Oliver Gould

Results 207 comments of Oliver Gould

Thanks for (re-)opening this issue! I've opened https://github.com/kube-rs/kube-rs/issues/1000 to discuss restoring support for the environment variable and https://github.com/Azure/AKS/issues/3183 to discuss how this feature is at odds with documented client contract.

And https://github.com/kubernetes/kubernetes/issues/112263 discusses reconciling client-go's behavior with the documentation.

This is fixed by kube-rs/kube-rs#1001. Once kube-rs v0.75 is released, we'll update dependencies. This is unlikely to be included in Linkerd stable-2.12.1, but it should be available for stable-2.12.2 (probably...

1. We do not want to support a "dynamic" trust root. There are a number of operational/security issues with making it possible to change trust roots at runtime without restarting...

The current scheme is derived from https://gateway-api.sigs.k8s.io/references/policy-attachment/#target-reference-api, which states: > Each Policy resource MUST include a single targetRef field. It MUST not target more than one resource at a time,...

It seems that there is a known bug in webpki and PR to fix it that's been open without review for >1 year: * https://github.com/briansmith/webpki/issues/20 * https://github.com/rustls/rustls/issues/841 * https://github.com/briansmith/webpki/pull/226 Given...

If I understand correctly, lifecycle hooks can't actually do this. From https://kubernetes.io/docs/concepts/containers/container-lifecycle-hooks/#hook-handler-execution: > `PreStop` > > This hook is called immediately before a container is terminated due to an **API...

@uralsemih Thanks for submitting a PR. > ### What problem are you trying to solve? > > The collector exposes port 4318(http) using OTLP. Can you elaborate on what problem...

> This behavior is not reproducible with a Hyper client and a Hyper server. If we ignore the specific case of `http1_header_read_timeout`, any server that closes a connection can cause...

One problem with this approach is that not all requests are safe to retry: once we've written a request, I don't think an OS error means that the request hasn't...