Hello,
I have a problem with Netbackup on an E450 running Solaris 8.
The E450 has a system disk, and a 100 Go mirror created
with Disksuite. There is a DB running on the mirror, that
needs to remain available at all times.
What we want to do :
- before backuping, lock the DB, put one submirror offline,
unlock the DB
- backup the offlined submirror (with the DB in a coherent
state)
- after backuping, put the submirror back online, and let
Disksuite deal with sync issues.
I have created bpstart_notify and bpend_notify scripts
accomplishing precisely what we need. So far, so good.
Unfortunately, when Netbackup backups the system, it
doesn't wait for the data to be sent to the server
before running the bpend_notify script (as it should, per
the Netbackup administrator's manual), instead running
that script right after the bpstart_notify.
I know the bpend_notify is supposed to wait only until
the data to backup is buffered, but, in our case, that
script runs less than half a second after the bpstart
script is over. I do not think our server can buffer and
send 50 megs (our backup test) in less than a second
(though I'm sure our users would love it :-)
Is there a known bug with Netbackup and Solaris 8? Or I
am overlooking something obvious?
TIA,
Daniel
--
Daniel Polombo
System/network engineer
Thomson-CSF Detexis
|