On Mar 25, 2010, at 5:31 PM, Len Philpot wrote:
> 
> That might explain some tape errors we've seen over the entire time we've 
> been running (Sun's OEM-ed) 7.2. I know... it should be replaced by 7.5.x 
> in a few weeks hopefully! We started seeing them immediately after 
> upgrading from 6.1 to 7.2, on the same hardware, same OS, two different 
> servers, etc., etc.
> 
> Networker would write just fine to a tape and then with a completely 
> random frequency fail to write a subsequent backup to the same tape at 
> some later point. It would load the tape, seek forward, keep on seeking 
> until it reached end of media, rewind and try a few more times. Usually, 
> it would then eject the tape, get another and backup successfully. Usually 
> (!) the delays introduced by such issues didn't cause problems.
> 
> My empirical theory (supported in concept by Sun and neither denied nor 
> ever fixed by EMC) was that some kind of final "end of data" marker was 
> not being written properly in some instances. This might explain the 
> /complete/ randomness of the errors: new tapes, old tapes, one pool, 
> another pool, ad infinitum, if that second EOF wasn't being reliably 
> written to tape. It also makes me feel better that any previous backups 
> written to said tapes are probably OK and shouldn't present issues if 
> being read for clones, etc. 

I forget the details, but a while ago, I vaguely remember that EMC issued an 
alert to install a NetWorker update for a very similar problem. If you search 
PowerLink for something like "EOF NetWorker" you might find the info.

To sign off this list, send email to listserv AT listserv.temple DOT edu and 
type "signoff networker" in the body of the email. Please write to 
networker-request AT listserv.temple DOT edu if you have any problems with this 
list. You can access the archives at 
http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER