Mailbox Store Will Not Mount After Backup Exec Mailbox Restore
We are getting ready to retire an old Exchange Server and I need to keep as much email data as possible for the Freedom Of Information Act. I have been restoring old backups to gather as many emails in to the mailboxes as possible before I Exmerge
them all out to .PST files. Everything was going fine until today when I was doing the last restore job. I am using Backup Exec version 11d. The restore job was successful but now I cannot mount the mailbox store again. When I try I
get a popup error that reads "The name of one of the database files in the store is invalid". Below that is "ID no: c1041731". It also logs the following errors in the application event log:
Event Type: Error
Event Source: ESE
Event Category: Logging/Recovery
Event ID: 454
Date: 10/25/2010
Time: 9:37:40 PM
User: N/A
Computer: EXCHANGE-SERVER
Description:
Information Store (4232) First Storage Group: Database recovery/restore failed with unexpected error -1022.
For more information, click
http://www.microsoft.com/contentredirect.asp.
Event Type: Error
Event Source: MSExchangeIS
Event Category: General
Event ID: 9518
Date: 10/25/2010
Time: 9:37:40 PM
User: N/A
Computer: EXCHANGE-SERVER
Description:
Error Disk IO error starting Storage Group /DC=local/DC=PAGECOVA/CN=Configuration/CN=Services/CN=Microsoft Exchange/CN=County Of Page Organization/CN=Administrative Groups/CN=First Administrative Group/CN=Servers/CN=EXCHANGE-SERVER/CN=InformationStore/CN=First
Storage Group on the Microsoft Exchange Information Store.
Storage Group - Initialization of Jet failed.
For more information, click
http://www.microsoft.com/contentredirect.asp.
Event Type: Error
Event Source: MSExchangeSA
Event Category: MAPI Session
Event ID: 9175
Date: 10/25/2010
Time: 9:38:23 PM
User: N/A
Computer: EXCHANGE-SERVER
Description:
The MAPI call 'OpenMsgStore' failed with the following error:
The Microsoft Exchange Server computer is not available. Either there are network problems or the Microsoft Exchange Server computer is down for maintenance.
The MAPI provider failed.
Microsoft Exchange Server Information Store
ID no: 8004011d-0526-00000000
For more information, click
http://www.microsoft.com/contentredirect.asp.
The server is loaded with Windows Server 2003 Standard with SP2. It is running Exchange 2003 version 6.5.7638.1.
If anyone can provide some insight in to what is wrong or how I can go about trying to repair it I would be very grateful.
Thanks,
Marty
October 25th, 2010 9:53pm
Could be that the database file in question is corrupted OR the disk you are attempting to restore to has I/O issues, i.e. I have seen people try to restore to cheap USB drives and sometimes it works but more often it fails so check the destination hardware
I would also clear out the destination to be free from other files to ensure nothing is getting commingled from another EDB
Push come to shove you can also try a repair i.e. ESEUTIL /P against that database but thats a last ditch effort and definitely ensure you run against a good disk
Check out these links as well
http://support.microsoft.com/kb/810198
NOTE: 1022 is a JetIO_Error
http://support.microsoft.com/kb/266361/EN-US/
More importantly
Be aware that ExMerge only supports the ANSI PST Format which has a 2GB limit
ExMerge will not warn you if you export a large mailbox and the PST goes over 2GB, it will just corrupt the PST and you won’t know it until you try to
open the PST at a later date and find that its corrupt and the data is gone
ANSI PST's actually start to degrade at about 1.8GB so I would not go over that limit in order to ensure the data is recoverable
To overcome this, check out Lucid8's DigiScope
http://www.lucid8.com/product/digiscope.asp since it supports UNICODE PST files up to 20GB and can export data from raw offline EDBs for Exchange 2010, 2007, 2003, 2000 and 5.5. Lots
of other cool features and capabilities in DigiScope as well. Here is a quick flash demo
http://l8downloads.com/FLASH/DSWEB/DigiScope_Overview.html
Troy Werelius
www.Lucid8.com
Free Windows Admin Tool Kit Click here and download it now
October 26th, 2010 10:30am
So should I run a CHKSDK on the hard drive to see if there are any hardware issues? We inherited this server and of course someone set it all up on the OS drive array.
October 26th, 2010 10:03pm
So everything is on one drive or? Maybe outline your system setup for me i.e. hardware, # drives, internal external OS etc
I would probably look at the event logs first to ensure all is well hardware Then ensure you have a good DR based backup in case things take a turn for the worse
Then if the event logs are clear run a CHKDSK because I have seen hardware on the edge die and things get worse by doing a CHKDSK. Not saying that will happen, but since you inherited the system and things are in an unknown state I would check
the health of the system all the way around, ensure you have a backup to roll back to etc
Troy Werelius
www.Lucid8.com
Free Windows Admin Tool Kit Click here and download it now
October 26th, 2010 10:11pm
Yes everything is on one drive array. Older IBM rack server. RAID 1 drive array (internal). Windows Server 2003 Standard (up to date SP).
The event logs do not indicate anything about hardware failures.
I will pull a complete DR backup of the machine. Would a built-in Windows backup be good enough or do I need to use Backup Exec?
October 26th, 2010 10:23pm
Whatever backup you feel most comfortable with, i.e. if the machine implodes what backup are you going to trust to recover from?
I assume that you followed the other steps above, i.e. cleared out the destination directories of older logs, EDB etc
Have you looked all all errors and warnings in the application log as well? Do you have lots of disk space on the LUN you are telling BE to recover to? Remember that it could be the database you are trying to recover is the problem as well
Does the Database you are restoring to RSG still exist i.e. is it still in operation on this server or?
Troy Werelius
www.Lucid8.com
Free Windows Admin Tool Kit Click here and download it now
October 26th, 2010 10:46pm
Hi Marty,
Please run the Eseutil /mh and post the results here.
Eseutil /M File Dump Mode
http://technet.microsoft.com/en-us/library/aa997795(EXCHG.65).aspx
If it is dirty shutdown, you may need to run the eseutil /p.
Eseutil /P Repair Mode
http://technet.microsoft.com/en-us/library/aa996773(EXCHG.65).aspx
Frank Wang
October 27th, 2010 2:34am
Microsoft(R) Exchange Server Database Utilities
Version 6.5
Copyright (C) Microsoft Corporation. All Rights Reserved.
Initiating FILE DUMP mode...
Database: c:\program files\exchsrvr\mdbdata\priv1.edb
File Type: Database
Format ulMagic: 0x89abcdef
Engine ulMagic: 0x89abcdef
Format ulVersion: 0x620,11
Engine ulVersion: 0x620,11
Created ulVersion: 0x620,9
DB Signature: Create time:04/11/2006 03:15:46 Rand:2126350 Computer:
cbDbPage: 4096
dbtime: 938150156 (0x37eb090c)
State: Dirty Shutdown
Log Required: 54437-54440 (0xd4a5-0xd4a8)
Streaming File: Yes
Shadowed: Yes
Last Objid: 156507
Scrub Dbtime: 0 (0x0)
Scrub Date: 00/00/1900 00:00:00
Repair Count: 0
Repair Date: 00/00/1900 00:00:00
Old Repair Count: 0
Last Consistent: (0xD434,278B,1A5) 12/23/2009 08:45:20
Last Attach: (0xD434,2794,1DB) 12/23/2009 08:47:49
Last Detach: (0x0,0,0) 00/00/1900 00:00:00
Dbid: 1
Log Signature: Create time:04/11/2006 02:43:34 Rand:98257 Computer:
OS Version: (5.2.3790 SP 2)
Previous Full Backup:
Log Gen: 54294-54298 (0xd416-0xd41a) - OSSnapshot
Mark: (0xD41B,8,16)
Mark: 12/22/2009 18:51:01
Previous Incremental Backup:
Log Gen: 0-0 (0x0-0x0)
Mark: (0x0,0,0)
Mark: 00/00/1900 00:00:00
Current Full Backup:
Log Gen: 0-0 (0x0-0x0)
Mark: (0x0,0,0)
Mark: 00/00/1900 00:00:00
Current Shadow copy backup:
Log Gen: 0-0 (0x0-0x0)
Mark: (0x0,0,0)
Mark: 00/00/1900 00:00:00
cpgUpgrade55Format: 0
cpgUpgradeFreePages: 0
cpgUpgradeSpaceMapPages: 0
ECC Fix Success Count: none
Old ECC Fix Success Count: none
ECC Fix Error Count: none
Old ECC Fix Error Count: none
Bad Checksum Error Count: none
Old bad Checksum Error Count: none
Operation completed successfully in 3.0 seconds.
Free Windows Admin Tool Kit Click here and download it now
October 27th, 2010 8:02am
At this point I trust the Windows backup more than BE. All files in the MDBDATA folder are current. No warnings or errors in the application log besides the ones already mentioned.
Plenty of disk space. Very good point. Yes the database does still exist. I have been restoring to it from old backups for the last couple of weeks and then exporting the mailboxes.
October 27th, 2010 8:08am
Since this was restored from online backup the database is in dirty shutdown mode. you need logs from Log Required: 54437-54440
1. what path did you specified for Restore.env prior to start restoring the database? by default it goes to c:\temp\folder
if you find so then you need to follow certain steps to replay those logs to database.
Let me know if you find that folderVinod |CCNA|MCSE 2003 +Messaging|MCTS|ITIL V3|
Free Windows Admin Tool Kit Click here and download it now
October 27th, 2010 8:16am
This DB is definitely in a dirty state and you may end up having to run a ESEUTIL /P as outlined in my original response step 4 and Step 5 of my last post
HOWEVER ALL backups are in a dirty state when taken and upon restore they attempt to rollup and put them into a clean state, hence my concern over the 1022 which is an I/O based error. Even though you had previous backups that recovered fine to
same destination this backup may be hitting a bad block upon restore or you developed a new HD issue at this time. This is reaching for sure but the 1022 I/O error is basically saying its having trouble accessing the data because of disk I/O related
issues OR because the DB is really damaged. if you want to confirm if its system related or the backup of the files are corrupt
Clean out the destination directory to ensure nothing is commingled from another backup upon restore
Ensured there are no disk related errors i.e. bad blocks etc and for this you may want to run CHKDSK. Then restore the data again
If you receive the same results then yes the DB itself contained corruption during the backup process or the backup process hiccuped at some point so I would then run a ESEUTIL /P against the database to see if its salvageable
Troy Werelius
www.Lucid8.com
October 27th, 2010 8:19am
I was restoring a handful of mailboxes, but not the entire store.
Free Windows Admin Tool Kit Click here and download it now
October 27th, 2010 11:44am
ok but in order to do that you had to
restore the EDB and logs to the RSG. use the MS tolls to extract the specific mailboxes to PST or to merge into live exchange server
Was this what you were doing or??? Troy Werelius
www.Lucid8.com
October 27th, 2010 12:05pm
When using backup exec it stages the entire database and restores mails/mailboxes back into the production store.
/MartinExchange is a passion not just a collaboration software.
Free Windows Admin Tool Kit Click here and download it now
October 27th, 2010 12:20pm
Honestly I am not sure of the actual process with this latest version of BackupExec I have been using. I was just using the Exchange agent to restore the mail for certain users from different tapes. I know in order for BE 11d to restore a single
mailbox it has to load up the entire database first.
October 27th, 2010 12:21pm
Well I want to ensure we are talking about the same thing here i.e.
Are you using BE to do granular restores via their GRT technology Or are you using BE to restore the EDB, Logs etc to the RSG and then using MS tools to export to PST, recover etc?
Troy Werelius
www.Lucid8.com
Free Windows Admin Tool Kit Click here and download it now
October 27th, 2010 12:42pm
Whatever BE 11d uses by default with the Exchange agent. Using Exmerge to export to .PST.
October 27th, 2010 1:46pm
ok then yes my current comments stand i.e. to validate if this is a DB or DISK issue I would
Clean out the destination directory to ensure nothing is commingled from another backup upon restore, i.e. remove all logs, EDBs etc from the RSG path
Ensured there are no disk related errors i.e. bad blocks etc and for this you may want to run CHKDSK. Then restore the data again
If you receive the same results then yes the DB itself contained corruption during the backup process or the backup process hiccuped at some point so I would then run a ESEUTIL /P against the database to see if its salvageable
If you want to super validate this then repeat steps # 1 and 2 with a previously recovered database. If it works then its the database if not its the environment/disk etc
Troy Werelius
www.Lucid8.com
Free Windows Admin Tool Kit Click here and download it now
October 27th, 2010 1:50pm
When I try to run eseutil /p it tells me that the database is not up to date. If I run eseutil /r to try and recover it and bring it up to date it tells me that the current log file is missing. I am thinking I should go back and run the restore
process again. Sound logical?
October 27th, 2010 3:29pm
Yes and follow steps 1-4 above as discussedTroy Werelius
www.Lucid8.com
Free Windows Admin Tool Kit Click here and download it now
October 27th, 2010 4:38pm
I tried that process and still end up with the same result. Then when I go to run ESEUTIL the repair option tells me the database is not up to date. The recovery option tells me that a log file is missing.
November 4th, 2010 7:48am
Yes BackupExec does use the c:\temp\ folder for making copies of the logs during the restore process. The log files numbers make no sense though. All of my log files start with E00...
Free Windows Admin Tool Kit Click here and download it now
November 4th, 2010 7:49am
Does it tell you which log file is missing? If the log is truly not there then you will need to run an eseutil /p Troy Werelius
www.Lucid8.com
November 4th, 2010 8:00am
Here is the line about the log file:
"Log Required: 54437-54440 (0xd4a5-0xd4a8)"
But there are no log files named even close to that.
Free Windows Admin Tool Kit Click here and download it now
November 7th, 2010 9:37pm
If the required logs are truly not there then you will need to run an eseutil /p
Troy Werelius
www.Lucid8.com
November 8th, 2010 2:28pm
When I try to do that I get the following error:
"Operation terminated with error -1022 <JET_errDiskIO, Disk IO error>
Free Windows Admin Tool Kit Click here and download it now
November 10th, 2010 7:13am
ok so this points to two possible issues;
The disk that you are running this database on has some serious contention because of speed of the media or possibly the media is going bad which is causing an I/O issue. So what type of disk did you recover this DB onto?
The database is seriously damaged and unrecoverable. To confirm the above I would try to move the files to another fast disk that you know is good and try again. You may even need to put this on another machine to do the repair since its possible the issue could also be within the controller but before
we go there tell me about the disk this DB is on, i.e. type speed, history etc and also check your event logs to see if any Disk issues are being reported
Troy Werelius
www.Lucid8.com
November 10th, 2010 9:38am