UNSOLVED
'Cannot find volume needed' recover error
Hello,
We are trying to recover a file.
The backup is successful.However we can see in the attached print screen from Networker and for 'mlnh101-backup_DB' Save Set a double entry in messages (and 0 size) .
So the first question is if these double messages is something normal.
Then in /nsr/index we can see for this client 2 indexes (different sizes and creation dates)
drwx------ 3 root root 512 Jul 30 00:02 mlnh101
drwx------ 3 root root 512 Jul 30 01:01 mlnh101-backup
If we try to recover a file from this client the error is shown below
root@mlnh101> recover
Current working directory is /
recover> add environment.txt
/
1 file(s) marked for recovery
recover> recover
Recovering 1 file into its original location
Volumes needed:
Unknown
Total estimated disk space needed for recover is 1 KB.
Requesting 1 file(s), this may take a while...
70582:recover: Requesting 0 recover sessions from server.
53365:recover: NSR server `mlncbr01-backup': Cannot find the volume needed
5766:recover: waiting 30 seconds then retrying
53365:recover: NSR server `mlncbr01-backup': Cannot find the volume needed
Next step was to delete mlnh101-The same error occured
Then we deleted mlnh101-backup and could not find index at all.So we concluded that the right index is the mlnh101-backup.
Running nsrck and mminfo outputs are normal
There are no double client ids in the mminfo for the client and the respective save sets.
Any ideas?I 've tried all the recommendations of previous posts for the same error.
Thanks
Anna
1070
0
Posted July 29th, 2013 10:00
Yes, some 'metadata' save sets (usually from db backups which use a NW module) exist which are almost 0 bytes in size.
Unfortunately you did not specify any details like OS and NW version, so there is a lot guessware possible. So let me guess where you are working ...
What you obviously want to perform is a file recovery. However the save sets you show in the screenshot are for sure not refering to filesystem backups as they would show the "absolute_pathname". These are DB alias names. So the files must belong to another save set.
Let me suggest that you run "volumes" after you marked your files for recovery. This should tell you the volume where these files are stored. It might give you a clue while NW cannot access it.
I think you can 'merge' 2 indexes by extending the client's Aliases list (Globals (1 of 2)) tab.
1070
0
Posted August 26th, 2013 01:00
This is typical issue where there is a disconnect between index and media database caused by configuration issue (like missing alias) or DNS broken. You can do restore in this case using ssid restore, but you should aim in fixing this issue for good. There might be couple of variants of this problem and depending on that, remedy will differ in small details too so engaging your support would be recommended (after that restore from backup will work for backups after issue has been fixed and for previous ones you can use ssid restore when needed).
masonb
445 Posts
1070
0
Posted July 29th, 2013 09:00
Anna,
I don’t see a “double entry” for your client unless you are talking about one entry for the DB backup and the other being the backup of the clients index backup?
Are minh101 and min101-backup separate clients or the same client but these are the aliases for these?
If they are the same client I would suggest you do have clientid issue and therefore may be best contacting EMC Support officially for confirmation. If you have nsrget already loaded on your NetWorker server from a previous support call then please run this and supply the output file to the SR when opening.
If not you can download it from ftp://nwc:nwc@ftp.emc.com.
The zip file is for Windows servers and the tar file for all flavors of Unix.
Regards
Bill Mason