Nuru

@dependabot rebase

My runners are exiting with this error if they remain idle more than 2 minutes. Deployed via `actions-runner-controller` [Helm Chart](https://github.com/actions/actions-runner-controller/blob/master/charts/actions-runner-controller/Chart.yaml) v0.23.7.

/remove-lifecycle stale

@njtran wrote: > Alternatively, the approach we're now running with for v1 is using consolidateAfter to filter out nodes for consolidation. When you set consolidationPolicy: WhenUnderutilized and (for example) set...

@njtran wrote: > > When a Node appears underutilized > > Hey @Nuru, this is one of the toughest things to compute with the approach in this PR. Finding if...

@jmdeal I'm wondering if we might be subject to this bug, or a similar bug, or if Karpenter properly handles this use case. It is difficult to catch happening, and...

@chiluk wrote > This patch breaks throttled metrics on Cent 7. That's why I haven't approved it yet. Throttled periods should be 50+ for both good and bad kernels but...

Unfortunately, without the offending code, I'm seeing the same behavior on the unpatched kernel where we are seeing throttling problems under Kubernetes, Debian 4.9.168-1+deb9u5 (2019-08-11): Throttled for: 51 CPU Usage...

@chiluk I forked this project intending to make a PR, but it seems like I may be hijacking your intent. What I have at [Nuru/fibtest](https://github.com/Nuru/fibtest) is strictly meant as a...

I tried the https://github.com/abutaha/aws-es-proxy but it does not seem ready for production. It has 13 open issues at the moment, 2 of which are deal killers for me: it fails...