John Belamaric

Are you using the version here in `master`? There is a newer version, right now just in branch `1.1.0-pre` that has some helpers for this. I think this new version...

Agreed we should make go-lint happy.

If I have 1000 clusters (or 10s of thousands!), a metric to correlate failures is WAY better than digging into individual node config options on individual clusters.

Maybe? I’ll let Joe chime in, he is closer to the details. Do we have an admission duration metric already? We can already tell if the feature gate is enabled...

/retest

Just needs a TOC update

Yes, it is merged right now as `provisional`, so the PRR tooling doesn't kick in - Rob & I discussed this. Before it moves to `implementable` it needs PRR. @robscott...