Migrate to sitrep mechanism for T5X and PAXML MGMN tests #401
There are no files selected for viewing
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
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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,134 @@ | ||
| name: ~Generate sitrep for Multi-Node Multi-GPU tests | ||
|
|
||
| on: | ||
| workflow_call: | ||
| inputs: | ||
| BADGE_FILENAME: | ||
| type: string | ||
| description: 'Name of the endpoint JSON file for shields.io badge' | ||
| required: true | ||
| ARTIFACT_NAME: | ||
| type: string | ||
| description: 'Name of the artifact zip file' | ||
| required: true | ||
| FW_NAME: | ||
| type: string | ||
| description: 'Name of the framework being used' | ||
| required: true | ||
| outputs: | ||
| STATUS: | ||
| description: 'Summary of all tests run for the workflow. Set to "success" when all metrics per job and all jobs pass, whereas a single metric failure or job error sets the status to "failure"' | ||
| value: ${{ jobs.sitrep.outputs.STATUS }} | ||
|
|
||
| jobs: | ||
| sitrep: | ||
| runs-on: ubuntu-22.04 | ||
| outputs: | ||
| STATUS: ${{ steps.gen-sitrep.outputs.STATUS }} | ||
| steps: | ||
| - name: Check out repository | ||
| uses: actions/checkout@v3 | ||
|
|
||
| - name: Download all artifacts from the previous jobs | ||
| uses: actions/download-artifact@v3 | ||
|
|
||
| - name: Write exit status summary | ||
| id: exit-status | ||
| shell: bash -x -e {0} | ||
| run: | | ||
| EXIT_STATUSES="${{ inputs.FW_NAME }}-${GITHUB_RUN_ID}-*/*-status.json" | ||
| EXIT_STATUS_SUMMARY_FILE="exit_status_summary.json" | ||
| echo -e "\n\n## ${{ inputs.FW_NAME }} MGMN+SPMD Test Status" >> $EXIT_STATUS_SUMMARY_FILE | ||
| cat <<EOF >>$EXIT_STATUS_SUMMARY_FILE | ||
| | Test Case | State | Exit Code | | ||
| | --- | --- | --- | | ||
| EOF | ||
|
|
||
| for i in $EXIT_STATUSES; do | ||
| # Files are named <FW_NAME>-<GHID>-<NAME>/<NAME>-status.json | ||
| echo "| $(echo $i | cut -d/ -f1 | awk -F- '{print $NF}') | $(jq -r .state $i) | $(jq -r .exitcode $i)" | ||
| done | tee -a $EXIT_STATUS_SUMMARY_FILE | ||
|
|
||
| echo "Test statuses:" | ||
| jq -rc 'input_filename,.' $EXIT_STATUSES | ||
|
|
||
| echo "EXIT_STATUS_SUMMARY_FILE=$EXIT_STATUS_SUMMARY_FILE" >> ${GITHUB_OUTPUT} | ||
|
|
||
| - name: Write metrics summary | ||
| id: metrics | ||
| shell: bash -x -e {0} | ||
| run: | | ||
| METRICS_SUMMARY_FILE="metrics_summary.json" | ||
| echo -e "\n\n## ${{ inputs.FW_NAME }} MGMN Test Metrics" >> $METRICS_SUMMARY_FILE | ||
| for i in metrics-test-log/*_metrics.json; do | ||
| echo $i | cut -d'.' -f1 | ||
| echo '```json' | ||
| jq . $i | ||
| echo '```' | ||
| done | tee -a $METRICS_SUMMARY_FILE | ||
|
|
||
| echo "METRICS_SUMMARY_FILE=$METRICS_SUMMARY_FILE" >> ${GITHUB_OUTPUT} | ||
|
|
||
| - name: Generate sitrep | ||
| id: gen-sitrep | ||
| shell: bash -x -e {0} | ||
| run: | | ||
| source .github/workflows/scripts/to_json.sh | ||
|
|
||
| EXIT_STATUSES="${{ inputs.FW_NAME }}-${GITHUB_RUN_ID}-*/*-status.json" | ||
|
|
||
| passed_tests=$(jq -r '. | select ((.state == "COMPLETED") and (.exitcode == "0")) | .state' $EXIT_STATUSES | wc -l) | ||
| failed_tests=$(jq -r '. | select ((.state != "COMPLETED") or (.exitcode != "0")) | .state' $EXIT_STATUSES | wc -l) | ||
| total_tests=$(ls $EXIT_STATUSES | wc -l) | ||
|
|
||
| METRICS_LOG=metrics-test-log/report.jsonl | ||
| all_outcomes() { | ||
| cat $METRICS_LOG | jq -r '. | select((.["$report_type"] == "TestReport") and (.when == "call")) | .outcome' | ||
| } | ||
| cnt_type() { | ||
| cat $METRICS_LOG | jq '. | select((.["$report_type"] == "TestReport") and (.when == "call") and (.outcome | contains("'${1}'"))) | .outcome' | wc -l | ||
| } | ||
| pytest_failed_tests=$(cnt_type failed) | ||
| pytest_passed_tests=$(cnt_type passed) | ||
| pytest_total_tests=$(all_outcomes | wc -l) | ||
|
|
||
| if ([[ $failed_tests -eq 0 ]] && [[ $total_tests -gt 0 ]] && \ | ||
| [[ $pytest_failed_tests -eq 0 ]] && [[ $pytest_total_tests -gt 0 ]]); then | ||
| status=success | ||
| badge_color=brightgreen | ||
| elif [[ $passed_tests -eq 0 ]] || [[ $pytest_passed_tests -eq 0 ]]; then | ||
| status=failure | ||
| badge_color=red | ||
| else | ||
| status=failure | ||
| badge_color=yellow | ||
| fi | ||
| badge_message="${passed_tests}/${total_tests} jobs | ${pytest_passed_tests}/${pytest_total_tests} metrics" | ||
|
|
||
| badge_label='Upstream Tests' | ||
| summary="# ${{ inputs.FW_NAME }} MGMN Test: $badge_message" | ||
| summary+=`cat ${{ steps.exit-status.outputs.EXIT_STATUS_SUMMARY_FILE }}` | ||
| summary+=`cat ${{ steps.metrics.outputs.METRICS_SUMMARY_FILE }}` | ||
|
|
||
| to_json \ | ||
| summary \ | ||
| total_tests passed_tests failed_tests \ | ||
| badge_label badge_color badge_message \ | ||
| > sitrep.json | ||
|
|
||
| schemaVersion=1 \ | ||
| label="${badge_label}" \ | ||
| message="${badge_message}" \ | ||
| color="${badge_color}" \ | ||
| to_json schemaVersion label message color \ | ||
| > ${{ inputs.BADGE_FILENAME }} | ||
|
|
||
| echo "STATUS='${status}'" >> ${GITHUB_OUTPUT} | ||
|
|
||
| - name: Upload artifacts | ||
| uses: actions/upload-artifact@v3 | ||
| with: | ||
| name: ${{ inputs.ARTIFACT_NAME }} | ||
| path: | | ||
| sitrep.json | ||
| ${{ inputs.BADGE_FILENAME }} |
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
Oops, something went wrong.
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is similar to our current practice, which uses a final postprocessing job to determine whether the overall MGMN test suite succeeds.
It has the problem that error feedback is delayed and made it difficult to investigate which individual tests failed.
For the new sitrep reporting system, let's do the following:
continue-on-erroroption can be helpful here.This helps to localize the error status for easier debugging and tracking.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
If a single job fails, what should be the outcome of the entire workflow? Also, I think we still need this output for downstream jobs like
triageorpublish containersince those depend on this overall outcome.There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
For the MGMN usecase, the tests are submitted to slurm via SSH. To mark an individual job as failure I can inspect the exit code and mark it accordingly. We would still need the overall status though, if there's any other recommended way to get this output, I'm happy to incorporate that.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Makes sense. How about we make individual jobs fail while also let the overall status be the Boolean AND of the individual job states?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Actually, a single job failure is already indicated. (ref: https://github.com/NVIDIA/JAX-Toolbox/actions/runs/7053172928/job/19199718748). The STATUS has to be derived not only from the job result, but also from its metrics so it is not possible to set status just based on the job states.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
At face value, the specific error message from the job that you provided looks coincidental. I understand that it may be the result of the true underlying error, but the message itself does not clearly indicate that. Could we improve on it?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Also could you please add a natural language description of the logic regarding
so that it serves as a form of documentation?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Not sure if there's a way to provide a more informative error based on the failure, might have to look at it outside the scope of this PR.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Done in 40aebf5