I have a 2 node NW6.5 SP6 cluster with local sys volumes and all other
volumes cluster enabled on an FC san. Each pool has several volumes. I
want to try bumping up the nss readaheadblks value, probably to 16 to start.

Should I put the readaheadblks command in the load script for the volume
resource, with the command for each volume preceding it's "mount
Volume_name VOLID=xxx"? Or should I put the command for each volume in
the nssstart.ncf file on both servers, even though some volumes listed
won't be mounted on one of those servers? Thanks.