Handle NFS mounts during 5.1 cluster member shutdown

From: ^IT_MPG_UNIX_ADMIN <^IT__MPG__UNIX__ADMIN_at_ccmail.allegiance.net>
Date: Wed, 07 Mar 2001 11:19:42 -0600

Hello all,

We are running Tru-64 5.1/TruCluster 5.1 on a two member cluster. I need some
help on configuring automatic re-mounting of all NFS filesystems on the other
cluster member when the server cluster is shut down.

For example, there are two machines HostA and HostB in the cluster. The df
command on HostA for NFS mounted filesystem /usr/mount shows:

# df -k /usr/mount
Filesystem 1024-blocks Used Available Capacity Mounted on
p01ap01:/usr/mount 20938912 10529756 3793840 74% /usr/mount

and cfgmfr shows HostA as the server

# cfsmgr /usr/mount

 Domain or filesystem name = /usr/mount
 Server Name = HostA
 Server Status : OK

#

HostB does not automatically re-mount NFS filesystem /usr/mount after HostA is
shutdown. This explanation is in the Tru-Cluster 5.1 release notes:


"When a cluster member shuts down (using the shutdown command), the /sbin/rc0
run-level transition script unmounts all NFS file systems in the cluster that
are not in active use. To provide automatic remounting of the file systems, use
AutoFS or automount."

We would like to keep the NFS filesystems defined in /etc/fstab, but use AutoFs
to re-mount the filesystems on another member in case of server member shutdown.
Defining the NFS filesystems both in /etc/fstab and AutoFS has caused conflicts
requiring a reboot. Anyone have any advice on how to handle this?

Thanks,

Marc Brandon
Allegiance Healthcare
1400 Waukegan Rd.
McGaw Park, IL 60085
E:mail:unixadmin_at_allegiance.net
Received on Wed Mar 07 2001 - 17:25:17 NZDT

This archive was generated by hypermail 2.4.0 : Wed Nov 08 2023 - 11:53:41 NZDT