NDMP backups failing with 52/84 media errors

[Junaid_Mohammad]

I have got 5 dedicated tape drives for NDMP on our master server which is running 5.1 MP5. I am finding that our NDMP backups are constantly failing with 52 errors, complaining about media being in a down drive, which is not the case or might get 84 - media write errors. This is only occurring for my NDMP backups and nothing else.

Checking the logs we are still getting media errors on NDMP backups (i.e. 84's) and also lots of errors like the following when a large number of media id's are rejected and then a status 52:-

05/06/2006 05:04:57 - Warning bptm(pid=12797) media id 601468 is unavailable (in a DOWN drive, misplaced, or not available for another reason), attempting retry with a different media id
05/06/2006 05:05:01 - mounting 601159
05/06/2006 05:05:18 - Warning bptm(pid=12797) media id 601159 is unavailable (in a DOWN drive, misplaced, or not available for another reason), attempting retry with a different media id
05/06/2006 05:05:22 - mounting 601555
05/06/2006 05:05:32 - Error bptm(pid=12797) media manager terminated during mount of media id 601555, possible media mount timeout
05/06/2006 05:05:32 - Error bptm(pid=12797) media manager terminated by parent process
05/06/2006 05:05:33 - end writing
timed out waiting for media manager to mount volume(52)


Has anybody see this before and do you have answer

Thanks

Junaid.

[zippy]

Junaid,

Your NDMP server should have a dedicated tape drive for all its backup.

Check your configuration.

http://symantec.atgnow.com/availability/Login.jsp?SearchTerm=NDMP&fcode=FE_51332&displayFcode=NBUESVR&ddOS=&ProductVersion=&show_pdf=1


JD

[Junaid_Mohammad]

I do have 5 dedicated tapes drives, which solely used for NDMP and nothing else.

The problem started to occur recently, in the last couple of months. The system administrator for the NDMP server has recently installed the latest patch for NDMP, and that�s when this first started to occur.

We get Bit threshold exceed on the san, we were getting that before the upgrade, as well, but we never received a 52 error message, backup always completed successfully.

I am confident that the NDMP configuration is correct; we have had backups and restores working successfully for a year.

Do you have any advice?

Thanks

Junaid

[patrickkuah]

can you reconfigure the authentication and retry again??

PK

[Junaid_Mohammad]

PK backups are working for NDMP, but we are occasionally getting 84 or 52 errors. I don't believe performing an authentication will make any difference.

# /usr/openv/volmgr/bin/set_ndmp_attr -list
Record Type: Authentication
NDMP Server:
Username : ndmp
Password :

I do not get any robotic information as below (e.g), does this matter.

Record Type: Robotic Information
NDMP Server: stripes
Device: c2t3l0
Controller: 2
SCSI Id: 3
SCSI LUN: 0

Thanks

Junaid

[Ankur_Kumar]

Dear Junaid,
As per Veritas NDMP remote backup fails with an error 84 cause the medias are write protected but due to a bug in Netbackup and in consideration with remote NDMP backups, the medias could not be freezeed which should be normally done when netbackup encounters a write protect enable media. Since the media is not been freezed netbackup keeps trying mounting and using the media for backup hence it fails with an error 84 or 83 which at the core is very much and 84.

The workarounds to this issue is

a) manually freeze the medias which are write protected.
b) Restrict yourself from using media which are write protected

But what wonders me is the issue was suppossed to be resolved with the release of the patch Netbackup 5.1 MP5.

In reference to the error 52 could you please check the type of medias been used and the category or class of drive been put to service for the backup. I believe its not the problem with the drive but with the medias in question. Whatddya say Bob, looking forward to your suggestion in these regard.


Chiao Come Va
Ankur Kumar

[Junaid_Mohammad]

Ankur,
None of the media are write protected

Thanks

Junaid

[Ankur_Kumar]

Hi Junaid,

The issue lies in the hardware level, that would be the media in use and not anything related to netbackup for NDMP, as well as could you please pass on the bptm log file for the NDMP drives and status code 52.


chiao
Ankur Kumar

[Junaid_Mohammad]

BPTM verbose for the master server is set to 5,

cat log.061206 | grep -i DOWN

11:37:26.306 <8> write_backup: media id 601337 is unavailable (in a DOWN drive, misplaced, or not available for another reason), attempting retry with a different media id
11:37:32.571 <2> set_job_details: LOG 1150108652 8 bptm 23600 media id 601095 is unavailable (in a DOWN drive, misplaced, or not available for another reason), attempting retry with a different media id.

Above is the kind of errors that are being displayed

Thanks

Junaid

[Junaid_Mohammad]

Problem has been resloved, we modfied the SAN switch from 2GB to 1GB.

[Ankur_Kumar]

You should check up the firmware version of the drives, and its compatibily with the uograded patch on the NetApp box.

Chiao Ankur Kumar

[Stumpr2]

Thanks for the update

[Ankur_Kumar]

Hi Junaid,
Thanks for the update, well do check up on the SPEED WRITE feature on the switch ports, if its enabled then ocurrence of 84 errors are pretty much possible.

Chiao
Ankur Kumar

[Junaid_Mohammad]

Problem has been resloved

[Stumpr2]

thanks