Read beyond EOT rst8?
Yukon Kid
jae at c-art.UUCP
Sun Dec 10 05:36:29 AEST 1989
OK, suppose I inadvertently, somehow, put an EOT at the beginning
of a valuable QIC-24 tape. (IDIOT! I know, I know %^) I REALLY
need some of the information previously recorded by
tar cvpf /dev/rst8
on this tape. Is there any way I can get at the data beyond what
looks to be an EOT at the very beginning of the tape? The tape
behaves as if I've overwritten an rst8 tar tape, but with a
zero length file, or something. My only feedback, whether
tar [rst8/0], dd, or mt, is:
tar: tape read error: I/O error
or variations, with `I/O error' consistent. mt's fsf 1
is not useful. dd with conv/noerror tries forever to get
past whatever's been put on the tape, but doesn't get anywhere.
(If I use /dev/nrst8 with this, and afterwards do a rewind, there
is no rewind. It hasn't got anywhere. I/O error keeps happening
ad infinitum.) retension doesn't help.
I wouldn't betray my foolishness unless it were important to
me. I do know there is good data beyond the start of the tape.
(My guess about how an EOT, if that's what it is, got at the front
of the tape is that my drive, after a cleaning, wouldn't write at all
for a short time, and I had tried to write on this IMPORtant [financial
backup!] tape.) My tape drive itself seems to be fine now - reading,
writing my own & other machines tapes.
I am quite willing to physically cut the tape, or mis-align the head,
or whatever. I guess the rst8 format writes all over the tape, and
it maybe is impossible to synch up with beginning of block (?), but
I have to ask. Anybody got an idea?
Thanks for anything,
-john
Ps: Oh, if I can get raw data in my machine via dd, or something,
I would be HAPPY to cut it up further using dd & od, etc.
Pps: Having duplicated the problem on a throwaway tape (I THINK),
I try to write an `eof', thinking to overwite an `eot', I get
hardware error - st0: illegal command.
John Eadie Computing Art Inc (416) 536-9951
e-mail: jeadie at sun.COM | jae at c-art.UUCP | {uunet,suncan}!c-art!jae
`these deepest problems are in fact *not* problems at all.' - Wittgenstein
More information about the Comp.unix.questions
mailing list