Solved

Failed to process chunk since 11.24.19 upgrade

  • 23 February 2022
  • 4 replies
  • 2716 views

Userlevel 1
Badge +5

Yesterday our environment did an automatic upgrade from 11.20.xx to 11.24.29.

Previously I have had absolutely zero issues with data verification and backups and aux copies and all jobs show as passing data verification.

Now I’m seeing errors/issues on aux copies and data verification along the lines of:


Failed to process chunk [2840311] in media [V_98073], at the time of error in library [Disk Library] and mount path [[ma02] C:\Mount\maglib04], for storage policy [Backup to Disk] copy [Primary] MediaAgent []: Data read from media appears to be invalid..

We’re using global dedupe and lots of synthetic full backups so I’m trying a proper full backup on a problem job to see what happens.

Before I raise with support does anyone have any suggestions or thoughts please?

This is a Windows MA environment where there have been zero issues this literally started to happen after the 11.24.29 install.

Obviously this doesn’t seem good :(

icon

Best answer by Mike Struening RETIRED 22 March 2022, 21:36

View original

If you have a question or comment, please create a topic

4 replies

Userlevel 7
Badge +23

Hey @Paul Hutchings , hope all is well!

Assuming the Media Agents are upgraded as well, I would suggest you open a support case right away.  You can take a look at the path in the message to confirm the chunk file/folder is still there in that Volume folder, though even if it is there, that doesn’t solve anything for you.

It COULD be one job, but before marking any job to not copy, you should have support investigate.

Can you share the case number was created?

Thanks!

Userlevel 1
Badge +5

Hi Mike, thank you, it’s ticket 220223-575.

Seems new incremental jobs are running and aux copying fine but I haven’t run a new synthetic full or full on a problem job yet.

Seems to occur at random and whenever I look the chunk file seems to be there with a .idx file.

This is something that has literally only happened after the 11.24 upgrade went on.

Userlevel 2
Badge +5

As a side note and perhaps something to look at, can you make sure that the MAs can talk to each other without any issues? can you use cvping between them?

I have had this issue in the past and one of the things we discovered was that we had a group “ firewalled”  in CV with a specific setting applied that would occasionally cause communications issues.

That is one, the second thing I would check is the last time you ran a DDB Verification FULL job against the DDB. Reason I ask is because DDB verification jobs depending on how you run them, they make sure that new backups only refer to valid blocks.

https://documentation.commvault.com/11.24/expert/12398_data_verification_online_help.html

The support ticket is a great idea to make sure and address the root cause here.

 

Userlevel 7
Badge +23

Sharing the solution:

A full DV2 was run and any failed jobs marked bad which resolved the issue.