Remove test that only works with amdgpu driver #15
Workflow file for this run
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| name: Test Omniwatch | |
| on: [push] | |
| jobs: | |
| test: | |
| name: Test Omniwatch integration | |
| runs-on: ubuntu-22.04 | |
| steps: | |
| - name: Check out repository code | |
| uses: actions/checkout@v4 | |
| - name: Install pytest | |
| run: sudo apt-get install -y python3-pytest | |
| - name: Comment out GPU devices (not available in GitHub) | |
| run: sed -i "/devices:/,+2 s/^/#/" test/docker/slurm/compose.yaml | |
| - name: Disable SMI collector (won't work in GitHub) | |
| run: > | |
| sed -i "s/enable_rocm_smi = True/enable_rocm_smi = False/" \ | |
| test/docker/slurm/omniwatch.slurm | |
| - name: Start containerized environment | |
| run: docker compose -f test/docker/slurm/compose.yaml up -d | |
| - name: Wait for Prometheus | |
| run: timeout 1m bash -c 'until $(curl -o /dev/null --fail -s localhost:9090/metrics); do echo "Waiting for Prometheus..."; sleep 5; done' | |
| - name: Wait for Omniwatch | |
| run: timeout 15m bash -c 'until [[ $(curl -s -g "localhost:9090/api/v1/series?match[]={instance=\"node:8000\"}" | jq ".data|length") != 0 ]]; do echo "Waiting for Omnistat..."; docker compose -f test/docker/slurm/compose.yaml logs | tail; curl -s -g "localhost:9090/api/v1/series?match[]={instance=\"node:8000\"}" | jq ".data|length"; sleep 15; done; docker compose -f test/docker/slurm/compose.yaml logs; curl -s -g "localhost:9090/api/v1/series?match[]={instance=\"node:8000\"}"; curl -s -g "localhost:9090/api/v1/series?match[]={instance=\"node:8000\"}" | jq ".data|length"' | |
| - name: Sleep | |
| run: sleep 60 | |
| - name: Check | |
| run: > | |
| echo "QUERY"; | |
| curl -s -g "localhost:9090/api/v1/series?match[]={instance=\"node:8000\"}"; | |
| echo ""; | |
| echo "LEN"; | |
| curl -s -g "localhost:9090/api/v1/series?match[]={instance=\"node:8000\"}" | jq ".data|length"; | |
| echo ""; | |
| docker exec slurm-node-1 ps aux; | |
| echo "ERROR"; | |
| docker exec slurm-node-1 cat /tmp/gunicorn-error.log; | |
| echo "ACCESS"; | |
| docker exec slurm-node-1 cat /tmp/gunicorn-access.log; | |
| - name: Install test dependencies | |
| run: pip3 install prometheus_api_client | |
| - name: Run tests | |
| run: pytest-3 -v test |