Objective
Add the new evaluation-specific metrics and graphs introduced in the referenced PerfStar PRs to the performance dashboard.
References
PerfStar PR:
Requirements
The dashboard should provide a clear, user-friendly view of evaluation performance trends and be suitable for presentation during Evaluation Sync meetings. Users should be able to understand performance changes and regressions at a glance.
The dashboard should include:
- Evaluation time broken down by individual evaluation phases.
- Total evaluation time.
- Total build time
- Time before MSBuild submission (dotnet.cli.process_start_to_msbuild_submission.duration).
- Historical trends showing improvements and regressions over time.
- Percentage change between builds for all key evaluation metrics.
Build Comparison Experience
Provide functionality to select and compare two specific builds. For each selected build, display:
- MSBuild commit SHA and a link to the commit.
- SDK commit SHA and a link to the commit used in the build.
- Key metric values for both builds.
- Percentage improvement or regression between the selected builds.
The comparison view should make it easy to answer questions such as:
- What changed between these two builds?
- Which evaluation metrics improved or regressed?
- What was the percentage improvement for each metric?
Objective
Add the new evaluation-specific metrics and graphs introduced in the referenced PerfStar PRs to the performance dashboard.
References
PerfStar PR:
Requirements
The dashboard should provide a clear, user-friendly view of evaluation performance trends and be suitable for presentation during Evaluation Sync meetings. Users should be able to understand performance changes and regressions at a glance.
The dashboard should include:
Build Comparison Experience
Provide functionality to select and compare two specific builds. For each selected build, display:
The comparison view should make it easy to answer questions such as: