Replies: 2 comments 4 replies
-
We had similar problems. But after configuring server and client timeouts for the watch streams as mentioned in #698 it was fixed. Something like this @kopf.on.startup()
def configure(settings: kopf.OperatorSettings, **_):
settings.watching.server_timeout = 120
settings.watching.client_timeout = 150 |
Beta Was this translation helpful? Give feedback.
1 reply
-
I have a some cluster-wide kopf operators based on...
...that, after a as-yet-to-be-determined number of days, stop responding to any new events. Just like @hapatrick, if bounce the operator Pod the missing events are picked up and acted upon. I will try the timeouts suggested and report in a few weeks if the operator stability has improved. |
Beta Was this translation helpful? Give feedback.
3 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
We have noticed that sometimes some of our handlers are not getting called when we expect them to. E.g. we have a handler on the status field of an object, the status field is updated but the handler never gets called. If we restart the operator, the handler is called as expected after the restart.
This is a rare event; most of the time things work as expected.
Has anyone else noticed this? It is certainly possible this is a problem with our code rather than kopf, but if so I don't see what we are doing wrong. Any suggestions on how to remedy this or what to look at to understand what's happening are appreciated.
Beta Was this translation helpful? Give feedback.
All reactions