How to clean up team_build_events_XXX and pipeline_build_events_XXX


#1

I guess we suffer from this problem too. (mentioned in the release log of 4.0.0) The CPU usage doubled from 50 to almost 100% in the last 30 days and we have a lot of these tables: team_build_events_70, pipeline_build_events_600

Which of the 600 tables can I safely drop?


#2

As far as I understand this, the 70 team_build_events are correct, as we have currently 70 teams. But the pipeline_build_events are way to many.

Can I drop all “pipeline_build_events_xxx” where xxx is not found in the pipelines table?

What should I do with the builds referenced in the pipeline_build_events_xxx ?