abc import Sequence
from langchain.schema.messages import BaseMessage
from onyx.agents.agent_sdk.message_types import AgentSDKMessage
from onyx.agents.agent_sdk.message_types import AssistantMessageWithContent
from onyx.agents.agent_sdk.message_types import ImageContent
from onyx.agents.agent_sdk.message_types import InputTextContent
from onyx.agents.agent_sdk.message_types import SystemMessage
from onyx.agents.agent_sdk.message_types import UserMessage
# TODO: Currently, we only support native API input for images. For other
# files, we process the content and share it as text in the message. In
# the future, we might support native file uploads for other types of files.
def base_messages_to_agent_sdk_msgs(
msgs: Sequence[BaseMessage],
is_responses_api: bool,
) -> list[AgentSDKMessage]:
return [_base_message_to_agent_sdk_msg(msg, is_responses_api) for msg in msgs]
def base_messages_to_agent_sdk_msgs(msgs: Sequence[BaseMessage]) -> list[dict]:
return [_base_message_to_agent_sdk_msg(msg) for msg in msgs]
def _base_message_to_agent_sdk_msg(
msg: BaseMessage, is_responses_api: bool
) -> AgentSDKMessage:
def _base_message_to_agent_sdk_msg(msg: BaseMessage) -> dict:
message_type_to_agent_sdk_role = {
"human": "user",
"system": "system",
@@ -32,136 +20,59 @@ def _base_message_to_agent_sdk_msg(
# Convert content to Agent SDK format
content = msg.content
if isinstance(content, str):
# For system/user/assistant messages, use InputTextContent
if role in ("system", "user"):
input_text_content: list[InputTextContent | ImageContent] = [
InputTextContent(type="input_text", text=content)
]
if role == "system":
# SystemMessage only accepts InputTextContent
system_msg: SystemMessage = {
"role": "system",
"content": [InputTextContent(type="input_text", text=content)],
}
return system_msg
else: # user
user_msg: UserMessage = {
"role": "user",
"content": input_text_content,
}
return user_msg
else: # assistant
assistant_msg: AssistantMessageWithContent
if is_responses_api:
from onyx.agents.agent_sdk.message_types import OutputTextContent
assistant_msg = {
"role": "assistant",
"content": [OutputTextContent(type="output_text", text=content)],
}
else:
assistant_msg = {
"role": "assistant",
"content": [InputTextContent(type="input_text", text=content)],
}
return assistant_msg
elif isinstance(content, list):
# For lists, we need to process based on the role
if role == "assistant":
# For responses API, use OutputTextContent; otherwise use InputTextContent
assistant_content: list[InputTextContent | OutputTextContent] = []
if is_responses_api:
from onyx.agents.agent_sdk.message_types import OutputTextContent
for item in content:
if isinstance(item, str):
assistant_content.append(
OutputTextContent(type="output_text", text=item)
)
elif isinstance(item, dict) and item.get("type") == "text":
assistant_content.append(
OutputTextContent(
type="output_text", text=item.get("text", "")
)
)
else:
raise ValueError(
f"Unexpected item type for assistant message: {type(item)}. Item: {item}"
)
else:
for item in content:
if isinstance(item, str):
assistant_content.append(
InputTextContent(type="input_text", text=item)
)
elif isinstance(item, dict) and item.get("type") == "text":
assistant_content.append(
InputTextContent(
type="input_text", text=item.get("text", "")
)
)
else:
raise ValueError(
f"Unexpected item type for assistant message: {type(item)}. Item: {item}"
)
assistant_msg_list: AssistantMessageWithContent = {
"role": "assistant",
"content": assistant_content,
# Convert string to structured text format
structured_content = [
{
"type": "input_text",
"text": content,
}
return assistant_msg_list
else: # system or user - use InputTextContent
input_content: list[InputTextContent | ImageContent] = []
for item in content:
if isinstance(item, str):
input_content.append(InputTextContent(type="input_text", text=item))
elif isinstance(item, dict):
item_type = item.get("type")
if item_type == "text":
input_content.append(
InputTextContent(
type="input_text", text=item.get("text", "")
)
)
elif item_type == "image_url":
# Convert image_url to input_image format
image_url = item.get("image_url", {})
if isinstance(image_url, dict):
url = image_url.get("url", "")
else:
url = image_url
input_content.append(
ImageContent(
type="input_image", image_url=url, detail="auto"
)
)
else:
raise ValueError(f"Unexpected item type: {item_type}")
else:
raise ValueError(
f"Unexpected item type: {type(item)}. Item: {item}"
)
]
elif isinstance(content, list):
# Content is already a list, process each item
structured_content = []
for item in content:
if isinstance(item, str):
structured_content.append(
{
"type": "input_text",
"text": item,
}
)
elif isinstance(item, dict):
# Handle different item types
item_type = item.get("type")
if role == "system":
# SystemMessage only accepts InputTextContent (no images)
text_only_content = [
c for c in input_content if c["type"] == "input_text"
]
system_msg_list: SystemMessage = {
"role": "system",
"content": text_only_content, # type: ignore[typeddict-item]
}
return system_msg_list
else: # user
user_msg_list: UserMessage = {
"role": "user",
"content": input_content,
}
return user_msg_list
if item_type == "text":
# Convert text type to input_text
structured_content.append(
{
"type": "input_text",
"text": item.get("text", ""),
}
)
elif item_type == "image_url":
# Convert image_url to input_image format
image_url = item.get("image_url", {})
if isinstance(image_url, dict):
url = image_url.get("url", "")
else:
url = image_url
structured_content.append(
{
"type": "input_image",
"image_url": url,
"detail": "auto",
}
)
else:
raise ValueError(f"Unexpected item type: {type(item)}. Item: {item}")
else:
raise ValueError(
f"Unexpected content type: {type(content)}. Content: {content}"
)
return {
"role": role,
"content": structured_content,
}

View File

@@ -1,125 +0,0 @@
"""Strongly typed message structures for Agent SDK messages."""
from typing import Literal
from typing import NotRequired
from typing_extensions import TypedDict
class InputTextContent(TypedDict):
type: Literal["input_text"]
text: str
class OutputTextContent(TypedDict):
type: Literal["output_text"]
text: str
TextContent = InputTextContent | OutputTextContent
class ImageContent(TypedDict):
type: Literal["input_image"]
image_url: str
detail: str
# Tool call structures
class ToolCallFunction(TypedDict):
name: str
arguments: str
class ToolCall(TypedDict):
id: str
type: Literal["function"]
function: ToolCallFunction
# Message types
class SystemMessage(TypedDict):
role: Literal["system"]
content: list[InputTextContent] # System messages use input text
class UserMessage(TypedDict):
role: Literal["user"]
content: list[
InputTextContent | ImageContent
] # User messages use input text or images
class AssistantMessageWithContent(TypedDict):
role: Literal["assistant"]
content: list[
InputTextContent | OutputTextContent
] # Assistant messages use output_text for responses API compatibility
class AssistantMessageWithToolCalls(TypedDict):
role: Literal["assistant"]
tool_calls: list[ToolCall]
class AssistantMessageDuringAgentRun(TypedDict):
role: Literal["assistant"]
id: str
content: (
list[InputTextContent | OutputTextContent] | list[ToolCall]
) # Assistant runtime messages receive output_text from agents SDK for responses API compatibility
status: Literal["completed", "failed", "in_progress"]
type: Literal["message"]
class ToolMessage(TypedDict):
role: Literal["tool"]
content: str
tool_call_id: str
class FunctionCallMessage(TypedDict):
"""Agent SDK function call message format."""
type: Literal["function_call"]
id: NotRequired[str]
call_id: str
name: str
arguments: str
class FunctionCallOutputMessage(TypedDict):
"""Agent SDK function call output message format."""
type: Literal["function_call_output"]
call_id: str
output: str
class SummaryText(TypedDict):
"""Summary text item in reasoning messages."""
text: str
type: Literal["summary_text"]
class ReasoningMessage(TypedDict):
"""Agent SDK reasoning message format."""
id: str
type: Literal["reasoning"]
summary: list[SummaryText]
# Union type for all Agent SDK messages
AgentSDKMessage = (
SystemMessage
| UserMessage
| AssistantMessageWithContent
| AssistantMessageWithToolCalls
| AssistantMessageDuringAgentRun
| ToolMessage
| FunctionCallMessage
| FunctionCallOutputMessage
| ReasoningMessage
)

View File

@@ -1,36 +0,0 @@
from typing import Any
from agents.models.openai_responses import Converter as OpenAIResponsesConverter
# TODO: I am very sad that I have to monkey patch this :(
# Basically, OpenAI agents sdk doesn't convert the tool choice correctly
# when they have a built-in tool in their framework, like they do for web_search
# and image_generation.
# Going to open up a thread with OpenAI agents team to see what they recommend
# or what we can fix.
# A discussion is warranted, but we likely want to just write our own LitellmModel for
# the OpenAI agents SDK since they probably don't really care about Litellm and will
# prioritize functionality for their own models.
def monkey_patch_convert_tool_choice_to_ignore_openai_hosted_web_search() -> None:
if (
getattr(OpenAIResponsesConverter.convert_tool_choice, "__name__", "")
== "_patched_convert_tool_choice"
):
return
orig_func = OpenAIResponsesConverter.convert_tool_choice.__func__ # type: ignore[attr-defined]
def _patched_convert_tool_choice(cls: type, tool_choice: Any) -> Any:
# Handle OpenAI hosted tools that we have custom implementations for
# Without this patch, the library uses special formatting that breaks our custom tools
# See: https://platform.openai.com/docs/api-reference/responses/create#responses_create-tool_choice-hosted_tool-type
if tool_choice == "web_search":
return {"type": "function", "name": "web_search"}
if tool_choice == "image_generation":
return {"type": "function", "name": "image_generation"}
return orig_func(cls, tool_choice)
OpenAIResponsesConverter.convert_tool_choice = classmethod( # type: ignore[method-assign, assignment]
_patched_convert_tool_choice
)

View File

@@ -2,7 +2,6 @@ import asyncio
import queue
import threading
from collections.abc import Iterator
from collections.abc import Sequence
from typing import Generic
from typing import Optional
from typing import TypeVar
@@ -12,7 +11,6 @@ from agents import RunResultStreaming
from agents import TContext
from agents.run import Runner
from onyx.agents.agent_sdk.message_types import AgentSDKMessage
from onyx.utils.threadpool_concurrency import run_in_background
T = TypeVar("T")
@@ -43,7 +41,7 @@ class SyncAgentStream(Generic[T]):
self,
*,
agent: Agent,
input: Sequence[AgentSDKMessage],
input: list[dict],
context: TContext | None = None,
max_turns: int = 100,
queue_maxsize: int = 0,

View File

@@ -14,10 +14,10 @@ from onyx.agents.agent_search.models import GraphConfig
from onyx.agents.agent_search.shared_graph_utils.agent_prompt_ops import (
trim_prompt_piece,
)
from onyx.configs.constants import DocumentSource
from onyx.prompts.agents.dc_prompts import DC_OBJECT_NO_BASE_DATA_EXTRACTION_PROMPT
from onyx.prompts.agents.dc_prompts import DC_OBJECT_SEPARATOR
from onyx.prompts.agents.dc_prompts import DC_OBJECT_WITH_BASE_DATA_EXTRACTION_PROMPT
from onyx.secondary_llm_flows.source_filter import strings_to_document_sources
from onyx.utils.logger import setup_logger
from onyx.utils.threadpool_concurrency import run_with_timeout
@@ -61,12 +61,10 @@ def search_objects(
if agent_1_independent_sources_str is None:
raise ValueError("Agent 1 Independent Research Sources not found")
document_sources = strings_to_document_sources(
[
x.strip().lower()
for x in agent_1_independent_sources_str.split(DC_OBJECT_SEPARATOR)
]
)
document_sources = [
DocumentSource(x.strip().lower())
for x in agent_1_independent_sources_str.split(DC_OBJECT_SEPARATOR)
]
agent_1_output_objective = extract_section(
agent_1_instructions, "Output Objective:"
@@ -121,7 +119,7 @@ def search_objects(
try:
llm_response = run_with_timeout(
30,
primary_llm.invoke_langchain,
primary_llm.invoke,
prompt=msg,
timeout_override=30,
max_tokens=300,

View File

@@ -155,7 +155,7 @@ def research_object_source(
try:
llm_response = run_with_timeout(
30,
primary_llm.invoke_langchain,
primary_llm.invoke,
prompt=msg,
timeout_override=30,
max_tokens=300,

View File

@@ -76,7 +76,7 @@ def consolidate_object_research(
try:
llm_response = run_with_timeout(
30,
primary_llm.invoke_langchain,
primary_llm.invoke,
prompt=msg,
timeout_override=30,
max_tokens=300,

View File

@@ -1,7 +1,6 @@
from enum import Enum
from pydantic import BaseModel
from pydantic import ConfigDict
from onyx.agents.agent_search.dr.enums import DRPath
from onyx.agents.agent_search.dr.sub_agents.image_generation.models import (
@@ -75,7 +74,8 @@ class OrchestratorTool(BaseModel):
cost: float
tool_object: Tool | None = None # None for CLOSER
model_config = ConfigDict(arbitrary_types_allowed=True)
class Config:
arbitrary_types_allowed = True
class IterationInstructions(BaseModel):

View File

@@ -41,7 +41,7 @@ from onyx.agents.agent_search.shared_graph_utils.utils import write_custom_event
from onyx.agents.agent_search.utils import create_question_prompt
from onyx.chat.chat_utils import build_citation_map_from_numbers
from onyx.chat.chat_utils import saved_search_docs_from_llm_docs
from onyx.chat.memories import get_memories
from onyx.chat.memories import make_memories_callback
from onyx.chat.models import PromptConfig
from onyx.chat.prompt_builder.citations_prompt import build_citations_system_message
from onyx.chat.prompt_builder.citations_prompt import build_citations_user_message
@@ -98,6 +98,14 @@ from onyx.utils.logger import setup_logger
logger = setup_logger()
def _format_tool_name(tool_name: str) -> str:
"""Convert tool name to LLM-friendly format."""
name = tool_name.replace(" ", "_")
# take care of camel case like GetAPIKey -> GET_API_KEY for LLM readability
name = re.sub(r"(?<=[a-z0-9])(?=[A-Z])|(?<=[A-Z])(?=[A-Z][a-z])", "_", name)
return name.upper()
def _get_available_tools(
db_session: Session,
graph_config: GraphConfig,
@@ -490,9 +498,11 @@ def clarifier(
if graph_config.tooling.search_tool
else None
)
memories = get_memories(user, db_session)
memories_callback = make_memories_callback(user, db_session)
assistant_system_prompt = handle_company_awareness(assistant_system_prompt)
assistant_system_prompt = handle_memories(assistant_system_prompt, memories)
assistant_system_prompt = handle_memories(
assistant_system_prompt, memories_callback
)
chat_history_string = (
get_chat_history_string(
@@ -552,7 +562,7 @@ def clarifier(
# if there is only one tool (Closer), we don't need to decide. It's an LLM answer
llm_decision = DecisionResponse(decision="LLM", reasoning="")
if llm_decision.decision == "LLM" and research_type != ResearchType.DEEP:
if llm_decision.decision == "LLM":
write_custom_event(
current_step_nr,
@@ -644,14 +654,11 @@ def clarifier(
if context_llm_docs:
persona = graph_config.inputs.persona
if persona is not None:
prompt_config = PromptConfig.from_model(
persona, db_session=graph_config.persistence.db_session
)
prompt_config = PromptConfig.from_model(persona)
else:
prompt_config = PromptConfig(
default_behavior_system_prompt=assistant_system_prompt,
custom_instructions=None,
reminder="",
system_prompt=assistant_system_prompt,
task_prompt="",
datetime_aware=True,
)
@@ -680,7 +687,7 @@ def clarifier(
@traced(name="clarifier stream and process", type="llm")
def stream_and_process() -> BasicSearchProcessedStreamResults:
stream = graph_config.tooling.primary_llm.stream_langchain(
stream = graph_config.tooling.primary_llm.stream(
prompt=create_question_prompt(
cast(str, system_prompt_to_use),
cast(str, user_prompt_to_use),
@@ -695,58 +702,55 @@ def clarifier(
should_stream_answer=True,
writer=writer,
ind=0,
search_results=context_llm_docs,
final_search_results=context_llm_docs,
displayed_search_results=context_llm_docs,
generate_final_answer=True,
chat_message_id=str(graph_config.persistence.chat_session_id),
)
# Deep research always continues to clarification or search
if research_type != ResearchType.DEEP:
full_response = stream_and_process()
if len(full_response.ai_message_chunk.tool_calls) == 0:
full_response = stream_and_process()
if len(full_response.ai_message_chunk.tool_calls) == 0:
if isinstance(full_response.full_answer, str):
full_answer = (
normalize_square_bracket_citations_to_double_with_links(
full_response.full_answer
)
)
else:
full_answer = None
# Persist final documents and derive citations when using in-context docs
final_documents_db, citations_map = (
_persist_final_docs_and_citations(
db_session=db_session,
context_llm_docs=context_llm_docs,
full_answer=full_answer,
if isinstance(full_response.full_answer, str):
full_answer = (
normalize_square_bracket_citations_to_double_with_links(
full_response.full_answer
)
)
else:
full_answer = None
update_db_session_with_messages(
db_session=db_session,
chat_message_id=message_id,
chat_session_id=graph_config.persistence.chat_session_id,
is_agentic=graph_config.behavior.use_agentic_search,
message=full_answer,
token_count=len(llm_tokenizer.encode(full_answer or "")),
citations=citations_map,
final_documents=final_documents_db or None,
update_parent_message=True,
research_answer_purpose=ResearchAnswerPurpose.ANSWER,
)
# Persist final documents and derive citations when using in-context docs
final_documents_db, citations_map = _persist_final_docs_and_citations(
db_session=db_session,
context_llm_docs=context_llm_docs,
full_answer=full_answer,
)
db_session.commit()
update_db_session_with_messages(
db_session=db_session,
chat_message_id=message_id,
chat_session_id=graph_config.persistence.chat_session_id,
is_agentic=graph_config.behavior.use_agentic_search,
message=full_answer,
token_count=len(llm_tokenizer.encode(full_answer or "")),
citations=citations_map,
final_documents=final_documents_db or None,
update_parent_message=True,
research_answer_purpose=ResearchAnswerPurpose.ANSWER,
)
return OrchestrationSetup(
original_question=original_question,
chat_history_string="",
tools_used=[DRPath.END.value],
query_list=[],
available_tools=available_tools,
assistant_system_prompt=assistant_system_prompt,
assistant_task_prompt=assistant_task_prompt,
)
db_session.commit()
return OrchestrationSetup(
original_question=original_question,
chat_history_string="",
tools_used=[DRPath.END.value],
query_list=[],
available_tools=available_tools,
assistant_system_prompt=assistant_system_prompt,
assistant_task_prompt=assistant_task_prompt,
)
# Continue, as external knowledge is required.

View File

@@ -41,16 +41,18 @@ def process_llm_stream(
should_stream_answer: bool,
writer: StreamWriter,
ind: int,
search_results: list[LlmDoc] | None = None,
final_search_results: list[LlmDoc] | None = None,
displayed_search_results: list[LlmDoc] | None = None,
generate_final_answer: bool = False,
chat_message_id: str | None = None,
) -> BasicSearchProcessedStreamResults:
tool_call_chunk = AIMessageChunk(content="")
if search_results:
if final_search_results and displayed_search_results:
answer_handler: AnswerResponseHandler = CitationResponseHandler(
context_docs=search_results,
doc_id_to_rank_map=map_document_id_order(search_results),
context_docs=final_search_results,
final_doc_id_to_rank_map=map_document_id_order(final_search_results),
display_doc_id_to_rank_map=map_document_id_order(displayed_search_results),
)
else:
answer_handler = PassThroughAnswerResponseHandler()
@@ -76,7 +78,7 @@ def process_llm_stream(
):
tool_call_chunk += message # type: ignore
elif should_stream_answer:
for response_part in answer_handler.handle_response_part(message):
for response_part in answer_handler.handle_response_part(message, []):
# only stream out answer parts
if (
@@ -92,7 +94,7 @@ def process_llm_stream(
if not start_final_answer_streaming_set:
# Convert LlmDocs to SavedSearchDocs
saved_search_docs = saved_search_docs_from_llm_docs(
search_results
final_search_results
)
write_custom_event(
ind,

View File

@@ -30,7 +30,6 @@ from onyx.db.connector import DocumentSource
from onyx.db.engine.sql_engine import get_session_with_current_tenant
from onyx.prompts.dr_prompts import BASE_SEARCH_PROCESSING_PROMPT
from onyx.prompts.dr_prompts import INTERNAL_SEARCH_PROMPTS
from onyx.secondary_llm_flows.source_filter import strings_to_document_sources
from onyx.server.query_and_chat.streaming_models import SearchToolDelta
from onyx.tools.models import SearchToolOverrideKwargs
from onyx.tools.tool_implementations.search.search_tool import (
@@ -129,11 +128,10 @@ def basic_search(
if re.match(date_pattern, implied_start_date):
implied_time_filter = datetime.strptime(implied_start_date, "%Y-%m-%d")
specified_source_types: list[DocumentSource] | None = (
strings_to_document_sources(search_processing.specified_source_types)
if search_processing.specified_source_types
else None
)
specified_source_types: list[DocumentSource] | None = [
DocumentSource(source_type)
for source_type in search_processing.specified_source_types
]
if specified_source_types is not None and len(specified_source_types) == 0:
specified_source_types = None

View File

@@ -66,7 +66,7 @@ def custom_tool_act(
base_question=base_question,
tool_description=custom_tool_info.description,
)
tool_calling_msg = graph_config.tooling.primary_llm.invoke_langchain(
tool_calling_msg = graph_config.tooling.primary_llm.invoke(
tool_use_prompt,
tools=[custom_tool.tool_definition()],
tool_choice="required",
@@ -125,7 +125,7 @@ def custom_tool_act(
query=branch_query, base_question=base_question, tool_response=tool_str
)
answer_string = str(
graph_config.tooling.primary_llm.invoke_langchain(
graph_config.tooling.primary_llm.invoke(
tool_summary_prompt, timeout_override=TF_DR_TIMEOUT_SHORT
).content
).strip()

View File

@@ -65,7 +65,7 @@ def generic_internal_tool_act(
base_question=base_question,
tool_description=generic_internal_tool_info.description,
)
tool_calling_msg = graph_config.tooling.primary_llm.invoke_langchain(
tool_calling_msg = graph_config.tooling.primary_llm.invoke(
tool_use_prompt,
tools=[generic_internal_tool.tool_definition()],
tool_choice="required",
@@ -113,7 +113,7 @@ def generic_internal_tool_act(
query=branch_query, base_question=base_question, tool_response=tool_str
)
answer_string = str(
graph_config.tooling.primary_llm.invoke_langchain(
graph_config.tooling.primary_llm.invoke(
tool_summary_prompt, timeout_override=TF_DR_TIMEOUT_SHORT
).content
).strip()

View File

@@ -117,8 +117,10 @@ def image_generation(
# save images to file store
file_ids = save_files(
urls=[],
base64_files=[img.image_data for img in image_generation_responses],
urls=[img.url for img in image_generation_responses if img.url],
base64_files=[
img.image_data for img in image_generation_responses if img.image_data
],
)
final_generated_images = [

View File

@@ -1,5 +1,3 @@
from collections.abc import Sequence
from exa_py import Exa
from exa_py.api import HighlightsContentsOptions
@@ -49,9 +47,9 @@ class ExaClient(WebSearchProvider):
]
@retry_builder(tries=3, delay=1, backoff=2)
def contents(self, urls: Sequence[str]) -> list[WebContent]:
def contents(self, urls: list[str]) -> list[WebContent]:
response = self.exa.get_contents(
urls=list(urls),
urls=urls,
text=True,
livecrawl="preferred",
)

View File

@@ -1,5 +1,4 @@
import json
from collections.abc import Sequence
from concurrent.futures import ThreadPoolExecutor
import requests
@@ -56,7 +55,7 @@ class SerperClient(WebSearchProvider):
for result in organic_results
]
def contents(self, urls: Sequence[str]) -> list[WebContent]:
def contents(self, urls: list[str]) -> list[WebContent]:
if not urls:
return []

View File

@@ -78,7 +78,7 @@ def web_search(
def _search(search_query: str) -> list[WebSearchResult]:
search_results: list[WebSearchResult] = []
try:
search_results = list(provider.search(search_query))
search_results = provider.search(search_query)
except Exception as e:
logger.error(f"Error performing search: {e}")
return search_results

View File

@@ -21,7 +21,6 @@ from onyx.configs.agent_configs import TF_DR_TIMEOUT_SHORT
from onyx.context.search.models import InferenceSection
from onyx.prompts.dr_prompts import INTERNAL_SEARCH_PROMPTS
from onyx.utils.logger import setup_logger
from onyx.utils.url import normalize_url
logger = setup_logger()
@@ -39,17 +38,12 @@ def is_summarize(
node_start_time = datetime.now()
# build branch iterations from fetch inputs
# Normalize URLs to handle mismatches from query parameters (e.g., ?activeTab=explore)
url_to_raw_document: dict[str, InferenceSection] = {}
for raw_document in state.raw_documents:
normalized_url = normalize_url(raw_document.center_chunk.semantic_identifier)
url_to_raw_document[normalized_url] = raw_document
# Normalize the URLs from branch_questions_to_urls as well
urls = [
normalize_url(url)
for url in state.branch_questions_to_urls[state.branch_question]
]
url_to_raw_document[raw_document.center_chunk.semantic_identifier] = (
raw_document
)
urls = state.branch_questions_to_urls[state.branch_question]
current_iteration = state.iteration_nr
graph_config = cast(GraphConfig, config["metadata"]["config"])
research_type = graph_config.behavior.research_type

View File

@@ -1,13 +1,9 @@
from abc import ABC
from abc import abstractmethod
from collections.abc import Sequence
from datetime import datetime
from enum import Enum
from pydantic import BaseModel
from pydantic import field_validator
from onyx.utils.url import normalize_url
class ProviderType(Enum):
@@ -20,14 +16,9 @@ class ProviderType(Enum):
class WebSearchResult(BaseModel):
title: str
link: str
snippet: str | None = None
author: str | None = None
published_date: datetime | None = None
@field_validator("link")
@classmethod
def normalize_link(cls, v: str) -> str:
return normalize_url(v)
snippet: str | None = None
class WebContent(BaseModel):
@@ -37,17 +28,12 @@ class WebContent(BaseModel):
published_date: datetime | None = None
scrape_successful: bool = True
@field_validator("link")
@classmethod
def normalize_link(cls, v: str) -> str:
return normalize_url(v)
class WebSearchProvider(ABC):
@abstractmethod
def search(self, query: str) -> Sequence[WebSearchResult]:
def search(self, query: str) -> list[WebSearchResult]:
pass
@abstractmethod
def contents(self, urls: Sequence[str]) -> list[WebContent]:
def contents(self, urls: list[str]) -> list[WebContent]:
pass

View File

@@ -4,8 +4,6 @@ from onyx.agents.agent_search.dr.sub_agents.web_search.models import (
from onyx.agents.agent_search.dr.sub_agents.web_search.models import (
WebSearchResult,
)
from onyx.chat.models import DOCUMENT_CITATION_NUMBER_EMPTY_VALUE
from onyx.chat.models import LlmDoc
from onyx.configs.constants import DocumentSource
from onyx.context.search.models import InferenceChunk
from onyx.context.search.models import InferenceSection
@@ -77,23 +75,3 @@ def dummy_inference_section_from_internet_search_result(
chunks=[],
combined_content="",
)
def llm_doc_from_web_content(web_content: WebContent) -> LlmDoc:
"""Create an LlmDoc from WebContent with the INTERNET_SEARCH_DOC_ prefix"""
return LlmDoc(
# TODO: Is this what we want to do for document_id? We're kind of overloading it since it
# should ideally correspond to a document in the database. But I guess if you're calling this
# function you know it won't be in the database.
document_id="INTERNET_SEARCH_DOC_" + web_content.link,
content=truncate_search_result_content(web_content.full_content),
blurb=web_content.link,
semantic_identifier=web_content.link,
source_type=DocumentSource.WEB,
metadata={},
link=web_content.link,
document_citation_number=DOCUMENT_CITATION_NUMBER_EMPTY_VALUE,
updated_at=web_content.published_date,
source_links={},
match_highlights=[],
)

View File

@@ -106,7 +106,7 @@ def extract_ert(
try:
llm_response = run_with_timeout(
KG_ENTITY_EXTRACTION_TIMEOUT,
primary_llm.invoke_langchain,
primary_llm.invoke,
prompt=msg,
timeout_override=15,
max_tokens=300,
@@ -176,7 +176,7 @@ def extract_ert(
try:
llm_response = run_with_timeout(
KG_RELATIONSHIP_EXTRACTION_TIMEOUT,
primary_llm.invoke_langchain,
primary_llm.invoke,
prompt=msg,
timeout_override=15,
max_tokens=300,

View File

@@ -202,7 +202,7 @@ def analyze(
llm_response = run_with_timeout(
KG_STRATEGY_GENERATION_TIMEOUT,
# fast_llm.invoke,
primary_llm.invoke_langchain,
primary_llm.invoke,
prompt=msg,
timeout_override=5,
max_tokens=100,

View File

@@ -169,7 +169,7 @@ def _get_source_documents(
try:
llm_response = run_with_timeout(
KG_SQL_GENERATION_TIMEOUT,
llm.invoke_langchain,
llm.invoke,
prompt=msg,
timeout_override=KG_SQL_GENERATION_TIMEOUT_OVERRIDE,
max_tokens=KG_SQL_GENERATION_MAX_TOKENS,
@@ -321,7 +321,7 @@ def generate_simple_sql(
try:
llm_response = run_with_timeout(
KG_SQL_GENERATION_TIMEOUT,
primary_llm.invoke_langchain,
primary_llm.invoke,
prompt=msg,
timeout_override=KG_SQL_GENERATION_TIMEOUT_OVERRIDE,
max_tokens=KG_SQL_GENERATION_MAX_TOKENS,
@@ -451,7 +451,7 @@ def generate_simple_sql(
try:
llm_response = run_with_timeout(
KG_SQL_GENERATION_TIMEOUT,
primary_llm.invoke_langchain,
primary_llm.invoke,
prompt=msg,
timeout_override=KG_SQL_GENERATION_TIMEOUT_OVERRIDE,
max_tokens=KG_SQL_GENERATION_MAX_TOKENS,

View File

@@ -94,7 +94,7 @@ def construct_deep_search_filters(
try:
llm_response = run_with_timeout(
KG_FILTER_CONSTRUCTION_TIMEOUT,
llm.invoke_langchain,
llm.invoke,
prompt=msg,
timeout_override=15,
max_tokens=1400,

View File

@@ -137,7 +137,7 @@ def process_individual_deep_search(
try:
llm_response = run_with_timeout(
KG_OBJECT_SOURCE_RESEARCH_TIMEOUT,
primary_llm.invoke_langchain,
primary_llm.invoke,
prompt=msg,
timeout_override=KG_OBJECT_SOURCE_RESEARCH_TIMEOUT,
max_tokens=300,

View File

@@ -127,7 +127,7 @@ def filtered_search(
try:
llm_response = run_with_timeout(
KG_FILTERED_SEARCH_TIMEOUT,
llm.invoke_langchain,
llm.invoke,
prompt=msg,
timeout_override=30,
max_tokens=300,

View File

@@ -1,7 +1,6 @@
from uuid import UUID
from pydantic import BaseModel
from pydantic import ConfigDict
from sqlalchemy.orm import Session
from onyx.agents.agent_search.dr.enums import ResearchType
@@ -26,7 +25,8 @@ class GraphInputs(BaseModel):
structured_response_format: dict | None = None
project_instructions: str | None = None
model_config = ConfigDict(arbitrary_types_allowed=True)
class Config:
arbitrary_types_allowed = True
class GraphTooling(BaseModel):
@@ -41,7 +41,8 @@ class GraphTooling(BaseModel):
force_use_tool: ForceUseTool
using_tool_calling_llm: bool = False
model_config = ConfigDict(arbitrary_types_allowed=True)
class Config:
arbitrary_types_allowed = True
class GraphPersistence(BaseModel):
@@ -56,7 +57,8 @@ class GraphPersistence(BaseModel):
# message were flushed to; only needed for agentic search
db_session: Session
model_config = ConfigDict(arbitrary_types_allowed=True)
class Config:
arbitrary_types_allowed = True
class GraphSearchConfig(BaseModel):
@@ -85,4 +87,5 @@ class GraphConfig(BaseModel):
# Only needed for agentic search
persistence: GraphPersistence
model_config = ConfigDict(arbitrary_types_allowed=True)
class Config:
arbitrary_types_allowed = True

View File

@@ -1,5 +1,4 @@
from pydantic import BaseModel
from pydantic import ConfigDict
from onyx.chat.prompt_builder.schemas import PromptSnapshot
from onyx.tools.message import ToolCallSummary
@@ -39,7 +38,8 @@ class ToolChoice(BaseModel):
id: str | None
search_tool_override_kwargs: SearchToolOverrideKwargs = SearchToolOverrideKwargs()
model_config = ConfigDict(arbitrary_types_allowed=True)
class Config:
arbitrary_types_allowed = True
class ToolChoiceUpdate(BaseModel):

View File

@@ -82,7 +82,7 @@ def trim_prompt_piece(config: LLMConfig, prompt_piece: str, reserved_str: str) -
def build_history_prompt(config: GraphConfig, question: str) -> str:
prompt_builder = config.inputs.prompt_builder
persona_base = get_persona_agent_prompt_expressions(
config.inputs.persona, db_session=config.persistence.db_session
config.inputs.persona
).base_prompt
if prompt_builder is None:
@@ -126,9 +126,7 @@ def build_history_prompt(config: GraphConfig, question: str) -> str:
def get_prompt_enrichment_components(
config: GraphConfig,
) -> AgentPromptEnrichmentComponents:
persona_prompts = get_persona_agent_prompt_expressions(
config.inputs.persona, db_session=config.persistence.db_session
)
persona_prompts = get_persona_agent_prompt_expressions(config.inputs.persona)
history = build_history_prompt(config, config.inputs.prompt_builder.raw_user_query)

View File

@@ -76,7 +76,7 @@ def stream_llm_answer(
else:
citation_processor = None
for message in llm.stream_langchain(
for message in llm.stream(
prompt,
timeout_override=timeout_override,
max_tokens=max_tokens,
@@ -156,7 +156,7 @@ def invoke_llm_json(
) and supports_response_schema(llm.config.model_name, llm.config.model_provider)
response_content = str(
llm.invoke_langchain(
llm.invoke(
prompt,
tools=tools,
tool_choice=tool_choice,
@@ -224,7 +224,7 @@ def get_answer_from_llm(
else:
llm_response = run_with_timeout(
timeout,
llm.invoke_langchain,
llm.invoke,
prompt=msg,
timeout_override=timeout_override,
max_tokens=max_tokens,

View File

@@ -10,7 +10,6 @@ from typing import TypedDict
from langchain_core.messages import BaseMessage
from langgraph.types import StreamWriter
from sqlalchemy.orm import Session
from onyx.agents.agent_search.shared_graph_utils.models import BaseMessage_Content
from onyx.agents.agent_search.shared_graph_utils.models import (
@@ -135,24 +134,18 @@ def format_entity_term_extraction(
def get_persona_agent_prompt_expressions(
persona: Persona | None,
db_session: Session,
) -> PersonaPromptExpressions:
if persona is None:
return PersonaPromptExpressions(
contextualized_prompt=ASSISTANT_SYSTEM_PROMPT_DEFAULT, base_prompt=""
)
# Pull custom instructions if they exist for backwards compatibility
prompt_config = PromptConfig.from_model(persona, db_session=db_session)
system_prompt = (
prompt_config.custom_instructions
or prompt_config.default_behavior_system_prompt
)
# Prompts are now embedded directly on the Persona model
prompt_config = PromptConfig.from_model(persona)
datetime_aware_system_prompt = handle_onyx_date_awareness(
prompt_str=system_prompt,
prompt_str=prompt_config.system_prompt,
prompt_config=prompt_config,
add_additional_info_if_no_tag=bool(persona and persona.datetime_aware),
add_additional_info_if_no_tag=persona.datetime_aware,
)
return PersonaPromptExpressions(
@@ -275,7 +268,7 @@ def summarize_history(
try:
history_response = run_with_timeout(
AGENT_TIMEOUT_LLM_HISTORY_SUMMARY_GENERATION,
llm.invoke_langchain,
llm.invoke,
history_context_prompt,
timeout_override=AGENT_TIMEOUT_CONNECT_LLM_HISTORY_SUMMARY_GENERATION,
max_tokens=AGENT_MAX_TOKENS_HISTORY_SUMMARY,

View File

@@ -2,20 +2,29 @@ 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 onyx.auth.constants import API_KEY_LENGTH
from onyx.auth.constants import API_KEY_PREFIX
from onyx.auth.constants import DEPRECATED_API_KEY_PREFIX
from onyx.auth.schemas import UserRole
from onyx.auth.utils import get_hashed_bearer_token_from_request
from onyx.configs.app_configs import API_KEY_HASH_ROUNDS
from shared_configs.configs import MULTI_TENANT
_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-Onyx-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
@@ -28,10 +37,34 @@ class ApiKeyDescriptor(BaseModel):
def generate_api_key(tenant_id: str | None = None) -> str:
if not MULTI_TENANT or not tenant_id:
return API_KEY_PREFIX + secrets.token_urlsafe(API_KEY_LENGTH)
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_LENGTH)}"
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:
@@ -41,30 +74,30 @@ def _deprecated_hash_api_key(api_key: str) -> str:
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):
if api_key.startswith(_API_KEY_PREFIX):
return hashlib.sha256(api_key.encode("utf-8")).hexdigest()
if api_key.startswith(DEPRECATED_API_KEY_PREFIX):
if api_key.startswith(_DEPRECATED_API_KEY_PREFIX):
return _deprecated_hash_api_key(api_key)
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) :]
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:]
return _API_KEY_PREFIX + api_key[:4] + "********" + api_key[-4:]
def get_hashed_api_key_from_request(request: Request) -> str | None:
"""Extract and hash API key from Authorization header.
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
Accepts both "Bearer <key>" and raw key formats.
"""
return get_hashed_bearer_token_from_request(
request,
valid_prefixes=[API_KEY_PREFIX, DEPRECATED_API_KEY_PREFIX],
hash_fn=hash_api_key,
allow_non_bearer=True, # API keys historically support both formats
)
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)

View File

@@ -1,15 +0,0 @@
"""Authentication constants shared across auth modules."""
# API Key constants
API_KEY_PREFIX = "on_"
DEPRECATED_API_KEY_PREFIX = "dn_"
API_KEY_LENGTH = 192
# PAT constants
PAT_PREFIX = "onyx_pat_"
PAT_LENGTH = 192
# Shared header constants
API_KEY_HEADER_NAME = "Authorization"
API_KEY_HEADER_ALTERNATIVE_NAME = "X-Onyx-Authorization"
BEARER_PREFIX = "Bearer "

View File

@@ -1,155 +0,0 @@
import time
from typing import Any
from urllib.parse import urlencode
from uuid import UUID
import requests
from sqlalchemy.orm import Session
from onyx.db.models import OAuthConfig
from onyx.db.models import OAuthUserToken
from onyx.db.oauth_config import get_user_oauth_token
from onyx.db.oauth_config import upsert_user_oauth_token
from onyx.utils.logger import setup_logger
logger = setup_logger()
class OAuthTokenManager:
"""Manages OAuth token retrieval, refresh, and validation"""
def __init__(self, oauth_config: OAuthConfig, user_id: UUID, db_session: Session):
self.oauth_config = oauth_config
self.user_id = user_id
self.db_session = db_session
def get_valid_access_token(self) -> str | None:
"""Get valid access token, refreshing if necessary"""
user_token = get_user_oauth_token(
self.oauth_config.id, self.user_id, self.db_session
)
if not user_token:
return None
token_data = user_token.token_data
# Check if token is expired
if OAuthTokenManager.is_token_expired(token_data):
# Try to refresh if we have a refresh token
if "refresh_token" in token_data:
try:
return self.refresh_token(user_token)
except Exception as e:
logger.warning(f"Failed to refresh token: {e}")
return None
else:
return None
return token_data.get("access_token")
def refresh_token(self, user_token: OAuthUserToken) -> str:
"""Refresh access token using refresh token"""
token_data = user_token.token_data
response = requests.post(
self.oauth_config.token_url,
data={
"grant_type": "refresh_token",
"refresh_token": token_data["refresh_token"],
"client_id": self.oauth_config.client_id,
"client_secret": self.oauth_config.client_secret,
},
headers={"Accept": "application/json"},
)
response.raise_for_status()
new_token_data = response.json()
# Calculate expires_at if expires_in is present
if "expires_in" in new_token_data:
new_token_data["expires_at"] = (
int(time.time()) + new_token_data["expires_in"]
)
# Preserve refresh_token if not returned (some providers don't return it)
if "refresh_token" not in new_token_data and "refresh_token" in token_data:
new_token_data["refresh_token"] = token_data["refresh_token"]
# Update token in DB
upsert_user_oauth_token(
self.oauth_config.id,
self.user_id,
new_token_data,
self.db_session,
)
return new_token_data["access_token"]
@classmethod
def token_expiration_time(cls, token_data: dict[str, Any]) -> int | None:
"""Get the token expiration time"""
expires_at = token_data.get("expires_at")
if not expires_at:
return None
return expires_at
@classmethod
def is_token_expired(cls, token_data: dict[str, Any]) -> bool:
"""Check if token is expired (with 60 second buffer)"""
expires_at = cls.token_expiration_time(token_data)
if not expires_at:
return False # No expiration data, assume valid
# Add 60 second buffer to avoid race conditions
return int(time.time()) + 60 >= expires_at
def exchange_code_for_token(self, code: str, redirect_uri: str) -> dict[str, Any]:
"""Exchange authorization code for access token"""
response = requests.post(
self.oauth_config.token_url,
data={
"grant_type": "authorization_code",
"code": code,
"client_id": self.oauth_config.client_id,
"client_secret": self.oauth_config.client_secret,
"redirect_uri": redirect_uri,
},
headers={"Accept": "application/json"},
)
response.raise_for_status()
token_data = response.json()
# Calculate expires_at if expires_in is present
if "expires_in" in token_data:
token_data["expires_at"] = int(time.time()) + token_data["expires_in"]
return token_data
@staticmethod
def build_authorization_url(
oauth_config: OAuthConfig, redirect_uri: str, state: str
) -> str:
"""Build OAuth authorization URL"""
params: dict[str, Any] = {
"client_id": oauth_config.client_id,
"redirect_uri": redirect_uri,
"response_type": "code",
"state": state,
}
# Add scopes if configured
if oauth_config.scopes:
params["scope"] = " ".join(oauth_config.scopes)
# Add any additional provider-specific parameters
if oauth_config.additional_params:
params.update(oauth_config.additional_params)
# Check if URL already has query parameters
separator = "&" if "?" in oauth_config.authorization_url else "?"
return f"{oauth_config.authorization_url}{separator}{urlencode(params)}"

View File

@@ -1,60 +0,0 @@
"""Personal Access Token generation and validation."""
import hashlib
import secrets
from datetime import datetime
from datetime import timedelta
from datetime import timezone
from urllib.parse import quote
from fastapi import Request
from onyx.auth.constants import PAT_LENGTH
from onyx.auth.constants import PAT_PREFIX
from onyx.auth.utils import get_hashed_bearer_token_from_request
from shared_configs.configs import MULTI_TENANT
def generate_pat(tenant_id: str | None = None) -> str:
"""Generate cryptographically secure PAT."""
if MULTI_TENANT and tenant_id:
encoded_tenant = quote(tenant_id)
return f"{PAT_PREFIX}{encoded_tenant}.{secrets.token_urlsafe(PAT_LENGTH)}"
return PAT_PREFIX + secrets.token_urlsafe(PAT_LENGTH)
def hash_pat(token: str) -> str:
"""Hash PAT using SHA256 (no salt needed due to cryptographic randomness)."""
return hashlib.sha256(token.encode("utf-8")).hexdigest()
def build_displayable_pat(token: str) -> str:
"""Create masked display version: show prefix + first 4 random chars, mask middle, show last 4.
Example: onyx_pat_abc1****xyz9
"""
# Show first 12 chars (onyx_pat_ + 4 random chars) and last 4 chars
return f"{token[:12]}****{token[-4:]}"
def get_hashed_pat_from_request(request: Request) -> str | None:
"""Extract and hash PAT from Authorization header.
Only accepts "Bearer <token>" format (unlike API keys which support raw format).
"""
return get_hashed_bearer_token_from_request(
request,
valid_prefixes=[PAT_PREFIX],
hash_fn=hash_pat,
allow_non_bearer=False, # PATs require Bearer prefix
)
def calculate_expiration(days: int | None) -> datetime | None:
"""Calculate expiration at 23:59:59.999999 UTC on the target date. None = no expiration."""
if days is None:
return None
expiry_date = datetime.now(timezone.utc).date() + timedelta(days=days)
return datetime.combine(expiry_date, datetime.max.time()).replace(
tzinfo=timezone.utc
)

View File

@@ -64,7 +64,6 @@ from onyx.auth.email_utils import send_user_verification_email
from onyx.auth.invited_users import get_invited_users
from onyx.auth.invited_users import remove_user_from_invited_users
from onyx.auth.jwt import verify_jwt_token
from onyx.auth.pat import get_hashed_pat_from_request
from onyx.auth.schemas import AuthBackend
from onyx.auth.schemas import UserCreate
from onyx.auth.schemas import UserRole
@@ -110,12 +109,13 @@ from onyx.db.models import AccessToken
from onyx.db.models import OAuthAccount
from onyx.db.models import Persona
from onyx.db.models import User
from onyx.db.pat import fetch_user_for_pat
from onyx.db.saml import get_saml_account
from onyx.db.users import get_user_by_email
from onyx.redis.redis_pool import get_async_redis_connection
from onyx.redis.redis_pool import get_redis_client
from onyx.server.utils import BasicAuthenticationError
from onyx.utils.logger import setup_logger
from onyx.utils.secrets import extract_hashed_cookie
from onyx.utils.telemetry import create_milestone_and_report
from onyx.utils.telemetry import optional_telemetry
from onyx.utils.telemetry import RecordType
@@ -206,10 +206,6 @@ def anonymous_user_enabled(*, tenant_id: str | None = None) -> bool:
def verify_email_is_invited(email: str) -> None:
if AUTH_TYPE in {AuthType.SAML, AuthType.OIDC}:
# SSO providers manage membership; allow JIT provisioning regardless of invites
return
whitelist = get_invited_users()
if not whitelist:
return
@@ -1068,7 +1064,17 @@ async def _check_for_saml_and_jwt(
user: User | None,
async_db_session: AsyncSession,
) -> User | None:
# If user is None, check for JWT in Authorization header
# Check if the user has a session cookie from SAML
if AUTH_TYPE == AuthType.SAML:
saved_cookie = extract_hashed_cookie(request)
if saved_cookie:
saml_account = await get_saml_account(
cookie=saved_cookie, async_db_session=async_db_session
)
user = saml_account.user if saml_account else None
# If user is still None, check for JWT in Authorization header
if user is None and JWT_PUBLIC_KEY_URL is not None:
auth_header = request.headers.get("Authorization")
if auth_header and auth_header.startswith("Bearer "):
@@ -1085,12 +1091,6 @@ async def optional_user(
) -> User | None:
user = await _check_for_saml_and_jwt(request, user, async_db_session)
# check if a PAT is present (before API key)
if user is None:
hashed_pat = get_hashed_pat_from_request(request)
if hashed_pat:
user = await fetch_user_for_pat(hashed_pat, async_db_session)
# check if an API key is present
if user is None:
try:

Some files were not shown because too many files have changed in this diff Show More