Hi!
> >Generally everything seems to work fine, but sometimes I get these
> >messages in /var/log/messages:
> >...
> >Jun 20 09:57:00 kuller /kernel: (sa0:ahc1:0:6:0): 32768-byte
> >tape record bigger than suplied buffer
>
> When do you get these errors? When you're doing an amdump (i.e. writing
> to the tape)? When you're labeling a new tape? When you're doing a
> recovery (reading the tape)? When you're running amcheck (which would
> also be reading)? What happens to whatever you're trying to do, i.e.
> does Amanda fail in some way?
The way this error occurs is as follows:
I (logged in as root) run amcheck:
su amanda -c "amcheck DAILY"
But I don't have the right tape in the drive. Say, last night's
backup was done to DAILY03 and the next backup should go to
DAILY04. I have the DAILY03 tape in drive when starting
amcheck. Amcheck says that the tape DAILY03 is active tape, ejects
the tape and sends me mail requesting to load Amanda tape (I use
chg-manual). I load the correct tape (DAILY04), amcheck completes
successfully and I am returned to shell prompt. A few seconds after
that I receive the buffer error message from kernel as described in
my previous mail.
The error does not occur if I insert the correct tape before
starting amcheck.
As you can see, there is no serious problem, I can do backups and
restores just fine. I'm just interested to know if there is
something I can do to avoid these errors, either by configuring
Amanda or my FreeBSD kernel.
BTW, if anyone on amanda-users list sees this, please drop me a
line off-list. I haven't received any messages from the list for
several days now and for whatever reason [EMAIL PROTECTED] seems
to ignore me (I get no replies to commands sent when I try to
investigate the issue).
--
Toomas
