Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

enable pprof on the metrics endpoint #1038

Open
jrwren opened this issue Aug 15, 2023 · 3 comments
Open

enable pprof on the metrics endpoint #1038

jrwren opened this issue Aug 15, 2023 · 3 comments

Comments

@jrwren
Copy link

jrwren commented Aug 15, 2023

Problem Statement

Memory usage of supertubes control plane in our clusters is very high. (1924Mi) It would be nice to be able to take a heap dump using pprof.

Proposed Solution

  •   mgr.AddMetricsExtraHandler("/debug/pprof/", http.HandlerFunc(pprof.Index))
    

in main.go around line 143

Alternatives Considered

I've not considered alternatives. I am not an expert and do not know them.

Additional Context

It seems like many other operators do this for good reason.

@jrwren jrwren mentioned this issue Aug 15, 2023
4 tasks
@bartam1
Copy link
Contributor

bartam1 commented Aug 15, 2023

Dear @jrwren !
This is not supertubes control plane.
Please give us more information about the problem.
Thank you!

@jrwren
Copy link
Author

jrwren commented Aug 15, 2023

@bartam1 apologies. I misunderstood this: https://banzaicloud.com/docs/supertubes/overview/

The problem I'm having is that I have a cluster with nearly 4000 secrets of a size large enough that they seem to take a large amount of memory. For all of our in-house developed operators, we've disabled secret caching (or used WatchesMetadata instead of Watches). In one case our operator memory usage fell by ~96% from ~2500MiB to ~50MiB (sorry, I don't recall exact numbers).

I thought supertubes-control-plane was built on koperator. If that isn't the case or if this is the wrong place, then I'll try to find support for supertubes-control-plane elsewhere. Thanks.

@bartam1
Copy link
Contributor

bartam1 commented Aug 15, 2023

@bartam1 apologies. I misunderstood this: https://banzaicloud.com/docs/supertubes/overview/

The problem I'm having is that I have a cluster with nearly 4000 secrets of a size large enough that they seem to take a large amount of memory. For all of our in-house developed operators, we've disabled secret caching (or used WatchesMetadata instead of Watches). In one case our operator memory usage fell by ~96% from ~2500MiB to ~50MiB (sorry, I don't recall exact numbers).

I thought supertubes-control-plane was built on koperator. If that isn't the case or if this is the wrong place, then I'll try to find support for supertubes-control-plane elsewhere. Thanks.

Supertubes control plane is a different operator (Its purpose is to install the Supertubes product's components) . The Supertube product is using Koperator to manage Kafka clusters.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants