Dr. Blinn suggested, in nicer words, that I take a hammer to the tape. He
pointed me towards `/usr/field/tapex -m`, 'The Tape Exerciser Program' to
check whether or not I had any readable data on the tape. There was none. He
suggested the use of tapex after every backup, just to confirm that the drive
is actually able to read what it wrote, at least once.
Selden Ball pointed me towards `/usr/sbin/uerf -R -o full | more` in order to
read /usr/adm/binary.errlog. Here I found that the drive which failed had
been dying for months. On the tape server this had been filled up with errors
regarding the tape drive at around the time the dump would have been
occuring. The DEC_Event package on the associated products cd is designed to
monitor binary.errlog, and would probably be a good idea for the future. He
also pointed me to /sbin/scu -f /dev/[tape] to read information from the tape
drive. As for the data, its gone.
-Lucas Carey
> > We had a drive fail and the backup script tried to back it up anyway.
> > now the tape is inaccessable
> > mt status:
> > mt_erreg 0x3 Nonrecoverable medium error.
> > ASC/ASQ = 14/3 End of data not found
> >
> > restore -ib 64:
> > [a long wait]
> > Tape read error while trying to set up tape
> > continue? [yn] y
> > Tape is not a dump tape
> >
> > Tru64 v4.0f, the tape consists of 4 rdumps followed by 3 dumps. Is there
> > anyway to get the data off this tape? Did attempting to rdump from a
> > failed device cause the tape to fail? or was it something else?
> > -Lucas Carey
Received on Mon Oct 08 2001 - 15:55:29 NZDT