Open oindrillac opened 1 year ago
Currently srcopsmetrics does not seem to save data onto S3 per Pull Request. Rather it seems to save an entire repository's data after a run completes. This can cause a lot of intermediate progress to get dumped if the run fails in between. We are encountering this problem when using srcopsmetrics in a github action workflow which cancels long running jobs https://github.com/redhat-et/time-to-merge-tool/issues/9 https://github.com/aicoe-aiops/ocp-ci-analysis/issues/603#issuecomment-1308958352
srcopsmetrics
The problem is also described here: https://github.com/thoth-station/mi/issues/573
When run in the default S3 mode, enable srcopsmetrics to save entities per PR, or have a mode when enabled, does that.
The alternative that we have considered using is the PyGithub API directly for downloading data https://github.com/redhat-et/time-to-merge-tool/issues/9
Problem statement
Currently
srcopsmetricsdoes not seem to save data onto S3 per Pull Request. Rather it seems to save an entire repository's data after a run completes. This can cause a lot of intermediate progress to get dumped if the run fails in between. We are encountering this problem when using srcopsmetrics in a github action workflow which cancels long running jobs https://github.com/redhat-et/time-to-merge-tool/issues/9 https://github.com/aicoe-aiops/ocp-ci-analysis/issues/603#issuecomment-1308958352The problem is also described here: https://github.com/thoth-station/mi/issues/573
Proposal description
When run in the default S3 mode, enable srcopsmetrics to save entities per PR, or have a mode when enabled, does that.
Alternatives
The alternative that we have considered using is the PyGithub API directly for downloading data https://github.com/redhat-et/time-to-merge-tool/issues/9