Arvind Iyengar

Results 34 comments of Arvind Iyengar

Identical to backend issue filed in https://github.com/rancher/rancher/issues/35311.

ProjectHelmCharts supports this with a construct called a GenericMap https://github.com/aiyengar2/helm-project-operator/blob/3032ca2bb64964f96bb446657d246826ce78a670/pkg/apis/helm.cattle.io/v1alpha1/project.go#L20, which also implements `toYaml`. If k3s-io/helm-controller wants to introduce this feature, I could move it over the rancher/wrangler as a...

Hi @gbonnefille, this PR is still a work in progress so I’m not sure yet. But splitting it into multiple PRs sounds like a good idea!

@brandond sorry for the delay! Rebased the current version of helm-controller to the latest master, can you take a look when you get a chance? Responded to the comments too.

@brandond if this is helpful, the contents of `pkg/controller/chart/chart.go` is just a renamed version of `pkg/helm/controller.go` with a bunch of code rewritten. The GH diff makes it seem like it...

> I've tried to approach this a few times, but honestly it's too much for me personally to review in a single PR. 42 commits and 3k changed files is...

https://github.com/k3s-io/helm-controller/pull/158 is the final PR that brings master to parity with the contents of this PR, so I will be closing out this PR in favor of that PR.

@gmintoco thanks for calling this out! Based on https://github.com/etcd-io/etcd/pull/12122/files#diff-051b19bef7add661d9678608040a89efaead1aa9d3063323effdfedae11c9978R76-R100, it seems like you are correct, although the etcd team seems to replace `BY (job, instance, grpc_service, grpc_method)` with `without (grpc_type,...

@ccittadino-ctic the issue was closed out in upstream without resolution. As @Tejeev mentioned, the upstream chart is still out of date at etcd v3.4.x: https://github.com/prometheus-community/helm-charts/blob/8318ce1d92e8c3dd487d89066d0f7b2405c65653/charts/kube-prometheus-stack/templates/prometheus/rules-1.14/etcd.yaml#L2

We are waiting on `kube-prometheus-stack` to absorb the latest dashboards from etcd; once `kube-prometheus-stack` has made the changes, we'll rebase on the latest upstream chart to get in the fix.