What To Do Next - IBM Storwize V7000 Unified Problem Determination Manual

10. Use the chdisk CLI command to restart all disks used by the file system. For
11. Use the mountfs CLI command to mount the file system. For example:

What to do next

Rebooting the file modules if none of the disks are up but all file volumes are
online:
To reboot the file modules if the multi-pathing driver may have failed following a
recovery of the control enclosure:
1. Identify the passive and the active management nodes from the Description
2. Wait until both nodes show OK in the Connection status column of the output
3. Resume the file module back into the cluster using the CLI command:
4. Reboot the file module that is the active management node using the CLI
5. Log back on to the Storwize V7000 Unified CLI. Wait until both nodes show
6. Resume the file module back into the cluster using the CLI command:
7. Then wait for GPFS to be active on both file modules in the output of the CLI
8. Check that the file systems are mounted by using the lsmount -r management
9. See Checking the GPFS file system mount on each file module if any file
Note that the management GUI can become very slow when the Storwize V7000
CLI is restricted. When you log on to the management GUI, it issues a warning
that the Storwize V7000 CLI is restricted. The management GUI runs the fix
274
Storwize V7000 Unified: Problem Determination Guide 2073-720
8. If none of the disks are up but all file volumes are online then the multi
pathing driver in the file modules may have failed and the best way to
recover is to reboot the file modules one after the other using the procedure
below.
9. If some of the disks are not up but the volumes are online then restart all
disks used by a file system before you continue to mount it.
example
chdisk <comma separated list of disk names> --start
mountfs <file system name>
column in the output from the CLI command:
lsnode -r
Reboot the file module that is the passive management node using the CLI
command:
stopcluster -node <node name> -retart
from the CLI command:
lsnode -r
resumenode <node name>
command. The active management node fails over to the file module that you
rebooted first.
stopcluster -node <node name> -restart
OK in the Connection status column of the output from the CLI command:
lsnode -r
resumenode <node name>
command:
lsnode -r
CLI command:
lsmount -r
systems are not mounted.