GitLab Kubernetes Agent throws confusing errors when it is configured via web socket
Summary
Steps to reproduce
- configure GitLab Kubernetes Agent according to the guide GitLab Kubernetes Agent
- use web socket (wss or ws) when specifying
kube-address
in the section Install the Agent into the cluster - check the logs of the agent pod
What is the current bug behavior?
It will throw the warning messages / errors like these ones:
{"level":"warn","time":"2020-10-30T10:02:39.173Z","msg":"GetConfiguration.Recv failed","error":"rpc error: code =
Unavailable desc = transport is closing"}
{"level":"warn","time":"2020-10-30T10:02:39.173Z","msg":"GetObjectsToSynchronize.Recv
failed","project_id":"path/to/project","error":"rpc error: code = Unavailable desc = transport is closing"}
What is the expected correct behavior?
Probably the message should not be shown or it should be re-phrased in order not to confuse users.
Relevant logs and/or screenshots
Example from the logs is mentioned above.
Additional information
- reproduced with GitLab Helm Chart version 4.5.1
- kas-server version v.0.0.8
- the messages are not shown when kas is configured to work using grpc
Edited by Alexandr Tanayno