need k8s.pod.memory.working_set_memory_limit_utilization metric

open-telemetry / opentelemetry-collector-contrib

Contrib repository for the OpenTelemetry Collector

https://opentelemetry.io

Apache License 2.0

3.06k stars 2.36k forks source link

need k8s.pod.memory.working_set_memory_limit_utilization metric #33562

Closed halleystar closed 2 weeks ago

halleystar commented 4 months ago

Component(s)

receiver/kubeletstats

Is your feature request related to a problem? Please describe.

need k8s.pod.memory.working_set_memory_limit_utilization, beacuse k8s pod oom is Obtained through working_set_memory / memory limit。But there is only k8s.pod.memory_limit_utilization.

Describe the solution you'd like

need k8s.pod.memory.working_set_memory_limit_utilization metric

Describe alternatives you've considered

need k8s.pod.memory.working_set_memory_limit_utilization metric

Additional context

No response

github-actions[bot] commented 4 months ago

Pinging code owners:

receiver/kubeletstats: @dmitryax @TylerHelmuth

See Adding Labels via Comments if you do not have permissions to add labels yourself.

ChrsMark commented 4 months ago

Hey @halleystar! Isn't the container's memory_working_set the crucial one here instead of the Pod's one? The OOM killer is triggered based on the container's memory_working_set metric I think, right?

Since we have multiple memory metrics I guess it would make sense to have their *memory*utilization equivalent ones. This one also has some good justification, but I would love other's opinions here as well.

We should also consider discussing it as part of the https://github.com/open-telemetry/semantic-conventions/issues/1032 as well.

halleystar commented 4 months ago

yes! your are right. It should be container's memory_working_set. I think we should have a relation memoryutilization at the same time. Otherwise we should provide container request and limit memory to calculate memory usage to observe whether OOM is about to occur.

github-actions[bot] commented 2 months ago

This issue has been inactive for 60 days. It will be closed in 60 days if there is no activity. To ping code owners by adding a component label, see Adding Labels via Comments, or if you are unsure of which component this issue relates to, please ping @open-telemetry/collector-contrib-triagers. If this issue is still relevant, please ping the code owners or leave a comment explaining why it is still relevant. Otherwise, please close it.

Pinging code owners:

receiver/kubeletstats: @dmitryax @TylerHelmuth @ChrsMark

See Adding Labels via Comments if you do not have permissions to add labels yourself.

github-actions[bot] commented 2 weeks ago

This issue has been closed as inactive because it has been stale for 120 days with no activity.