Ayoub Mrini

Done. The remaining failing test `TestOnlyProviderStaleTargetsAreDropped` is due to https://github.com/prometheus/prometheus/issues/13312.

https://github.com/prometheus/prometheus/issues/13312 was fixed in main, now the CI is green again.

How about ``` .PHONY: docker-current-arch docker-current-arch: promu $(PROMU) crossbuild -p linux/$(GOHOSTARCH) $(MAKE) npm_licenses $(MAKE) common-docker-$(GOHOSTARCH) ``` I didn't test that but I think it should run the needed commands. I'm...

Are you still willing to work on this @sbdtu5498? @kartikaysaxena would like to take over.

Thanks to @roidelapluie, we have https://prometheus.io/docs/prometheus/latest/command-line/promtool/ (generated) now. We can enrich it and add links to relevant materials you shared. I think we can reword this issue.

Adding another limiting parameter would also add more complication and confusion (https://github.com/prometheus/prometheus/issues/12646 is suggesting different limits for alerts and recording rules...). How about grouping the alerts/rules per group? Separate the...

Hello @bboreham, Thanks for your input. Yes, I'm aware of Mimir's implementation and how we add offsets for scraping. At first, I wanted us to agree on the best place...

I pushed a new version. It shifts the whole compact operation (not only the compactHead one). I have some questions, see comments above. I ran some benchmarks, and on busy...

> [ ] Look at the mmapHeadChunks starvation. See: https://github.com/prometheus/prometheus/pull/12532#issuecomment-1855622480 Actually, this is due to the fact that we create a new timer channel every time. https://github.com/prometheus/prometheus/blob/69abd6d9f65a6358025461510237366f1d939c70/tsdb/db.go#L980 For a busy...

> Actually, this is due to the fact that we create a new timer channel every time. >Line 980 in [69abd6d](https://github.com/prometheus/prometheus/commit/69abd6d9f65a6358025461510237366f1d939c70) >For a busy DB, compactc is always ready (due...