Hello all,
I have a disk that is giving some problems and am contemplating options to replace it. I
think I need to bounce my ideas off the collective wisdom here.
I have a disk that appears to be going bad. This disk is one of two in and ADVfs domain.
The usage is about 60% on each disk so I cannot just remove the one with rmvol and migrate
everything to the other. I wish it was that easy. And maybe it will turn out to be that
easy. The SCSI id's 0-6 are used on this machine. One thought I have for this is to
replace, temporarily, the tape drive, scsi 2, with another disk. I would then add this
disk with addvol to the domain. I would then rmvol the disk that is going bad, scsi id 6,
to get all the files off of it. At that point I could remove the failing disk and then
change the scsi id of the new disk from 2 to the id of the failing drive, id 6. My
problem with that seems to be that the ADVfs domain will be looking for id 2 instead of id
6? Am I right with that? I know I can go to /etc/fdmns/domain_name and change the pointer
there to look for id 6 again instead of id 2. Does this take care of what I need to do?
Is there anything on the drive(s) in the domain information that would prevent me from
doing it this way? The other way will be to do a dump of the domain, remove the whole
domain, replace the failing disk and then restore the whole domain. Method 2 here will
take at least 4-5 hours. I know the dump of the domain takes about 1 1/2 hours. I don't
really want to have the machine out of service that long. This machine happens to be our
main web server. At least with the addvol and rmvol I can keep the machine up while the
files move. Anyone done anything like this before or have some good ideas? TIA.
Mark
--
Mark Van Overbeke Systems Administrator
Computing Services
600 East 4th Street
University of Minnesota, Morris e-mail: markvo_at_mrs.umn.edu
Morris, MN 56267 Phone: 1-320-589-6378
--
Success is the progressive realization of your worthwhile goal or dream!
Received on Tue Apr 06 1999 - 13:46:22 NZST