Hi!
on one of our machines (DEC OSF 4.0D) /usr is an advfs fileset
spanning multiple (3) disks. one of these (a RZ28M) went south
yesterday (unrecoverable read errors).
I got another RZ28M, "addvol"ed it to usr_domain and tried to
"rmvol" the flaky one.
Unfortunately, rmvol stopped as soon as it hit a read error..
so i first ran "balance" to "fill up the new disk" and then
ran "rmvol" in a loop, manually removing the "bad file" that
stopped each pass.
That is, until rmvol got stuck on <unknown> entries..
To make matters worse, the drive is deteriorating rapidly, which means
that (most) attempts to mount /usr crash after multiple read errors ..
Question:
- how would advfs cope with a *missing* volume (as in "physically
removed")
- could the remaining data (on the other disks) be salvaged?
TiA,
Steffen
------------------------------------------------------------------------
Steffen M. Kremser Danet GmbH
kremser_at_danet.de Mobile Communications Technology
Tel: +49-6151-868 484 Gutenbergstr. 10
Fax: +49-6151-868 498 64331 Weiterstadt, Germany
PGP: 1024/4DF769B5 Fingerprint: 85866AC0420EEDB7E0E0E9D988E772E1
Received on Thu Jul 22 1999 - 08:48:57 NZST