On Thursday 30 September 2004 10:04, Stefan G. Weichinger wrote: >Hi, Gene, > >on Donnerstag, 30. September 2004 at 05:27 you wrote to > amanda-users: > >GH> Greetings; > >GH> I'm trying to recover some stuff that yum killed last night, but > in GH> running amrecover, I believe I am getting the equ of an EOT > message GH> from the attempt to extract. > >GH> The man page for ammt isn't at all clear about how one goes > about GH> 'rewinding' a virtual tape. Can somebody clue me in? > >You don't need this with chg-disk. > >GH> Never mind, I used the old trusted dd and got it all. > >GH> BUT, amrecover needs to be able to rewind the "tape" after > you've done GH> a 'settape'. Somehow that functionality seems to > be missing, or at GH> least well hidden. > >Not sure if I understand correctly, but I think you should do it > like this: > >In terminal1 you do > >amrecover <conf> .... >[..] >sethost >setdisk >setdate >add >extract > >amrecover asks for the tape <label-of-requested-tape> ... > >In terminal2 you do > >amtape <conf> label <label-of-requested-tape> > Humm, and this is to switch the data pointer to point to the right tape somehow?
Ok, I fooled around with it and found the syntax needed, its a bit different as you need to enter the exact tape name, not just the slot number as is done in other utilities. The biggest problem right now is that I didn't make a scratch dir in /tmp, so now I have about 20 megs of /etc stuffs scattered around in /tmp when I unpacked the tarball I needed. What I can kill and what I cannot are iffy at best. But I have the space, so... Hindsite is 20/10 you know. It would be nice if this was all handled within amrecover though, this feels a bit kludgey. Different subject, last nites run was partially fubar, the amverify says, about 1/3rd of the way down the list it starts spitting out errors about wrong file position and eventually exits. And that virtual tape has no end of tape file in it either. That apparently took place around 3am as I was catching some zz's. I'd installed a newer kernel, 2.6.9-rc3 before I went to bed. There is nothing in the logs that corresponds either with the job, or the time frame. I wonder if this kernel is flakey... It also used the 2nd unused 'tape', skipping one because I'd moved the data link without echoing the right number to /path/to/config/chg-disk-slot, so I've reverted that for tonights run. If this one errors similarly, I'll reboot to an older kernel, and bug the lkml about this. >In terminal1 you continue ... > >This does the trick for me ... >Is this what you needed? I think so, but is this in the archives docs someplace? Or the manpages? -- Cheers, Gene "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) 99.26% setiathome rank, not too shabby for a WV hillbilly Yahoo.com attorneys please note, additions to this message by Gene Heskett are: Copyright 2004 by Maurice Eugene Heskett, all rights reserved.
