Resolving Network Errors - IBM Storwize V7000 Unified Problem Determination Manual

2. Allow fifteen minutes for the error to disappear. If the error does not disappear,
Error for "The mount state of the file system /ibm/
Filesystem_Name changed to error level"
About this task
If the command lshealth -i gpfs_fs -r returns "The mount state of the file
system /ibm/Filesystem_Name changed to error level", complete the following
steps to resolve the issue.
Procedure
1. Verify that the other file module role displays Host State OK. Repair the host
2. Issue the command mountfs fileSystem.
3. Issue the command lsfs -r.
4. Issue the command lshealth -i gpfs_fs -r.
5. If the error persists, refer to the GPFS documentation to debug or repair the

Resolving network errors

Use the following information and examples to resolve network errors that are
identified by the health system.
If a network is not attached to an interface, the health center monitors all ports and
logs and displays any port failures.
To stop monitoring an unused port, use the attachnw command to attach the
interface that corresponds to that port. After this command is issued, you must
manually mark the displayed error events for the unused port as resolved. You can
use the GUI system details panel to manually mark the events.
Whether a port is monitored depends on the attached network configuration and
the interface that is used for the management network. If an interface is not in use
by an attached network, the interface is monitored if it is used by the management
network.
Important:
It is highly recommended, that for a given VLAN subnet definition, you must also
consistently provide the same VLAN ID (tag) for that subnet on any network
definition that you create within a common interface bond on the clustered system.
If you define a VLAN on your management network, and you have the same
VLAN subnet on the same interface bond for data connectivity, ensure that you
provide the exact same VLAN ID. Additionally, the switch configurations
supporting this VLAN should be same for all connections that the VLAN uses.
Ensure that you use trunk, access link, or native VLAN consistently on the switch
ports that are connected to all ports on the clustered system. The clustered system
shares the same VLAN subnet. Avoid defining the same VLAN with a tag ID
198
Storwize V7000 Unified: Problem Determination Guide 2073-720
attempt to reboot the passive management node. The issue should be resolved
after the reboot and within five minutes after the file module displays Host
State OK again.
state if necessary.
The command output should display The mount state of the file system
/ibm/gpfs1 was set back to normal level.
error.