Jonathan Innis

cc: @jackfrancis @tallaxes @Bryce-Soghigian @ellistarn @dtzar

This appeared to be mistakenly marked as stale. I think we need to update our `StaleBot` to deal with the new issue labels in `kubernetes-sigs`. I'll create an issue to...

/assign @jigisha620

@sidewinder12s We're planning on looking over all of our metrics right now to get a more comprehensive idea of where we have gaps so this is good feedback for that...

@Horiodino Assigned!

Seems like this one might be difficult for the same reason that #686 is difficult. The second you get multiple NodePools into the mix, we now have to figure out...

> Actually using the metric you'd likely use a rate or duration of elevated count to indicate Karpenter is unable to process some pod +1, on board with this. Textbook...

> It is important for us that PDBs are never violated. We are happy with nodes left in a partially drained state indefinitely. We have alerting which will highlight this...

> Or are there other CRD resources that we want to add/augment aggregated status data for We'd probably also want to scope this proposal to the `NodeClass` and `NodeClaim` as...

It's the CloudProvider-specific parameters for a NodeClaim. We should probably enforce readiness in the status as part of the contract for a NodeClaim and then we can check the readiness...