cancel
Showing results for 
Search instead for 
Did you mean: 

NBCC Error CONSISTENCY_ERROR Oper_12_1

Spartacus81
Level 6
Partner Accredited

Friends,

Facing NBU catalogue inconsistency in one my production enviornments see below. any suggesstions will highly be appreicated apart from logging the call with Symantec.

5.0  Beginning NetBackup catalog consistency check
       Start time = 2013-12-17 14:56:13
 
5.1  There were no tape media involved in active NetBackup jobs
5.3  Processing EMM database Volume attribute records, pass 1 (of 2),
     198 records to be processed
       Processed 198 EMM database Volume attribute records.
5.4  Checking for duplicate EMM server host names in Volume
     attribute data
5.5  Processing Image DB, pass 1 (of 2),
     5995 images to be processed
       5995 images processed on pass 1
       There were 0 images with at least one copy on hold detected.
5.6  Processing EMM database Media attribute records, pass 1 (of 3),
     108 records to be processed
       Processed 108 EMM database Media attribute records.
       There were 0 tape media detected that are on hold.
5.8  Check for duplicate media server names in the EMM database
       Media attribute data
5.9  Processing EMM database Media attribute records, pass 2 (of 3),
      108 records to be processed
5.10  Processing Image DB, pass 2 (of 2),
     5995 images to be processed
CONSISTENCY_ERROR Oper_12_1
 
5.11  NetBackup catalog consistency check completed
       End time = 2013-12-17 14:56:15
1 ACCEPTED SOLUTION

Accepted Solutions

mph999
Level 6
Employee Accredited

Oper_12_1 means that there were images found that had missing fragments, that is, the first fragment in the list of fragments for the image was not frag number 1.

I'd log a call and let NBCC do all the work personally ...

View solution in original post

9 REPLIES 9

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified

.... apart from logging the call with Symantec.

 

That is unfortunately the only option.

Only Symantec Support engineers have access to the tools to analyze the output files and suggest ways to fix it.

As a matter of interest - which nbcc version did you use?
How may output files were created?
If I remember correctly - there should be a text file with a bit more info that the output produced by the script.

 

Mark_Solutions
Level 6
Partner Accredited Certified

I have been doing this for a customer recently and it appears that nbcc deliberately doesn't tell you anything within its output files - you can sometimes google the code to get an idea (i have fixed most issues on my site that way plus a little help from friends) but the only real way is to log a support call and let them analyse it with the nbcca tool

You can try the -upgrade switch as well - at least that gives you an idea of it is it is "serious" or not.

The real rule with running nbcc is to only do it when absolutely nothing is running and no tapes mounted etc. - also with no tapes in a frozen state as that can throw and error

 

Yours looks to stop at the end of the emm inage checks which indicates that there is an issue there - it could be an image no longer owned by a media server or it could be a corrupt image - see if there has been a dbcorrupt file created in the db folder

Hope this helps - really you need to log a call to get it analysed

mph999
Level 6
Employee Accredited

Oper_12_1 means that there were images found that had missing fragments, that is, the first fragment in the list of fragments for the image was not frag number 1.

I'd log a call and let NBCC do all the work personally ...

Spartacus81
Level 6
Partner Accredited

Thanks Guys.. 

@Marriane the script created 17 files and the extract in my post is out of the file name nbcc-info.txt

nbcc version is 7.5.0.3

The file you are talking about have no details other than the error code..

@Mark thanks -  i didnt try the --upgrade switch when i was at customer site yesterday as i wasnt sure the seriousness of the issue. 

@MPH you made a good point.. after speaking to my customer prior to visiting him he said he did manaully remove some images from .../db/images area in order to free up some space and i was like SERIOUSLY i think what you are saying is make more sense. so the obvious question is how can i identify the images that has missing fragments? 

 

Mark_Solutions
Level 6
Partner Accredited Certified

i think the bpdbm log will list these, but maybe only when they actually come to expire - having said that with the items misisng all together i am not sure even that would be the case!

was there a corrupt file under the db directory?

I really think this needs to go to support to have it fully checked out

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified

bpdbm -consistency may also help.

See this forum post for details: https://www-secure.symantec.com/connect/forums/how-use-bpdbm-consitency

mph999
Level 6
Employee Accredited

If theer are only fragments missing, bpdbm -consistency won't pick these up ...

The way to identify them is run nbcc ...

It is can very difficult to fix nbcc issues without nbcc.

You are on 7.5.x - so the header files are in the catalog ...

So, the image_fragment table (at an educated guess) is missing fragment lines for some of the images.

That isn't going to be fixed without NBCC ...

 

Mark_Solutions
Level 6
Partner Accredited Certified

That is a good point Martin .. if it is a 7.5 install and the image files were deleted at some point to free up space - then maybe that happended before the 7.5 upgrade?

If so then under \veritas\netbackup\ should be the db.corrupt file created during the upgrade listing the bad images

That would be a handy way of finding them all if it does exist

mph999
Level 6
Employee Accredited

They might not be corrupt ...

If you have a image spanning two tapes ... and phase 1 import the second tape first.  Then you will end up with  perfectly valid (but not yet complete) entries for the image in the image table that is 'missing' fragment 1.

I don't think this is about missing image 'header' files (or their in DB equiv in 7.5) it's about just missing FRAG lines, or in 7.5, entries in the image frag tables.

No matter what you can work out, the only way this is going to be fixed would be via nbccr.

M