Using prometheus as metrics provider

[Jack-Kingdom]

What would you like to be added?

support using prometheus metrics info as metrics source.

Why is this needed?

metrics-scraper store metrics info for provide, and prometheus did this job too. We have no need store those data twice.

[k8s-triage-robot]

The Kubernetes project currently lacks enough contributors to adequately respond to all issues.

This bot triages un-triaged issues according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

• Mark this issue as fresh with /remove-lifecycle stale
• Close this issue with /close
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale

[k8s-triage-robot]

The Kubernetes project currently lacks enough active contributors to adequately respond to all issues.

This bot triages un-triaged issues according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

• Mark this issue as fresh with /remove-lifecycle rotten
• Close this issue with /close
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle rotten

[Jack-Kingdom]

/remove-lifecycle stale

[Jack-Kingdom]

/remove-lifecycle rotten

[k8s-triage-robot]

The Kubernetes project currently lacks enough contributors to adequately respond to all issues.

This bot triages un-triaged issues according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

• Mark this issue as fresh with /remove-lifecycle stale
• Close this issue with /close
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale

[Jack-Kingdom]

/remove-lifecycle rotten

[Jack-Kingdom]

/remove-lifecycle stale

[avanish23]

Hi @Jack-Kingdom
The ask seems to be able to disable the default metrics scraper that comes with the dashboard as you already have Prometheus collecting them for you.
Is my understanding right?

[Jack-Kingdom]

Yeah, I mean, we can disable the default metrics scraper and use prometheus as metrics source.

[floreks]

If I understand correctly you would want dashboard to support prometheus as data source and display its' metrics directly in the UI?

[Jack-Kingdom]

@floreks Right!

I found we use sqlite as storage in metrics scraper. Use relation-based db as time seriesed db is not a good idea, this may cause performance issues in large scale k8s.

we can make metrics scraper as a adapter layer to support different metrics source. In my mind, most of k8s deployment have their own metrics system.

[floreks]

Supporting other data sources is on our list for sure. Victoria metrics/prometheus would be first.

Metrics scraper can't really be compared to them since it is designed to scrape only basic metrics and for a very short time. Normally it only stores the last 15 minutes to visualize current usage and short time peaks. It also scrapes only from metrics server, where i.e. prometheus has its own node-level scraper. For what it is being used is totally fine.

[avanish23]

Oh great, I now understand the ask.
I will take this up and start working on it as I find more bandwidth.
/assign