In a previous article I wrote:
> Due to an inadvertant error (despite my best efforts), I have a tape
> full of data which was written over with nothing (just two EOFs).
>
> Is there ANY way to force an Exabyte 8500 to skip over this to the
> data on the tape? Any utilities, program source, or even physical
> procedures (short of tape splicing) would be appreciated (although
> I would prefer authoratative answers from anyone who has actually
> done this -- urban legends abound.)
>
> Please reply via email; I will summarize.
Essentially, the answer is "no."
(But see below for a solution to the problem.)
I got three replies, and an answer from my vendor.
"root@pauli.phys.washington.edu" said, there is no way. He then suggested
variations on the ideas above (including power cycling the drive while
it is still writing, so no EOF, and splicing tape).
"system@alchemy.chem.utoronto.ca" suggested I contact our vendor (OSS),
as he had heard they might have a PC program to change drive parameters
to permit fsf past EOT. (They didn't know anything about that.)
"scott@geom.umn.edu" said he had contacted Exabyte, and they had told him
to contact a data recovery service.
There is no substitute for backups, and data recovery services exist
for a reason. :-(
However, when I contacted our vendor, they suggested I contact Exabyte,
and that they will recover data from tapes. When I contacted them
myself, I was told that it was free of charge, but was considered
a "favour", as opposed to a "free service" -- I cannot emphasize enough
the difference: if this service gets abused, it will go away. If you
have backups, recover from backups. This should only be used as a
last resort. If you *need* to do this, contact them yourself, and ask
nicely. And don't tell where you heard about this. ;-)
-- Brian Dickson | University of Toronto Computing & Communications brian@canet.ca | Network & Operations Services +1 416 978 0993 (voice) | External Network Facilities Managment +1 416 978 6620 (fax) | 4 Bancroft Ave., Rm 101, Toronto, ON M5S 1C1
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:09:16 CDT