Skip to content
This repository has been archived by the owner on Nov 19, 2020. It is now read-only.

Watcher for CRD timeout in 90sec with EOF #116

Open
hardwaysun opened this issue Apr 9, 2019 · 3 comments
Open

Watcher for CRD timeout in 90sec with EOF #116

hardwaysun opened this issue Apr 9, 2019 · 3 comments

Comments

@hardwaysun
Copy link

Not sure it is K8s doing it or something in the package. When watch resource like validatingwebhookconfiguration, it return from the watch.Next with EOF every 40 min , but if I define crd and watch it, it return every 90 second, tried adjust http.Transport{IdleConnTimeout} or even the --min-request-timeout=1800 on apiserver, all doesn't help.

Intuitively it should be same as it is just a resource on apiserver we doing long poll, should not matter what we looking for, but struggled for 3 days, can't find the issue. See log below, mytestcrd return very 90sec, whild the validatingwebhoodconfiguration do so around 45min

2019-04-09T12:32:16.168|ERRO|CTL|resource.(*kubernetes).startWatchResource.func1: Watch failure - error=decode event: EOF resource=validatingwebhookconfiguration
........................
2019-04-09T13:21:58.73 |ERRO|CTL|resource.(*kubernetes).startWatchResource.func1: Watch failure - error=decode event: EOF resource=mytestcrds
2019-04-09T13:22:21.342|ERRO|CTL|resource.(*kubernetes).startWatchResource.func1: Watch failure - error=decode event: EOF resource=validatingwebhookconfiguration
2019-04-09T13:23:22.714|ERRO|CTL|resource.(*kubernetes).startWatchResource.func1: Watch failure - error=decode event: EOF resource=mytestcrds

@silenceli
Copy link

I meet the some problem. I don't known how to solve

@silenceli
Copy link

Not sure it is K8s doing it or something in the package. When watch resource like validatingwebhookconfiguration, it return from the watch.Next with EOF every 40 min , but if I define crd and watch it, it return every 90 second, tried adjust http.Transport{IdleConnTimeout} or even the --min-request-timeout=1800 on apiserver, all doesn't help.

Intuitively it should be same as it is just a resource on apiserver we doing long poll, should not matter what we looking for, but struggled for 3 days, can't find the issue. See log below, mytestcrd return very 90sec, whild the validatingwebhoodconfiguration do so around 45min

2019-04-09T12:32:16.168|ERRO|CTL|resource.(*kubernetes).startWatchResource.func1: Watch failure - error=decode event: EOF resource=validatingwebhookconfiguration
........................
2019-04-09T13:21:58.73 |ERRO|CTL|resource.(*kubernetes).startWatchResource.func1: Watch failure - error=decode event: EOF resource=mytestcrds
2019-04-09T13:22:21.342|ERRO|CTL|resource.(*kubernetes).startWatchResource.func1: Watch failure - error=decode event: EOF resource=validatingwebhookconfiguration
2019-04-09T13:23:22.714|ERRO|CTL|resource.(*kubernetes).startWatchResource.func1: Watch failure - error=decode event: EOF resource=mytestcrds

Do you find the answer now?

@hardwaysun
Copy link
Author

hardwaysun commented Mar 19, 2020 via email

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants