Hi,
We are experiencing a higher than desired rate of hard errors with
TypeIV DLT tapes. We are getting hard errors on about 1 to 1.5 tapes
for every ten tape we use. NSR will write to the tape until it hits an
error and will then mark the tape as full and then move onto the next
available tape in the jukebox. This leaves me with a number of tapes
in the jukebox containing only 300MB-5GB out of a theoretical 40GB
per tape.
I know it is possible to clone the contents of the damaged tape onto
another tape, and delete the damaged one out of the indexes prior
to its removal. Unfortunately this is a fairly time consuming process
and it deals with 'live' data.
What I would like to do it to try to catch this problem before we use
the tapes for backups and data archiving. Has anyone got any ideas on
how to verify the physical tape media? The tapes in question are
TypeIV DLT tapes in a TL812 jukebox controlled by NSR.
I have thought about doing an NSR backup to the tapes to be tested,
using local data drives and a tape pool that doesn't store its indexes
on disk. The tapes could then be re-labeled for production use, but
this is about as involved as cleaning up after the tapes with errors.
Does anyone have any better ideas?
BTW - Our fail rate with TypeIII tapes is about 1 out of 40. (Except
for a batch of bad tapes that we recently got which would fail on label
about 20% of the time.) The drives in the TL812 appear to be fine and
are cleaned regularly.
Thanks in advance,
Tom
--
+--------------------------------+------------------------------+
| Tom Webster | "Funny, I've never seen it |
| SysAdmin MDA-SSD ISS-IS-HB-S&O | do THAT before...." |
| webster_at_ssdpdc.mdc.com | - Any user support person |
+--------------------------------+------------------------------+
| Unless clearly stated otherwise, all opinions are my own. |
+---------------------------------------------------------------+
Received on Sun Aug 03 1997 - 02:39:56 NZST