Amlen fail to maintenance after switch and restart

[jochen123456]

Hi,

I have amlen installed within kubernetes in active/passive. It works fine except in one cluster. I think ot has something to do with the nfs volumes.

1. The psb-messagesight-0 goes to Maintenance after Restart of Pods. psb-messagesight-1 waits for sync.
2. I clean the store psb-messagesight-0. Now psb-messagesight-0 active.
3. I delete pod psb-messagesight-0. psb-messagesight-0 and psb-messagesight-1 goes to maintenance.
4. I clean the store psb-messagesight-0 and psb-messagesight-1. Now psb-messagesight-0 is active.

Can anyone help and point out what is the problem with the storage,

x.x.109.18:/psb_FS/psb-qa-psb-imaserver-data-psb-messagesight-0-pvc-7f22f362-4c70-4952-9186-04274e51188a on /var/lib/amlen-server type nfs4 (rw,relatime,vers=4.1,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,clientaddr=x.x.184.20,local_lock=none,addr=
x.x.109.18)
x.x.109.18:/psb_FS/psb-qa-psb-imaserver-log-psb-messagesight-0-pvc-ee319c4d-2a82-41ec-abf0-43d29834b853 on /var/lib/amlen-server/diag type nfs4 (rw,relatime,vers=4.1,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,clientaddr=x.x.184.20,local_lock=none,a
ddr=x.x.109.18)
x.x.109.18:/psb_FS/psb-qa-psb-imaserver-data-psb-messagesight-1-pvc-34925f78-0a0c-45d2-b66f-531b036712a8 on /var/lib/amlen-server type nfs4 (rw,relatime,vers=4.1,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,clientaddr=x.x.184.21,local_lock=none,addr=
x.x.109.18)
x.x.109.18:/psb_FS/psb-qa-psb-imaserver-log-psb-messagesight-1-pvc-79e99cfb-443d-45b1-81a4-4e7f1f1dd87d on /var/lib/amlen-server/diag type nfs4 (rw,relatime,vers=4.1,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,clientaddr=x.x.184.21,local_lock=none,a
ddr=x.x.109.18)

[jochen123456]

Uploading psb-messagesight-0-imatrace.log…

[jochen123456]

psb-messagesight-1-imatrace.log

[jonquark]

I was going to take a quick look at this, this morning - but weirdly, the "link" to the messagesight-0 log is a link to the issue noto to the file itself and I'd want logs from both machines (especially including a time that it goes into maintenance)