I have some older servers that have already been updated through quite
few versions of OES. This process never seems to go as smoothly as I
would like so for the last few years instead of doing in-place upgrades
I have just done a fresh install. The NSS volumes are on a separate
array so rebuilding the server doesn't effect the data volumes. With
OES11, when I do this the volumes will mount after the rebuild but the
information will not population in eDir. The docs indicate this is a
replica issue but all the servers hold the master replica for their
container. If I delete the volume and pool, I can recreate with the same
name but obviously I have to restore the data (which is the point of
leaving it). After some experimenting it seems I have to delete the
pool, not just the volume before I can recreate.

Process to replicate:

1)Shutdown existing server
2)Delete server info from eDir (there's quite a bit in this step, moving
replicas, making sure the info is out of all rings, etc)
3)Install OES11 using the same information as existing server
4)NSSMU mount existing volume (get -613 error)

I know this isn't a traditional upgrade procedure but it has worked well
for me for a few years now. We can do the server upgrade and have a
clean OS in just a few hours. It would take days if we have to move the
data off the NSS volumes, upgrade, then move it back.

Any suggestions on how to avoid or repair the -613 error aside from
deleting the pool?