Per k8s namespace billing information about k8s.io HOT 14 OPEN

/assign @thockin

from k8s.io.

Is it possible, what is the timeframe, how does it integrate with the other billing data

from k8s.io.

Is that this: https://cloud.google.com/kubernetes-engine/docs/how-to/cluster-usage-metering ?

from k8s.io.

Need clusters turned on and activated to try using this, believe it may be a beta thing. Believe data is dumped into bigquery and then we can extend the report.

from k8s.io.

Tim needs a chance to play with this first hand

from k8s.io.

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

from k8s.io.

Our issue lifecycles tend to be a bit longer than a k8s release.
/remove-lifecycle stale

from k8s.io.

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

from k8s.io.

/lifecycle frozen
/remove-lifecycle stale
/remove-lifecycle rotten

from k8s.io.

/assign @bartsmykla
Bart tried to copy the data studio report but was unable to connect to the data set. Will connect with @justinsb to figure out how to resolve permissions issues

from k8s.io.

/priority backlog
/remove-priority important-longterm

The shape of the billing shows this isn't really important compared to larger infra-sized billing buckets

from k8s.io.

/assign @Riaankl

from k8s.io.

/remove-priority backlog
/priority awaiting-more-evidence

from k8s.io.

The only reason I would think to bump this up in priority is more scope growth:

• I want to know which prow jobs are costing us the most on-cluster spend (vs. those that cost us by provisioning clusters in e2e projects)
• If the usage data we're collecting is able to tie together from arbitrary pod labels we should be able to do this (prow.k8s.io/job is applied to all Pods created from prowjobs)

If we find cluster usage in kubernetes-public as a whole is really high, maybe there's value in per-namespace to see which of our apps is generating the most usage. I suspect the usage as a whole is significantly less than our CI or artifact hosting spend though.

from k8s.io.