Closed liornoy closed 10 months ago
The the server doesn't have..
message is the same one any oc
command gives when it can't access something.
This issue will need someone like @tssurya, @npinaeva or someone from the networking team to review.
The following are lines of code that generate this error; so
https://github.com/openshift/must-gather/blob/master/collection-scripts/gather_network_logs#L78
https://github.com/openshift/must-gather/blob/master/collection-scripts/gather_network_logs#L91 (likely using related resources, to get to the qoses
object)
I think the source is here https://github.com/openshift/oc/blob/master/pkg/cli/admin/inspect/util.go#L118-L119 I have seen this warning for other resource types too (not networking-specific), so maybe we need to change log level for it?
@soltysh / @ingvagabund can you investigate here if changing the log level in inspect is correct; or if this is a collection issue (IE: a request for a missing/bad resource)?
The warning was introduced to avoid displaying an error when a non-existing resource is inspected. It's safe to increase the log level +1 over the default one. Just, it might be impossible to propagate the log level from the initial command down to the low level binary in the chain
@liornoy I assume all the must-gather logs can be found in any CI job?
@ingvagabund yes, each cnf-tests CI job of ours collects must-gather logs. But the logs about "egressfirewalls" and "egressqoses" don't reappear every time. I think I encountered that in a corner case where the cluster didn't setup correctly.
Issues go stale after 90d of inactivity.
Mark the issue as fresh by commenting /remove-lifecycle stale
.
Stale issues rot after an additional 30d of inactivity and eventually close.
Exclude this issue from closing by commenting /lifecycle frozen
.
If this issue is safe to close now please do so with /close
.
/lifecycle stale
I will close this for now.
Looking at the must-gather.log from my latest ci-test run, it seems like we throw INFO logs which make the log harder to read. See:
I file this issue to suggest muting the
the server doesn't have..
logs to provide a cleaner log output for the must-gather.