On Friday 09 August 2002 15:45, Chris Johnson wrote:
>Hi again group,
>    I'm fighting (and losing) a battle with a  SCO unix system and
>amanda client. Every thing appears to be working correctly
>(ha...ha...ha). 'amcheck' comes backup without errors. The server
> can backup all my linux and windows NT servers fine and without
> compliant. gnutar has recently been updated to 1.13.25. When I
> run 'amdump' I can see the tar "scheduling" command running on
> the SCO machine. I'm attempting to backup 4 filesystems all are
> on a compaq raid controller. two are about 9 GB and two are 45 GB
> and they are about 85-90% full. Every time I run the dump I can
>     'tail -f /tmp/amanda/sendsize.[date,time].debug' and watch it
> time out on the first 45 GB filesystem it tries to schedule.
>    Is there somewhere I can give the server a longer time-out
> value?

Yes, its one of the *timeout values in amanda.conf.  Depending on 
which phase, estimate or the actual dump is timing out, you can 
rather arbitrarily double those until the error goes away.  It 
sounds like the estimate phase is being a problem right now, so I'd 
start by doubleing the 'etimeout' value.  With 45 gigs mostly full, 
even doubling it might not be sufficient, depending on how fast the 
machine and its drives are.  You could make it arbitrarily large, 
then check the email to see how long it actually took and set 
etimeout to about 125% of that.

[...]

-- 
Cheers, Gene
AMD K6-III@500mhz 320M
Athlon1600XP@1400mhz  512M
99.10% setiathome rank, not too shabby for a WV hillbilly