Skip to content

Run responses and conversations auth tests with Kubernetes authentication #1

Run responses and conversations auth tests with Kubernetes authentication

Run responses and conversations auth tests with Kubernetes authentication #1

name: Integration Responses & Conversations Auth Tests
run-name: Run responses and conversations auth tests with Kubernetes authentication
on:
push:
branches:
- main
- 'release-[0-9]+.[0-9]+.x'
pull_request:
branches:
- main
- 'release-[0-9]+.[0-9]+.x'
paths:
- 'distributions/**'
- 'src/llama_stack/**'
- '!src/llama_stack_ui/**'
- 'tests/integration/**'
- 'uv.lock'
- 'pyproject.toml'
- 'requirements.txt'
- '.github/workflows/integration-responses-conversations-auth-tests.yml' # This workflow
- 'scripts/integration-responses-conversations-auth-tests.sh'
merge_group:
branches:
- main
- 'release-[0-9]+.[0-9]+.x'
concurrency:
group: ${{ github.workflow }}-${{ github.ref == 'refs/heads/main' && github.run_id || github.ref }}
cancel-in-progress: true
permissions:
contents: read
jobs:
test-matrix:
runs-on: ubuntu-latest
strategy:
matrix:
auth-provider: [oauth2_token]
fail-fast: false # we want to run all tests regardless of failure
steps:
- name: Checkout repository
uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
- name: Install dependencies
uses: ./.github/actions/setup-runner
- name: Install minikube
if: ${{ matrix.auth-provider == 'kubernetes' }}
uses: medyagh/setup-minikube@e9e035a86bbc3caea26a450bd4dbf9d0c453682e # v0.0.21
- name: Start minikube
if: ${{ matrix.auth-provider == 'oauth2_token' }}
run: |
minikube start
kubectl get pods -A
- name: Configure Kube Auth
if: ${{ matrix.auth-provider == 'oauth2_token' }}
run: |
kubectl create namespace llama-stack
for account in llama-stack-auth llama-stack-user1 llama-stack-user2; do
kubectl create serviceaccount $account -n llama-stack
kubectl create token $account -n llama-stack > $account-token
done
- name: Set Kubernetes Config
if: ${{ matrix.auth-provider == 'oauth2_token' }}
run: |
echo "KUBERNETES_API_SERVER_URL=$(kubectl get --raw /.well-known/openid-configuration| jq -r .jwks_uri)" >> $GITHUB_ENV
echo "KUBERNETES_CA_CERT_PATH=$(kubectl config view --minify -o jsonpath='{.clusters[0].cluster.certificate-authority}')" >> $GITHUB_ENV
echo "KUBERNETES_ISSUER=$(kubectl get --raw /.well-known/openid-configuration| jq -r .issuer)" >> $GITHUB_ENV
echo "KUBERNETES_AUDIENCE=$(kubectl create token llama-stack-auth -n llama-stack --duration=1h | cut -d. -f2 | base64 -d | jq -r '.aud[0]')" >> $GITHUB_ENV
echo "TOKEN=$(cat llama-stack-auth-token)" >> $GITHUB_ENV
- name: List response recordings (replay)
if: ${{ matrix.auth-provider == 'oauth2_token' }}
run: |
echo "Response recordings in checkout:"
ls -la tests/integration/responses/recordings/ || true
- name: Set Kube Auth Config and run server
env:
OPENAI_API_KEY: "dummy-key-for-replay-mode"
# Enable replay so the server uses checked-in recordings instead of calling the API
LLAMA_STACK_TEST_INFERENCE_MODE: "replay"
# Pin recording dir to workspace so the server finds tests/integration/*/recordings/
LLAMA_STACK_TEST_RECORDING_DIR: ${{ github.workspace }}/tests/integration/common
# Tell the server to sync test context from client headers so recordings
# are looked up in the correct per-test-directory (e.g. responses/recordings/)
LLAMA_STACK_TEST_STACK_CONFIG_TYPE: "server"
if: ${{ matrix.auth-provider == 'oauth2_token' }}
run: |
run_dir=$(mktemp -d)
cat <<EOF > $run_dir/config.yaml
version: '2'
image_name: kube
distro_name: kube
apis:
- files
- inference
- vector_io
- tool_runtime
- agents
providers:
files:
- provider_id: builtin-files
provider_type: inline::localfs
config:
storage_dir: $run_dir/files
metadata_store:
table_name: files_metadata
backend: sql_default
inference:
- provider_id: openai
provider_type: remote::openai
config:
api_key: \${env.OPENAI_API_KEY}
vector_io:
- provider_id: faiss
provider_type: inline::faiss
config:
persistence:
namespace: vector_io::faiss
backend: kv_default
tool_runtime:
- provider_id: file-search
provider_type: inline::file-search
config: {}
agents:
- provider_id: builtin
provider_type: inline::builtin
config:
persistence:
agent_state:
namespace: agents
backend: kv_default
responses:
table_name: responses
backend: sql_default
max_write_queue_size: 10000
num_writers: 4
storage:
backends:
kv_default:
type: kv_sqlite
db_path: $run_dir/kvstore.db
sql_default:
type: sql_sqlite
db_path: $run_dir/sql_store.db
stores:
metadata:
namespace: registry
backend: kv_default
inference:
table_name: inference_store
backend: sql_default
conversations:
table_name: openai_conversations
backend: sql_default
prompts:
namespace: prompts
backend: kv_default
models:
- model_id: openai/gpt-4o
model_type: llm
provider_id: openai
server:
port: 8321
auth:
access_policy:
- permit:
actions: [read]
resource: model::*
description: Any authenticated user can use configured models (inference, responses)
- permit:
actions: [create]
resource: sql_record::openai_files::*
description: Any authenticated user can create files
- permit:
actions: [read, delete]
resource: sql_record::openai_files::*
when:
- user is owner
description: Users can read and delete their own files
- permit:
actions: [create]
resource: sql_record::openai_conversations::*
description: Any authenticated user can create conversations
- permit:
actions: [read, update, delete]
resource: sql_record::openai_conversations::*
when:
- user is owner
description: Users can access their own conversations
- permit:
actions: [create]
resource: sql_record::conversation_items::*
description: Any authenticated user can create conversation items
- permit:
actions: [read, update, delete]
resource: sql_record::conversation_items::*
when:
- user is owner
description: Users can access items in their own conversations
- permit:
actions: [create]
resource: sql_record::responses::*
description: Any authenticated user can create responses
- permit:
actions: [read, update, delete]
resource: sql_record::responses::*
when:
- user is owner
description: Users can access their own responses
EOF
yq eval '.server.auth.provider_config.type = "${{ matrix.auth-provider }}"' -i $run_dir/config.yaml
yq eval '.server.auth.provider_config.tls_cafile = "${{ env.KUBERNETES_CA_CERT_PATH }}"' -i $run_dir/config.yaml
yq eval '.server.auth.provider_config.issuer = "${{ env.KUBERNETES_ISSUER }}"' -i $run_dir/config.yaml
yq eval '.server.auth.provider_config.audience = "${{ env.KUBERNETES_AUDIENCE }}"' -i $run_dir/config.yaml
yq eval '.server.auth.provider_config.jwks.uri = "${{ env.KUBERNETES_API_SERVER_URL }}"' -i $run_dir/config.yaml
yq eval '.server.auth.provider_config.jwks.token = "${{ env.TOKEN }}"' -i $run_dir/config.yaml
cat $run_dir/config.yaml
# avoid line breaks in the server log, especially because we grep it below.
export LLAMA_STACK_LOG_WIDTH=200
# OPENAI_API_KEY is set via env at step level
nohup uv run llama stack run $run_dir/config.yaml > server.log 2>&1 &
- name: Wait for Llama Stack server to be ready
run: |
echo "Waiting for Llama Stack server..."
for i in {1..30}; do
# Note: /v1/health does not require authentication
if curl -s -L http://localhost:8321/v1/health | grep -q "OK"; then
echo "Llama Stack server is up!"
if grep -q "Enabling authentication with provider: ${{ matrix.auth-provider }}" server.log; then
echo "Llama Stack server is configured to use ${{ matrix.auth-provider }} auth"
exit 0
else
echo "Llama Stack server is not configured to use ${{ matrix.auth-provider }} auth"
cat server.log
exit 1
fi
fi
sleep 1
done
echo "Llama Stack server failed to start"
cat server.log
exit 1
- name: Test auth
run: |
export INFERENCE_MODEL=openai/gpt-4o
# Run the auth tests
./scripts/integration-responses-conversations-auth-tests.sh
- name: Server log on failure
if: failure()
run: |
echo "=== Llama Stack server log (last 500 lines) ==="
tail -500 server.log