2.12.1 Required settings for monitoring shared disks (for Linux)
When the SNMP agent is installed in a cluster system, file system information about the shared disk might not be periodically acquired from the manager. This problem occurs because the target file system is not in the file /etc/fstab. The SNMP agent uses OS system calls to acquire file system information. These OS system calls acquire information for only the file systems in the file /etc/fstab. The solution to this problem is to add the information about the file system on the shared disk to the file /etc/fstab.
When the shared disk is monitored, the SNMP agent recognizes only the first field (block special devices) and the second field (mount points) among the various fields in the information about the file system on the shared disk in the file /etc/fstab.
This product does not recognize any other fields. Therefore, for details about how to set values for the other fields and how to check the values set for the other fields related to the shared disk in the file /etc/fstab, see the cluster software and OS documentation.
For example, if the cluster software name is HA monitor, the mount point of the shared disk is /mnt/test, and as a requirement for managing the shared disk, the shared disk is set to not be automatically mounted when the OS starts, the example setting in the file /etc/fstab is as follows.
- <Setting example>
/dev/sdb1 /mnt/test ext3 defaults,noauto 0 0
The above setting is an example for HA monitor cluster software. Before performing this setting, see the manuals and Release Notes for HA monitor.
If you use some other cluster software, perform the appropriate settings in the file /etc/fstab, according to the requirements of that cluster software and OS environment.