Skip to main content

Troubleshooting Cluster components

This article introduces the problems and troubleshooting ideas of Rainbond cluster components.

Troubleshoot ideas

If such problems occur, you can troubleshoot according to the following troubleshooting ideas.

rbd-api service exception

The rbd-api log reports an error:

websocketproxy: Error when copying from client to backend: websocket: close 1006 (abnormal closure): unexpected EOF

There are two possible situations where this problem occurs:

  1. rbd-api cannot connect to rbd-eventlog or rbd-node services. At this time, you need to check whether the network between POD is normal.
  2. The POD IP of rbd-eventlog or rbd-node has changed, but rbd-api has not been updated. In this case, the rbd-api service needs to be restarted.

rbd-monitor service exception

rbd-monitor is always in 0/1 state. In this state, the application view of Rainbond will not be displayed normally, because the corresponding application indicator data cannot be obtained.

Usually, in this case, you need to check the rbd-monitor log to see if there is any error message.

kubectl logs -fl name=rbd-monitor -n rbd-system

It can be solved by:

info

Removing PVC for rbd-monitor may result in loss of historical metrics.

kubectl delete pvc data-rbd-monitor-0 -n rbd-system
ctrl c end
kubectl delete sts rbd-monitor -n rbd-system
kubectl delete pod -l release=rainbond-operator -n rbd-system

Until the rbd-monitor service is recreated and started normally.