Date: Sun, 5 Nov 1995 22:44:39 -0500 (EST) From: John Brann <jbrann@panix.com> To: julian@ref.tfs.com (Julian Elischer) Cc: hardware@freebsd.org Subject: Re: Nasty SCSI tape problem Message-ID: <199511060344.WAA23152@panix3.panix.com> In-Reply-To: <199511042245.OAA03245@ref.tfs.com> from "Julian Elischer" at Nov 4, 95 02:45:01 pm
next in thread | previous in thread | raw e-mail | index | archive | help
> > use the 'B' option to dump I'm doing that - that's where the calculations don't make sense. According to the man page B=no. of blocks and b= blocksize. So, in the absence of other information i'd expect dump 0uBbf 160000 10 /dev/nrst0 /usr and dump 0uBbf 1600000 1 /dev/nrst0 /usr to take up the same amount of tape. In fact, dump calculates 2.4 tapes for the first command and 0.24 for the second. (I'd give you the actual output from dump, but I just tried to do that and my system froze up - see below.) If I let the command run, the intermediate messages, and demands for tapes, match the calculations. This isn't quite the end, though. If I make a dump according to the second command it appears to complete. If I then rewind the tape and try to wind to the end of the dump (a simple test of a clean tape file, besides, I'd like to use the rest of the tape) using these commands - mt rewind mt fsf 1 the rewind works normally. The second command starts the tape winding, but it never stops. It sounds as though the tape winds to the end, rewinds, and then starts winding again. I have left this running for 15 minutes. If I interrupt the command, the system gradually freezes - as below. I originally posted this to 'hardware' because I'm suspicious of the tape drive, but the fact that this behaviour seems confined to the dump command has me confused. If I tar a file on to a tape, I can use 'mt fsf 1' to move past it without trouble. I should also add that my system is, otherwise, very reliable. Nothing else causes crashes or freezes of this kind. Sorry to be so long-winded, but this is annoying and I'm concerned about only having tar-style backups. John > > Hi, > > > > I have a Conner CTMS tape drive. This is a SCSI device which uses high density > > QIC-type cartridges which hold 1.6 or 2Gb. I have an Adaptec 2940/PCI SCSI > > controller which runs the hard disk, tape and CDROM. The system has no IDE. > > I'm running 951020-SNAP. > > > > I'd like to use 'dump' to make backups, so I was experimenting with it, but > > having problems persuading it to store my 700Mb /usr partition on only one > > tape. For some reason the calculations of number of blocks / blocksize > > don't seem sensible. Anyway, after 'dump' told me how many tapes I was > > going to need, I hit CTRL-C to abort. At this point the disk light came on > > and the whole system gradually froze until it totally stopped. Using the > > reset button started up the usual BIOS boot cycle - but before the point at > > which the BIOS tells me that the disk drive is drive C: and proceeds to boot > > from it, the system hung up - the disk light flickered intermittently, but > > nothing else happened. After about 2 minutes I powered the system down, said > > a short prayer, and switched it on again. Everything came up without trouble. > >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199511060344.WAA23152>