Expand retry conditions for K8 logs #818
Conversation
AaronH88
force-pushed
the
goaway_catch_update
branch
from
625a5d2
to
5eb8c0f
Compare
AaronH88
force-pushed
the
goaway_catch_update
branch
2 times, most recently
from
46cabf7
to
2df4f86
Compare
AaronH88
force-pushed
the
goaway_catch_update
branch
from
2df4f86
to
978f1d6
Compare
|
Kudos, SonarCloud Quality Gate passed!
|
pkg/workceptor/kubernetes.go
Outdated
| @@ -275,7 +255,7 @@ func (kw *kubeUnit) kubeLoggingWithReconnect(streamWait *sync.WaitGroup, stdout | |||
| break | |||
| } | |||
| } | |||
| if err != nil { | |||
| if err != nil && remainingRetries > 0 { | |||
There was a problem hiding this comment.
I don't think the additional conditional && remainingRetries > 0 is needed here
I think the only way that the code reach this point is if we exhausted the retries?
There was a problem hiding this comment.
agreed. I don't think this conditional will ever hit
There was a problem hiding this comment.
Removed, i think Seth is right, now with the changes this condition will never be hit, the first err != nil will be triggered and this will never be called.
I have removed the whole check.
AaronH88
force-pushed
the
goaway_catch_update
branch
2 times, most recently
from
9591f7e
to
b6c9e6c
Compare
pkg/workceptor/kubernetes.go
Outdated
| *stdoutErr = err | ||
| kw.Error("Error reading from pod %s/%s: %s", podNamespace, podName, err) | ||
|
|
||
| return |
There was a problem hiding this comment.
these code will still need to be executed when the retries is exhausted
AaronH88
force-pushed
the
goaway_catch_update
branch
3 times, most recently
from
9c83f22
to
4295e36
Compare
AaronH88
force-pushed
the
goaway_catch_update
branch
from
4295e36
to
a931ac4
Compare
Fixes: ansible/awx#14293
Expanded retry conditions for K8 logs to retry at least 5 times regardless of the error returned. This is due to K8 api timings where the GOAWAY error may not be sent or sent already before log streams get attached.