- This test case should be tested on both yaml installation, chart installation (Helm and Rancher UI), as well as upgrade scenarios
- After install Longhorn using on of the above method, ssh into a worker node that has a longhorn-manager pod running
- check the mount point
/run/secrets/kubernetes.io/serviceaccount
by running:root@node-1:~# findmnt /run/secrets/kubernetes.io/serviceaccount
- Verify that there is no such mount point
- Kill the longhorn-manager pod on the above node and wait for it to be recreated and running
- check the mount point
/run/secrets/kubernetes.io/serviceaccount
by running:root@node-1:~# findmnt /run/secrets/kubernetes.io/serviceaccount
- Verify that there is no such mount point
- Repeat the step 5 to 7 a few times