zalando-incubator / zalando-incubator/stackset-controller
fix transport issue with WATCH
Nobody has claimed this yet.
- Dominant language
- Go
- Stars
- 183
- Forks
- 29
- Avg merge
- 3d 18h
- Merged PRs (30d)
- 5
Description
Original issue https://github.com/kubernetes/client-go/issues/374#issuecomment-716846816
This could cause no update for 30 minutes...
Recent comments suggest that we can fix it and use http2 Transport for connecting to kube-apiserver and set ReadIdleTimeout: https://github.com/golang/net/blob/release-branch.go1.15/http2/transport.go#L117
we would need to switch all controllers that rely on WATCH to Go1.15 in order to fix this
import "golang.org/x/net/http2"
func configureKubeConfig(apiServerURL *url.URL, timeout time.Duration, stopCh <-chan struct{}) (*rest.Config, error) {
tr := &http2.Transport{
....
ReadIdleTimeout: 10 * time.Second, // <--
}
...
}
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start from the configureKubeConfig entry point shown in the issue and inspect the controllers that rely on WATCH. Verify how the kube-apiserver transport is configured and whether the Go 1.15 HTTP/2 ReadIdleTimeout behavior addresses the reported 30-minute lack of updates. Done means affected watches continue receiving updates without the transport issue.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- go, kubernetes
- Domain
- infrastructure
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100