VERIFICATION:
- Check if there were any recent Failovers on the filer.
- Please run the below command.
# pz_resilverd -s
If you get a similar output then the filer is affected by this bug[SA:filername - Active /]# pz_resilverd -s
HA mode, exit
WORKAROUND:
This is a known bug CFS-8508 and we can follow the below steps to fix this.
- Double check and make sure this is an active filer and not an HA.
- Take a backup of the files, config-resilver.md5 & resilverd_ddt* under /cloudfs/.resilver/
- Once they are backed up, these files must be removed.
Example below:
[SA:Filername - Active .resilver]# ls -lh
total 3
-rw-r--r-- 1 root wheel 221M Sep 15 01:47 resilverd_ddt_dump > Backup this file and remove this
-rw-r--r-- 1 root wheel 0B Oct 18 11:22 config-resilver.md5 > Backup this file and remove this
drwxr-xr-x 2 root wheel 2B Mar 10 2022 status/
[SA:Filername - Active .resilver]# pwd
/cloudfs/.resilver
- Then run the below command to copy and take a backup of the below file.
# cp -p /opt/pixel8/vm_common/get-cc-role /opt/pixel8/vm_common/get-cc-role.orig
- Once the file has been copied, we need to modify this file using vi editor. We need to make the below change in line 26.
Before:
echo -n $ccMode
After:
echo -n "subordinate"
You can directly jump to Line 26 by running ":26" after the file is open in vi.
- Once the change is made, you can verify the fix by running pz_resilverd -s. The output should now display the CMP details.
- To monitor the progress after fixing, you may use the below command.
# pz_resilverd –s | grep last_consistent_snap
Example:
[SA:Filername - Active /]# pz_resilverd -s | grep last_consistent_snap
{last_consistent_snap=(175649,csp=197113,cmp=175649), last_trnd_snap=197113}