[kube-prometheus-stack] feature(kubelet): Add trackTimestampsStaleness

[Breee]

What this PR does / why we need it

This PR enables trackTimestampsStaleness for the cadvisor path to avoid stale Metrics provided by the cadvisor.

Which issue this PR fixes

(optional, in fixes #<issue number>(, fixes #<issue_number>, ...) format, will close that issue when PR gets merged)

• fixes #

Special notes for your reviewer

@SuperQ suggested this and might want to add comments.
slack conversation: https://cloud-native.slack.com/archives/C167KFM6C/p1733917170837469

Checklist

• DCO signed
• Chart Version bumped
• Title of the PR starts with chart name (e.g. [prometheus-couchdb-exporter])

[QuentinBisson]

LGTM

[SuperQ]

Sorry, but this needed a lot more work in order to be a complete change.

• The scrape interval needs to be adjusted to 10s in order to handle the inconsistent internal intervals of cAdvisor.
• This was only supposed to be applied to one endpoint, the cAdvisor metrics. Not all kubelet metrics endpoints.

@QuentinBisson Please take more thought and care to reviewing these kinds of changes.

[Breee]

Let's fix it then, i'm sorry that i missed the required changes. I also forgot to Mark the PR as draft because I wanted to wait for your Feedback

…

On Sat, Dec 14, 2024, 09:56 Ben Kochie ***@***.***> wrote: Sorry, but this needed a lot more work in order to be a complete change. - The scrape interval needs to be adjusted to 10s in order to handle the inconsistent internal intervals of cAdvisor. - This was only supposed to be applied to one endpoint, the cAdvisor metrics. Not all kubelet metrics endpoints. @QuentinBisson <https://github.com/QuentinBisson> Please take more thought and care to reviewing these kinds of changes. — Reply to this email directly, view it on GitHub <#5050 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AC3JPMIXYFMAOFNAXNF7CV32FPXDBAVCNFSM6AAAAABTPIOKZKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKNBTGAYDSNJTGQ> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[Breee]

@QuentinBisson would you be so Kind and revert the PR, so we can get it right? we still should bump the Version then for people that use automatic dependency Updates. Sorry for the trouble.

…

On Sat, Dec 14, 2024, 10:21 Julian Loeffler ***@***.***> wrote: Let's fix it then, i'm sorry that i missed the required changes. I also forgot to Mark the PR as draft because I wanted to wait for your Feedback On Sat, Dec 14, 2024, 09:56 Ben Kochie ***@***.***> wrote: > Sorry, but this needed a lot more work in order to be a complete change. > > - The scrape interval needs to be adjusted to 10s in order to handle > the inconsistent internal intervals of cAdvisor. > - This was only supposed to be applied to one endpoint, the cAdvisor > metrics. Not all kubelet metrics endpoints. > > @QuentinBisson <https://github.com/QuentinBisson> Please take more > thought and care to reviewing these kinds of changes. > > — > Reply to this email directly, view it on GitHub > <#5050 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AC3JPMIXYFMAOFNAXNF7CV32FPXDBAVCNFSM6AAAAABTPIOKZKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKNBTGAYDSNJTGQ> > . > You are receiving this because you authored the thread.Message ID: > ***@***.***> >

[SuperQ]

I think a bugfix PR would also be fine.