On Mon, 4 Nov 2002, Robert Albano wrote: > When I give the following command > >mt -f /dev/st0 rewind > I get the following message > >...device or resource busy.... > > On Friday Nov 1 the samba server was not responding so the office > manager powercycled the machine. The samba server is working but the > the tape drive no longer responds to the mt commands giving a "...device > or resource busy error...". There appear to be no dead tar or mt > commands when I do an > ps -ax > > Oct 30 23:10:26 server-1 kernel: Additional sense indicates Scsi parity > error > Oct 31 23:05:00 server-1 kernel: (scsi2:0:4:0) CHECK_CONDITION on > REQUEST_SENSE, returning an error. > Oct 31 23:05:02 server-1 last message repeated 2 times > Oct 31 23:05:02 server-1 kernel: scsi2 channel 0 : resetting for second > half of retries. > Oct 31 23:05:02 server-1 kernel: SCSI bus is being reset for host 2 > channel 0. > Oct 31 23:05:02 server-1 kernel: (scsi2:0:4:0) CHECK_CONDITION on > REQUEST_SENSE, returning an error. > Oct 31 23:05:02 server-1 last message repeated 2 times
In the initial kernel probe the device, which I presume is st0, returns an error. No amount of fiddling with mt(1) is going to work if the scsi device is scrammed. I'd give the physical scsi hardware a once-over to make sure the cables are not in askew nor in a place they're likely to overheat. Schedule a maintenence shutdown and gently powercycle the scsi hardware too (I'm assuming this tape unit is external). Make sure the external scsi hardware is on before the host is powered up. Also, make sure nobody's made off with a terminator or swapped cables on you. Load a blank or scratch tape and run: $ mt -f /dev/st0 status See what that gets you. If you still have a bunch of errors in /var/log/messages, put the tape unit on another system with a scsi interface and see if you get the same problems. If you have the spares laying around you can work through the deduction of which part is failing: drive, controller, cable, terminator... If it does turn out any hardware is failing, I recommend making sure you can read any existing tapes from another drive and making a new backup immediately. Good luck, -po _______________________________________________ Eug-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
