cancel
Showing results for 
Search instead for 
Did you mean: 

VCS 5.0 Multiple Failover Groups NFSRestart Agent

ArjunYK
Level 2

We have been trying to configure NFS Server in parallel mode,  but no luck yet.   Would like to know if anyone has succeeded in doing similar confiugration. Spent long hours with Symantec tech support, still no joy.

Here is what we are trying:

VCS5.0 in 3 node cluster. Solaris 10 Sparc.
Uses EMC SAN storage and VXFS filesystem for providing the data to be shared using NFS Share.
About  8 Failover service groups contains 4 or 5 NFS Share resources in each, along with other resources in it.
More than one service group is online in a node at a time, with Share resouces.
These 8 Failover service groups with Share resources needs to be able to failover to any other node
Cluster also contains other Failover Service groups, which does not contain any NFS Share resources, so they are not relevent here.

/// These are the resources in a group with NFS Shares.  (All 8 service groups are similar to this)
proj1_proxy_nfs
proj1_share_1
proj1_mount_1
proj1_dg_1
proj1_ipmulticnicb_1
proj1_nfsrestart
proj1_proxy_infra_multinicb


/// Resources in Parallel group with NFS Server resource and the MultiNICB
infra_NFSserver
infra_multinicb


This configuration works as expected. But, we are not very happy about this design, as the nfsd daemon is killed by the NFSRestart Agent each time the service group is offlined, and that's expected behaviour of NFSRestart Agent.

Can we get this done without having to kill the nfsd when a service group is switched or offllined ? Any suggestions ?


Thanks in advance
Arjun

1 REPLY 1

S_Herdejurgen
Level 4
Accredited Certified

Have you considered moving the multiple proj*_nfsrestart resources to the parallel service group as a single infra_nfsrestart resource?

-Seann