Closed janheise closed 12 months ago
After Bernd reduced the number of parallel jobs on the Jenkins workers, a lot of sporadic occurring container startup problems were not seen again for a couple of days. I keep it open as a reminder but am hopeful that we don't see this error again.
https://github.com/Graylog2/graylog2-server/pull/16133 could be another try to resolve issues that sometimes seem to occur with the performance plugin on CI.
Current Behavior
Sometimes a timeout occurs running the full backend tests. It looks as if all the containers start but index creation is not finished until the timeout occurs. The log output (see below, rather large) indicates, that the graylog container is up and running. We do not output the logs of the OpenSearch container with this type of error, so it's unclear if there is a hangup here or if it's just bad performance.
Possible Solution
The final parts of the GL logs: the "doesn't exist" repeats until the timeout
Stacktrace: