Events and other resources not deleted in OKD 4.13 #1867
Replies: 2 comments 3 replies
-
That is not supported, I'm not sure what configmap you mean
Is it for |
Beta Was this translation helpful? Give feedback.
-
|
Yes, I know it's not supported to change that value, 3h it's the default value. So it's suppose to store only the 3h events, not more And yes, I have all the "installer" configmaps and secrets, although I already delete some of them. |
Beta Was this translation helpful? Give feedback.
-
|
Any idea why it doesn't delete the old events? |
Beta Was this translation helpful? Give feedback.
-
|
We'd need must-gather to find out more - it could be one of known issues e.g. https://issues.redhat.com/browse/API-1650 |
Beta Was this translation helpful? Give feedback.
-
|
As for "what's eating my etcd space" see https://etcd.io/blog/2023/how_to_debug_large_db_size_issue/#what-data-occupies-most-of-the-storage-space - it should give you a list of objects. Later you can figure out which namespaces are these coming from, if its indeed from |
Beta Was this translation helpful? Give feedback.
-
Hi all,
OpenShift version
4.13.0-0.okd-2023-06-24-145750
Problem:
The events in all namespaces are not being deleted, there are events since 1 month ago and more. Even if in the ConfigMap config is: "event-ttl":["3h"]" in the apiserver. It should only store the 3h events.
The config maps and secrets in openshift-kube-apiserver created for every update, don't rotate either, I think it should store only the last 10 updates, and it doesn't delete the previous ones, I am in the 66:
And the old ones are still there:
Consequences:
Last week I get the etcd cluster full because of the thousands of objects in the cluster, events mostly (alarm:NOSPACE). I am deleting then right now manually.
Solution:
How can I rotate the events and resources correctly??
Kind regards
Beta Was this translation helpful? Give feedback.
All reactions