This PR fixes the issue that mission/edgecluster status are not correctly reported when the underlying edgecluster is not long responding but clusterd is working. The existing code only handles the cases where the clusterd cannot reach the cloud.
Before this change, if the connection between clusterd and cloud is OK, but the apiserver in the edgecluster is not longer responding, "kubectl get edgecluster" and "kubectl get missions" will get the following outputs, which are really confusing to the end users:
With this PR, the end user will see the following for the commands "kubectl get edgecluster" and "kubectl get missions" when the under cluster is not reachable.
pdgetrf
commented
3 years ago
This PR fixes the issue that mission/edgecluster status are not correctly reported when the underlying edgecluster is not long responding but clusterd is working. The existing code only handles the cases where the clusterd cannot reach the cloud.
Before this change, if the connection between clusterd and cloud is OK, but the apiserver in the edgecluster is not longer responding, "kubectl get edgecluster" and "kubectl get missions" will get the following outputs, which are really confusing to the end users:
With this PR, the end user will see the following for the commands "kubectl get edgecluster" and "kubectl get missions" when the under cluster is not reachable.