Maksim Nabokikh

icon indicating an email [email protected]

icon company @deckhouse icon location Germany icon location Technical Product Manager at @deckhouse mostly interested in cloud-native security and observability | Staff Engineer at @palark | Maintainer of @dexidp

Results 316 comments of Maksim Nabokikh

Add request count metric with status code label

@Aneurysm9 fyi

Add request count metric with status code label

From my point of view, it would be better to avoid using updown counter for errors, because you cannot apply rate functions to it. Adding status code label for histograms...

Add request count metric with status code label

@MrAlias @Aneurysm9, just a friendly reminder. Is it ok to add the status code label to the server latency metric?

Add request count metric with status code label

I've just added the status code label, so everything should be good now.

Add request count metric with status code label

Folks, any updates on this one? It would be best if we resolve all concerns about this PR before the next release 🙏

Add request count metric with status code label

I rebased this branch on the main. Just a friendly reminder, still waiting for the feedback 😸

Add the ability to expire internal metrics.

@tobz, I'd like to help execute this plan step by step because [our Kubernetes cluster installation suffers from the problem of metrcs number growth](https://github.com/vectordotdev/vector/issues/11225#issuecomment-1133047767). I have experience doing the same...

Add the ability to expire internal metrics.

I'm about to open a PR, but apparently it will only be a part of the future v0.24 release.

Add the ability to expire internal metrics.

@jszwedko @bruceg I did not manage to go far with it and stuck with the way to detect staleness (and I also lost time to update dependencies, what Bruce've already...

Local dev mode: show some kind of diff and reason why stage rebuild started

+1 to this issue. In Deckhouse, we use werf to build ~150 images. Each time many images are rebuilt because of cache issues, we struggle to investigate the root cause.