Solved

When doing restore error "The index cannot be accessed"

  • 3 March 2022
  • 8 replies
  • 2037 views

Badge +1

When trying to do a restore I get the error “The index cannot be accessed. Try again later. If the issue persists, contact support.” I checked the Data Recovery Job History and the errors state it’s trying to restore the index and the media (tape) isn’t available. The tapes it is looking for are bad and no longer exist.

This happens for jobs going to tape or disk. I would be fine with deleting the index and starting over, but I can’t figure out how to do that either.

Thanks for the help.

icon

Best answer by Hyder 4 March 2022, 23:23

View original

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

8 replies

Badge +1

If I understand correctly, since we are now using a new MA and new disk storage instead of tape a new index will be rebuilt that saves to disk instead of tape. Over the next couple weeks with the backups being done and the index being rebuilt everything should return normal. Is that correct?

Userlevel 3
Badge +8

Client backups will dictate primarily what index gets generated, the MA then decides where to write it, in your case its just 1 MA so its all-in-one-location. Are you using V1 or V2 indexing?

In either case the index is locally written to the MA as it has classically in both indexing versions, then it gets written to the media, mag lib and/or tape. We recall the index if its non-accessible when you’re doing a browse or need it for a backup job.

Badge +1

I should give more detail. I recently added a 2nd media agent and disk storage so we can backup directly to disk and then aux copy to tape. Before we were just doing backup to tape.

This shows my lack of depth in the CV backup system. Can’t we control where an index is stored. I also thought the new media agent/disk storage/storage policy would create a new index. Is that wrong?

 

Userlevel 3
Badge +8

If I understood you correctly, in this case the media is not readable, if that’s the case in future it would be the same, or am I not getting what you mean?

Index is written to tape so it can always be rebuilt different location or different MA

 

Badge +1

@Hyder even if I can do a “restore by jobs”, I’m concerned it will impact future jobs as well. How do I rebuild the index to a different location.

Userlevel 3
Badge +8

(intelligentDataAgent) File System, Exchange etc

If you have bad media this is a nightmare for sure to sort thru, if this is a FS iDA then you can dump the entire job out, irrespective of the index requirement as that just gives you Browse options.

Do a RestoreByJob ← this could get you partial data but at least it won’t tie you up regarding indices.

https://documentation.commvault.com/11.24/expert/12983_restore_by_jobs.html


 

Badge +1

Thanks for the response @Hyder . I’m not sure what you mean by iDA type. I’ve tried restoring from tape and disk both, if that is what you mean. We’ve have several tapes that are bad and we had to retire them.

I was able to add some of the tapes the index job was asking, but some are gone. With these tapes added I did get a different error, “ The end of data marker was reached during an operation “ Error code 40:78

I’d really like to find some way to start over on the index and have it on local disk. I don’t like being dependent on another tape when I try to do a restore.

Thoughts?

Userlevel 3
Badge +8

Hi @tobeyal what iDA type are you restoring data from. Also is this a job which is still on the Storage Policy Copy’s job history or is it aged? Just curious how do you know the media no longer exits, was it manually removed or something else?