cancel
Showing results for 
Search instead for 
Did you mean: 

Backup Exec mark as bad some LTO that are good

Fabio_Grasso
Level 2

Hello, I've a strange problem with my installation of backup exec. I have a new IBM TS3100 (LTO4)library connected (via SAS) with an IBM x3400 (using a dedicated HBA SAS - 25R8060).

On the x3400 I have a fresh W2k3 64bit installation with Symantec Backup Exec 12.5. Both O.S. and Backup Exec are upgraded to the last version, service pack and patch. I have also installed the symantec driver for the tape and upgrade the firmware of library, tape and HBA to the last release.

Everything work correctly, but about once a day the library mark a tape as bad and the Backup Exec stop to write on the tape. If I remove the bad LTO Backup Exec it works again.

Today I have removed the fifth "bad" LTO. It seems strange to me that in 10 LTO cartridges (all IBM original) purchased there are already 5 defective.

Thinking on an hardware problem I've open a ticket to IBM support. They've analyzed my log and told me that writing is blocked but can't understand if it's a physical problem on the tape or whether a software issue (BackupExec detects any error in writing and mark cartridges as bad).

IBM send me a tool (ITDT) that write directly on the tape. With this software there isn't any problem. I've tried to write for about 20 hours without any block. Moreover if I try to write on an LTO that BackupExec has marked as bad ITDT can write on them and after the write process the cartdrige "rises" and work again.

 

Everything makes me think of a question of dialogue between BackupExec and my library.
 
Any idea about?
 
Thank you very much,
      Fabio Grasso
1 ACCEPTED SOLUTION

Accepted Solutions

Fabio_Grasso
Level 2
Hello, sorry for my delay. I've finally solved the problem: IBM has send me a new LTO tape 10 days ago. After having replaced the backup returned to work.

I've write on tape about 10TB since the replace so I can tell that the problem has resolved.

Thanks to everybody for the support!

View solution in original post

10 REPLIES 10

Johanvdv
Level 4

Fabio,

Your description is a very good match of the problem I am currently investigating.

I have a Dell TL4000 library and since about a month or so, numerous tapes have been marked as 'bad'.

I frequently get write errors on these tapes. Retrying or using brand new tapes (Sony) does not resolve my issue.

I have run a test using EMC Retrospect 7.6 and that did NOT report any problems with the tapes that were flagged as bad by Symantec BE 12.5.

I currently have Dell support looking into hardware problems (ITDT diagnostic program) and Symantec support doing test backups.

 

I too have suspicions that the last driver upgrade by symantec (SP1 now at Rev 2213)  is causing this problem.

 

 

 

Larry_Fine
Moderator
Moderator
   VIP   

I assume some backup jobs are failing when the tape gets marked as bad.  What errors are in the job log?  Have you checked the Windows event logs for anything relevant?

Johanvdv, are you using LTO4 SAS also?

 

Johanvdv
Level 4

The errors that are reported to me by BE are :

...

Backup- D:
Storage device "IBM 3" reported an error on a request to write data to media.
Error reported:
A tape read/write error has occurred.  This is usually caused by dirty read/write heads in the tape ...
V-79-57344-34028 - A tape read/write error has occurred.  This is usually caused by dirty read/write...
...

 

There is absolutely nothing relevant in the eventviewer (no event 5/7/...).

And yes, my drive is an LTO4 SAS. A Dell TL-4000 is actually a rebranded IBM drive. 

 

Today, Dell replaced my tape drive. I am again rerunning my large job to validate any issues as originating from the drive.

 

Symantec support has until now not been helpfull in tracking the issue : they just keep insisting that the problems are related to a 'bad job' and want me to recreate the job (and selection list) and then try again.

I have already gone down that route using a small job and a large one.

At present they want me to run a large backup to a non-tape location (disk to disk using internal or external disks). Problem is that I do not have a 4TB test drive available.

The support engineer also claims that they have no real way to request more detailed info or traces on the write failure.

Since Dell agreed that the tape drive is showing issues, Dell replaced it within the support plan.

I will report back in a couple of days on the results so far.

 

 

Larry_Fine
Moderator
Moderator
   VIP   
You should be able to use the SCSI tracer to get more details about the error and confirm what the drive is reporting.
http://seer.entsupport.symantec.com/docs/285630.htm

teiva-boy
Level 6
 I wouldnt be using the Symantec drivers, but that is just me..  I personally want the hardware vendor ones that actually provide good performance, rather than conservative reliability.


Ben_L_
Level 6
Employee
Johanvdv,

Can you post a bediag if the problem comes back up so we can review the configuration? Or provide me with your case number so I can review the data.  I would suggest private messaging me the case number, but I haven't figured out how to do that with the new forums yet :)

Larry_Fine
Moderator
Moderator
   VIP   
Is BE marking the media as bad (or retired)?   Or does it just fail the job?

What happens if you move a retired tape into scratch and run a quick erase on it? 

Ken_Putnam
Level 6

Johanvdv,

Can you post a bediag if the problem comes back up so we can review the configuration? Or provide me with your case number so I can review the data.  I would suggest private messaging me the case number, but I haven't figured out how to do that with the new forums yet :)

Click on the user name
In the new window, select the bottom choice "Send this user a messag"
(now if I could only find smilies  |8^)

Fabio_Grasso
Level 2
Hello, sorry for my delay. I've finally solved the problem: IBM has send me a new LTO tape 10 days ago. After having replaced the backup returned to work.

I've write on tape about 10TB since the replace so I can tell that the problem has resolved.

Thanks to everybody for the support!

Johanvdv
Level 4

Similar to Fabio, I got a replacement tape drive and this fixed the issue.

I did not have to make any changes to my BE jobs.

Please note that in my case the tapes were not being marked as 'bad' or 'retired'.

I have run two cycles (totalling more than 6TB) and did not have any media errors (soft or hard). I am now re-testing the tapes that were logged with media errors by doing a full format on them an re-using these tapes.