Recovering from massive media errors

[sweater]

Here's what happened:



For some time, we've been getting some read and write errors in TSM when using our 3583. Now it's getting to the point that any new scratch tape gets set to read-only at 0.0% utilization because of a bad drive. We're working with IBM in narrowing that down.



Once we fix this, however, we're stuck with 90+ tapes that we don't trust. Anyone have some suggestions as how to best start moving data onto new media?



Also, are there any easy, brain-dead ways to make sure we catch these errors next time? Tossing another 90 tapes is not in the best interest of our dept...



Thanks in advance for any help!

 

[ndonato]

would that be the same reason my dlt7000 drive resizes the tape from 35 G compressed to 20 G compressed the tape itself is a DLT IV 35 G of course my mentor insists the drive is good. constantly complains bout cleaning too..



-guess

 

[TSMWHO]

If have constanlty been getting read/write errors. If CE states that hardware is ok, then make sure latest Firmware is loaded. Also make sure you are using the latest Ultirum device driviers for the 3583. Use IBM FTP site for that.





For the tapes that have been marked unavailble don't assume that they are bad. Update the volumes to readonly, move the data off, then checkout and relabel the volumes to TSM.

 

[sweater]

<TABLE BORDER=0 ALIGN=CENTER WIDTH=85%><TR><TD><font class="pn-sub">Quote:</font><HR></TD></TR><TR><TD><FONT class="pn-sub"><BLOCKQUOTE>If have constanlty been getting read/write errors. If CE states that hardware is ok, then make sure latest Firmware is loaded. Also make sure you are using the latest Ultirum device driviers for the 3583. Use IBM FTP site for that.</BLOCKQUOTE></FONT></TD></TR><TR><TD><HR></TD></TR></TABLE>



Got a link? :grin:



We're working two issues here, I think: IBM has failed to prove that our whole library isn't hosed (two drives replaced, firware updated and I still can't use drive0 to write to anything) and we're trying to weed out what tapes are useful and which are not. Our fear is that we had a bad drive for a while and that any tape running through there just got munged up.



Basically, this is a really really bad thing to happen to an organization that's just learning/deploying TSM...

 

[heada]

ftp://service.boulder.ibm.com/storage/358x/3583/



Not sure which file you need since I don't have any of these kind of drives. Your IBM CE should be able to tell you exactly what you need.



I had a similar problem with 3590 drives. Once the CE can get the drive working correctly (several tapes run through without issue) then perform an audit on each volume that you had problems on before. Any volume with errors, perform a move data on that volume to a new volume.



-Aaron

 

[sweater]

Yeah - spent a large part of yesterday updating firmware (I guess it takes longer through the web interface). No dice.



So we're now onto another drive swap-out, cables, etc. At least one of the drives is working fine, it seems, we're just not able to do any reclamation, etc due to disk space constraints. Otherwise I'd make another storage pool on disk and do that as if I only had one drive. Which, I guess for now, is the case...

 

[sweater]

So this has turned out to be cables - three drive replacements, me building another server, and 2 +1/2 weeks of troubleshooting later and it's the cables. Turns out, IBM wants you to use their cables - no one else's.



So at least I can now run reclamation, start backing up my copypool again, etc...