frame-omni-bencher short checks #5268
Conversation
| - '${{ steps.current-file.outputs.current-workflow-file }}' | ||
| rust: | ||
| - '**/*' | ||
| - '!.github/**/*' |
There was a problem hiding this comment.
Would it make sense to use forklift caching to decide if something changed?
There was a problem hiding this comment.
forklift will know about changes only during cargo build and this action is for skipping entire workflows if no rust code has changed
There was a problem hiding this comment.
Yea but there could be a forklift --did-something-change? Trying to manually re-list all files that could trigger a build seems more fragile. But maybe its not the objective of forklift to offer this...
Co-authored-by: Oliver Tale-Yazdi <[email protected]>
711d91a
|
cc @bkontur |
|
@bkontur yep, that's correct, this is not enabled yet. If you want to add it to the current CI you need to change gitlab ci files. |
|
@bkontur i think if you could fix 2-3 failing frame omni bencher checks with merging those fixes, may be you could just remove the gitlab part of it and remove this condition - then it will be fully migrated to GHA and frame omni bencher by default You can test those by adding |
|
@bkontur please don't add |
@mordamax don't worry, thanks for that conditions (exactly, what I was looking for: a47aff4), I will continue here with fixing and when everything works we will remove gitlab CI pipelines |
- Part of paritytech/ci_cd#1006 - Closes: paritytech/ci_cd#1010 - Related: paritytech#4405 - Possibly affecting how frame-omni-bencher works on different runtimes: paritytech#5083 Currently works in parallel with gitlab short benchmarks. Triggered only by adding `GHA-migration` label to assure smooth transition (kind of feature-flag). Later when tested on random PRs we'll remove the gitlab and turn on by default these tests --------- Co-authored-by: Oliver Tale-Yazdi <[email protected]>
- Part of https://github.com/paritytech/ci_cd/issues/1006 - Closes: https://github.com/paritytech/ci_cd/issues/1010 - Related: paritytech#4405 - Possibly affecting how frame-omni-bencher works on different runtimes: paritytech#5083 Currently works in parallel with gitlab short benchmarks. Triggered only by adding `GHA-migration` label to assure smooth transition (kind of feature-flag). Later when tested on random PRs we'll remove the gitlab and turn on by default these tests --------- Co-authored-by: Oliver Tale-Yazdi <[email protected]>
Part of https://github.com/paritytech/ci_cd/issues/1006
Closes: https://github.com/paritytech/ci_cd/issues/1010
Related: CI short benchmarking with
frame-omni-bencher#4405Possibly affecting how frame-omni-bencher works on different runtimes:
frame-omni-bencherfix for--repeat> 1 #5083Currently works in parallel with gitlab short benchmarks.
Triggered only by adding
GHA-migrationlabel to assure smooth transition (kind of feature-flag).Later when tested on random PRs we'll remove the gitlab and turn on by default these tests