kyverno
kyverno copied to clipboard
kyverno
[Bug] LeaderElection related logs should be Debug level instead
Kyverno Version
1.7.2
Description
While checking Kyverno logs, just noticed it prints out lots of leader election related logs:
kyverno-84ffd96fb5-fpn4p kyverno I0628 07:48:15.561563 1 round_trippers.go:553] GET https://10.233.0.1:443/apis/coordination.k8s.io/v1/namespaces/kyverno/leases/kyverno 200 OK in 3 milliseconds
kyverno-84ffd96fb5-fpn4p kyverno I0628 07:48:15.563599 1 round_trippers.go:553] GET https://10.233.0.1:443/apis/coordination.k8s.io/v1/namespaces/kyverno/leases/kyverno 200 OK in 1 milliseconds
kyverno-84ffd96fb5-fpn4p kyverno I0628 07:48:15.563740 1 status.go:31] webhookMonitor/WebhookStatusControl "msg"="annotation kyverno.io/webhookActive already set to 'true'" "name"="kyverno" "namespace"="kyverno"
kyverno-84ffd96fb5-7txh6 kyverno I0628 07:48:16.020497 1 round_trippers.go:553] GET https://10.233.0.1:443/apis/coordination.k8s.io/v1/namespaces/kyverno/leases/webhook-register 200 OK in 2 milliseconds
kyverno-84ffd96fb5-7txh6 kyverno I0628 07:48:16.025020 1 round_trippers.go:553] PUT https://10.233.0.1:443/apis/coordination.k8s.io/v1/namespaces/kyverno/leases/webhook-register 200 OK in 4 milliseconds
kyverno-84ffd96fb5-7txh6 kyverno I0628 07:48:16.025184 1 leaderelection.go:278] successfully renewed lease kyverno/webhook-register
Leader election logs prints lots of information in I level and our logging system pushes these logs to elasticsearch index storage unnecessarily.
It would be nice to set log level D (Debug) or something.
cc @developer-guy
Slack discussion
No response
Troubleshooting
- [X] I have read and followed the documentation AND the troubleshooting guide.
- [X] I have searched other issues in this repository and mine is not recorded.
@Dentrax I closed the PR because the changes I made through the PR wasn't expected changes as @realshuting suggested, Would you like to explain this issue a little bit more so that It could be clear what's the expectation for the issue because this issue is also unclear to @realshuting. As you mentioned in the issue for changing the the level of logs related to the leaderElection but all the leaderElection logs are printing relevant message. slack conversation link
Yes, actually:
$ helm list
NAME NAMESPACE REVISION UPDATED STATUS CHART APP VERSION
kyverno kyverno 2 2022-10-20 15:40:37.344646 +0300 +03 deployed kyverno-2.6.0 v1.8.0
It would be nice to set log level D (Debug) or something.
@Dentrax this does not exist (only I or E are available).
We are switching to structured logging, will be ready in 1.8.2 hopefully.
What -v level are you running ? Some logs here are not controlled by us but come from a k8s lib. Are you running with -v >= 5 ?
Are you running with -v >= 5?
I was passed -v=5, is there another flag to set log verbosity?
Nope but with -v=5 there’s nothing we can do, it’s hard coded in k8s lib.
@Dentrax can we close this as there's nothing much we can do when -v=5 ?
Oh I see. I thought I had to pass a flag to make verbosity level to info. I couldn't see any info log from kyverno containers. I did helm install with default values. If this was the expected behavior, so LGTM! We can close this one. 🤞
kyverno-84ffd96fb5-7txh6 kyverno I0628 07:48:16.025184 1 leaderelection.go:278] successfully renewed lease kyverno/webhook-register
~~This line comes from kyverno itself for example.~~
Actually no.
There's no reason kyverno should not output logs. Can you check on the pods ?
