Date: Mon, 22 May 1995 09:00:33 -0400 (EDT) From: Mark Hittinger <bugs@ns1.win.net> To: hackers@FreeBSD.org Subject: Re: kern/430: bug in tape drivers (fwd) Message-ID: <199505221300.JAA00995@ns1.win.net>
next in thread | raw e-mail | index | archive | help
> > It sets the CCB_ABORTED flag the first time, so you should only > get one "Try to abort" message per aborted transaction. More than > one message means more than one aborted transaction. We > ought to print out the ccb with %p. Will patch it and see what I get. > > > > The workaround (a gross and sleazy kludge I admit :-) ) lets me do backups > > without bitspraying my disks! > > Are you sure? > Which other transfers were aborted? Were they > retried? Are you sure you don't have > subtle corruption problems, and you just aren't getting gross disk > corruption and a system panic? > I've been doing it since early december. This is on my heaviest hit server so I doubt that something terribly ugly occurs here. At least I've not seen anything :-) I will pay more attention to this and put more printfs into the driver for more clues. Oh yeah, and I was getting gross disk corruption and system panics every time without the gross sleazy patch. That was the point of the gross sleazy patch :-) Several afternoons in fsck for too many hours. > I think the tape drive isn't disconnecting and the > SCSI bus is locked up. As long as you ARE getting "disk retry" messages > you are OK, though. If you aren't I think you may have problems. I got another email too about enabling scsi disconnection on the tape drive. I double checked and - dammit - it was not on for that device. I can't remember if that was important. This morning I will enable disconnection and run another backup - about 13 gigs of stuff to the drive. See what happens ect. Regards, Mark Hittinger bugs@win.net
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199505221300.JAA00995>