This post is more than 5 years old
76 Posts
0
6911
November 23rd, 2011 00:00
MDB restore doesn't work.
Hello guys,
My configuration is the following: Win 2008R2 with Exchange 2010 SP1, NWS 7.6.2.4 Build 669 and NMM 2.3.0.81
I have 2 mailbox server with DAG and 2 CAS/HUB.
The backup works fine, not the restore.
When i want restore MDB in RDB i have this error in NMM monitor:
NMM .. Starting Recovery. This operation may take a long time depending on the size of the data requested.
82296:nsrsnap_vss_recover:NMM .. Operation unit failed with error 'br-module: Tape Restore Failed...'.Possible cause: 1)Unsupported file system or 2)write-protected disc or 3)No space on disc or 4)Drive not found
powershell.Invoke Mount-Database failed.
Cannot invoke this function because the current host does not implement it.
81188:nsrsnap_vss_recover:
**************************************************************NMM .. Failed to mount Database [RDB01]. Client will not attempt any further mounts.
Recover completion status
Files failed to be restored for Microsoft Exchange 2010.
Exchange restore completed for MDB01.
**************************************************************
50325:nsrsnap_vss_recover:nsrsnap_vss_recover: Recovery operation succeeded.nsrsnap_vss_recover complete time: 21.11.2011 13:58:06
Recover program exited with status: 0
Restore destination has been reset for Mail Box Database MDB01
The options specified in the Recover Options Dialog have been reset back to their safe default values.
The unit where i restore the MDB is in NTFS format, permissions are good, no protected, not full and healthy.
I wait on support since 1 week...
If you can help me.
Thanks a lot
David.


broeste1
153 Posts
0
May 3rd, 2012 02:00
Hi Carlos,
Nope, the database en logfiles are on seperate path and drives. I have tried deleting everything and doing new restore but with same problems. I have just tried restoring once again and followed the Application/msexchange logs and the restore failed wiht this message
NMM .. Starting Recovery. This operation may take a long time depending on the size of the data requested.
66422:nsrsnap_vss_recover:NMM .. gather writer status after PostRestore returned status failure VSS_WS_FAILED_AT_POST_RESTORE for writer 'Microsoft Exchange Writer'. Writer failure 0x800423f3 -- VSS_E_WRITERERROR_RETRYABLE
67202:nsrsnap_vss_recover:NMM .. After PostRestore ... writer status indicates a failure -- 0x800423f3.
81188:nsrsnap_vss_recover:
NMM .. Failed to mount Database [mailrestore]. Client will not attempt any further mounts.
**************************************************************
Recover completion status
VSS validation and preprocessing failed for Microsoft Exchange 2010 with error 0x800423f3.
Exchange restore completed for Dako Mailbox DB - 10 GB limit.
There is nothing wrong with the vss writers accordingly to vssadmin list writers
Writer name: 'Microsoft Exchange Writer'
Writer Id: {76fe1ac4-15f7-4bcd-987e-8e1acb462fb7}
Writer Instance Id: {24e6d791-1f48-4b2a-a603-1cb7deddbc3b}
State: [1] Stable
Last error: No error
NWadmin1
89 Posts
0
May 3rd, 2012 03:00
Hi,
Alternate database recovery supported only in the same domain.
Cross-domain Exchange recovery is not supported. That is, if two hosts are in different domains, alternate database recovery cannot be performed.
Regards,
Mustafa
CarlosRojas
1.7K Posts
0
May 3rd, 2012 03:00
Hi Thomas,
Can you please let us know if, as Mustafa mentioned, you are trying a cross-domain or cross-platform restore?
Also please check the following KB:
https://solutions.emc.com/emcsolutionview.asp?id=esg127107
Thank you.
Carlos.
CarlosRojas
1.7K Posts
0
May 3rd, 2012 03:00
Hi Thomas,
I guess you are doing backup of passive DB's, right?
I would say that a delay in log replication from Active DB to passive DB could cause some inconsistency on the backup, and I suppose you have disabled the JET errors and DB consistency checks, right?
Thank you.
Carlos.
broeste1
153 Posts
0
May 3rd, 2012 03:00
I not trying cross domain recovery. This is a DAG setup with three physical nodes. The customer has 3 different databases "moving around" in the DAG setup.
I'm trying to do a restore of one of the other databases in the DAG setup and it's running now. It will take a couple of hours before it's
As Carlos posted, this might be caused by some missing exchange logs ind one of the backups. The resolution was to fix the logs issue and then do a database restore again. This is a Networker/Avamar setup and therefore we are doing full backups of the exchange databases everyday. So I would think that logs are being truncated everyday and therefore changing restore date to another should work but it doesn't look like it changes anything
broeste1
153 Posts
0
May 3rd, 2012 05:00
Hi Carlos,
Nope, actually they are backing up active databases. The consistency checks on DAG is disabled per default and has not been enabled. Do you think an consistency check would have discovered this log issue?
broeste1
153 Posts
0
May 3rd, 2012 23:00
Hi,
here is an update.
I tried restoring another database in the sam DAG setup and it worked without any problems, so I'm thinking that there is an issue with either the backed up .edb file or the logfiles. I just don't know how to complete rule out that this might be an NMM issue.
The consistency check would it have caught corruption in either edb or logfiles?
CarlosRojas
1.7K Posts
0
May 5th, 2012 22:00
Hi Thomas,
Yes, Consistency check should have discovered the delay in the logs replay, however you can check yourself the application and system event logs, even the exchange logs (in all nodes), and try to find out the issue with the logs replication.
Anyway you can enable the JET errors, but this can lead to lot:
NSR_CHECK_JET_ERRORS=values of backup failures
Specifies the JET database errors to check for in the event log before taking a snapshot.
Note: If errors are found, the snapshot fails.
Values include the following:
• 447 (Exchange Server 2007 only)
• 448 (Exchange Server 2007 only)
• 1018
• 1019
• 1022
• all
• none
Values can be separated with a comma, for example:
NSR_CHECK_JET_ERRORS=1018,1019
The default is none.
Thank you.
Carlos.