When Windows Server Backup is used to back up an Exchange Server 2010 Mailbox server that a a member of a Database Availability Group the backup result may report “Completed with warnings”.
In the details of the backup report an error is displayed that the application will not be available for recovery.
In addition the Application event log on the server will report Event ID 2137.
Log Name: Application
Source: MSExchangeRepl
Date: 24/11/2010 12:55:07 PM
Event ID: 2137
Task Category: Service
Level: Warning
Keywords: Classic
User: N/A
Computer: ex1.exchangeserverpro.local
Description:
RPC request to the Microsoft Exchange Information Store service for log truncation failed for database Mailbox Database 02EX1. Error: 4294965485
In the example above the server EX1 was not able to perform log truncation for Mailbox Database 02. Notice that Mailbox Database 02 is currently mounted on a different DAG member, not on EX1. In other words, the database that is mentioned in the event log error is a passive database copy on server EX1 at this time.
[PS] C:\>Get-MailboxDatabaseCopyStatus Name Status ---- ------ Mailbox Database 01EX1 Mounted Mailbox Database 02EX1 Healthy Archive MailboxesEX1 Mounted
This error occurs if an application volume being backed up includes passive database copies. There are two solutions for this issue:
- Only select volumes with active mailbox database copies for inclusion in the backup job. This is not ideal for all environments depending on their database storage layout and whether a mailbox database has been activated on a different server due to a fault or maintenance.
- Add a registry key to disable the Exchange Replica Writer and instead use the Exchange Store Writer for Windows Server Backup
To make this registry change open Regedit and navigate to HKEY_LOCAL_MACHINESOFTWAREMicrosoftExchangeServerV14ReplayParameters. Add a DWORD entry of “EnableVSSWriter” with a value of 0 (zero).
Next, restart the Microsoft Exchange Replication service.
[PS] C:\>Restart-Service MSExchangeRepl
Re-run the backup job and the result should be successful without errors this time.
Hi I have similar issue in my Exchange 2007. can i do the same?
Hi Paul and guys,
my DAG include 2-nodes MBX-01 and MBX-02 with totals 10 DBs. DB1-DB5 is active on MBX-01 and DB6-10 is active on MBX-02. I’ve chosen your second solution that: Add a registry on both MBX-01; MBX-02. On MBX-01, i created a schedule task to run Windows Server Backup Script to VSS Full Backup all the Database Volume (include Active and also Passive Database Copy). Note that i run the Backup Script on MBX-01. All Databases are backed up successfully but only DBs MOUNTING on MBX-01 that is truncated LOG (which means DB1-DB5 here).
So what is the problem here or is it the drawback when we use Exchange Store Writer for Windows Server Backup.
Thank you for your support Paul !
Do you see that event ID in your event log?
Hi Paul,
The Backup Process is completed successfully without any Event ID 2137. But the transaction logs are still not truncated if the MBX Server on which we run the wbadmin script host the Healthy Database Copy (Although i edited the registry as your guide).
So what is the problem here ? The case is:
– If we edit the registry: We can backup successfully without error but the LOG is still not truncated !!
– If we doesn’t edit the registry: We can’t backup successfully with event ID 2137 !
We have Exchange Server 2010 But we create a backup so they gave me this error. but you guide add in regedit so i try now i facing still this error. please help me.
I am waiting your reply or comment.
Dear Paul,
Thanks again for this nice explanation. I have these errors as well, event id 2137:
RPC request to the Microsoft Exchange Information Store service for log truncation failed for database YEF-NL1-2010AM1S162. Error: 4294965485.
They are filling up the event viewer of my mbx servers but that is the whole day and not (only) during the backup. We use Commvault for backup and when I look at one of the databases that is backed up, the logfiles are cleaned.
What can this be? Do I need to apply the fix as well?
Thnx!
i have two Mb servers in a DAG , and four Databases distributed equally on each MB servers.i did mount and activate all databases on one of the MB server ( all mailboxes are active on this server). i did a full vss windows backup on all four databases ,however , there was no log truncated from the server not even GB space cleared . i am having this for all the Databases. replication is healthy ? what else should i have to try
Which backup product?
Did the lastfullbackup timestamp update on the databases?
Were there any event log errors or warnings?
1-i use Windows server 2008R2 (windows backup) for doing full backup
2-yes everytime i took fullbackup , timestamp updates on the database
3-there is no error but information event id 9827 ,event id 2006
i have been three weeks doing research on this issue , Logs are filling up our drives , i cound’t find solution on this !!!!!!
thanks
Hi
I am having the same issue but I am using third party backup solution EMC Networker
can I use the second solution (Registry Modification)
You’ll need to check that with EMC Networker’s documentation.
Pingback: Backing up and Restoring databases in Exchange 2010 | Sail Away Systems Ltd
hi,
now in my Exchange Server is a similar situation, but my mailbox version of Exchange 2007, using cluster continuous replication CCR mailbox database backup using Windows Server 2003 Ntbackup, now I don’t know what to do
Hey Paul. I bought your Migration guide for 2010 , worked like a charm. I recently had to remount an Active DB with Circular logging because log files filled during mbx moves. I also removed both passive copies from the dag and erased their data in logs/db directories. Remounted the Active with Circular logging off and re-added the passives. After the seeds were complete and healthy , the passives are giving me these 2137 errors although i haven’t run any backup yet so I’m not sure what is trying to truncate the logs.
I’m trying manual reseeds of both copies to see if the error goes away.
Re-seeding didn’t solve it but a full VSS backup of the passive DB did.
Thank you for the excellent post. It worked on giving me a “Successful” backup, however it did not truncate the logs. I can’t find any errors in my Event Log.
What are the repercussions if I just delete them? They’re up to 9Gb and growing!
No you can’t just delete them. Make sure you’re selecting the entire volume(s) to back up, and setting the VSS setting to do a Full backup.
https://www.practical365.com/exchange-2010-mailbox-database-backup-restore-windows-server-backup
Yessir, I read that article first. My backup is set to “VSS Full Backup” and it is backing up the whole volume. But it is not truncating the log.
Here’s a screenshot of the successful backup:
http://prntscr.com/yhhtz
I checked the time period in the Event Log and there are not errors or warnings at that time.
I am baffled (and running out of space fast).
If this server is a member of a DAG make sure you’re (a) backing up each database on the active server, and (b) that replication is healthy for the database.
Can we do this on the symantec backup 2010 as well.
Hi, I have Exch 2010 3 DAGs with 3 databases. i did registry fix and VSS full backup but still i got the same backup error with event id 9782.
Q1 – do i have to apply this registry fix on all DAG servers or only on one of the DAG where backup is going to run?
Q2 – Though i will not be able to restore this backup but since i have whole volume backup and i already have 3 copies of each database, can i enable circular logging on all 3 databases to free up some 550 GB of logs which has been accumalated since some time. i dont have more free space and i need immediately some free space.
i need to know your opinion! by the way it took almost 48 hours to do 745 GB of volume backup.
thanks in advance.
You’re getting a different event ID than the one described in the article. I’d suggest looking closer at the event log errors on your server to see if they provide detail for why the backup is failing.
i have got event id 8230 and 565 during backup time. Also there was an event id: 489 & 518 for one corrupt/missing log file.
I checked event 8230 and registry has domain admin account configured and it has access to everything.
thanks,
If you’d like to post the full event log details in the forums https://www.practical365.com/forums that may make it easier.
Pingback: E2K10 – Notes-25052012 | Jonson Yang
Hi Paul,
I’m experiencing the same issue here and I’ve tried the fix you mentioned and then restarting the Exchange Replication service but I’ve not been successful, one thing to note, EnableVSSWriter was already set to 0 in the registry, if that means anything, so I guess I didn’t really do much other than restart that service.
I’m performing a VSS Full Backup on selected folder objects where Exchange is housed. If I choose to backup the entire volume which houses Exchange I get a different WSB error relating to not being able to back up a 2TB volume.
Do you perhaps have any hints on what else I could try out to solve the issue?
Thanks
-j-
Everything I read says the 2TB limit is something we have to live with (for now I guess).
And unfortunately you can’t do an Exchange backup with Windows Server Backup by choosing specific folders, it has to be an entire volume backup.
My suggestion would be to rearrange your Exchange storage layout to use volumes smaller than 2TB, or switch backup products. Sorry 🙂
Hi,
I am unable to take backup mailbox server 2010 .Mailbox is in dag .I get the error message “Consistency check failed and unable to take backup system state also
This worked for me. Thanks alot.
Pingback: Exchange 2010: Are Passive Database Copies Included in DAG Backups?
its the same as you described here. I am thinking if i should go for resetting the log sequence and then try backing up again. I will keep you posted bob.
I am unable to get this working.
After the registry tweak and starting the Microsoft Exchange Server Extension for Windows Server Backup and still my logs dont get purged. I did the same on my passive node also and tried to lock and load a full backup and no joy
event id’s 565 and 9782
I will try an find out a why and comment inere. between if anybody knows how to resolve this would appreciate
Hi Gaurav, sounds like you may be having a different issue than what is described in this article.
If you post your problem in the forums and provide some more details I’m sure we could help.
Hi i’m having the same issue on our Exchange Server but we don’t have more then one Exchange in our Enviroment.
We have only one Exchange Server and when i try to run a backup (with VSS Full Option) the logs are not being truncate.
This solution is the same with one or more then one Exchange Server?
Thanks.
Hi Burak, what errors or event log entries are you seeing for the failed backups?
hi,
when i try to start an exchange backup (i’m selecting only the exchange partition for it) after the consistency check an error message appears and says you can not recover exchange from this backup..so i read a lots of things about this and most of them for written for different scenerios..
so i need an advice about what should i do for succesfull backup up of the exchange server.
thanks.
Hi Paul,
Thank you for your articles. I like to read them
I have a backup question.
When I backup Exchange 2007 SP3 (Update Rollup 5 For Exchange Server 2007 Service pack 3), I get the same warning/problem you described above for Exchange server 2010, I tried the Registery fix,, but still I get the warning. and event ID 565; Event ID 9782; and Event ID 2007.. I dont use LCR and there is no passive copy
Is this still a bug in Exchange 2007 SP 3. What is the solution..
Regards…..
Henk from Holland
Hi Henk, sounds like you have a completely different problem. The exact details of your event log entries would be good to know.
Hi,
We have a DAG where SERVER1 is having EXCHDB1 and SERVER2 is having EXCHDB2.
EXCHDB1 on SERVER1 is the active MOUNTED db and EXCHDB2 on SERVER2 is passive HEALTHY db.
Each time when I run the Windows Server Backup (VSS full Backup) on SERVER1, I get the error message “Consistency check failed”, database backup completed with warnings and It does not truncate the logs.
I also get the error message of one of the log file being corrupt.
Please help
Hi Shabbir, if you’re commenting on this article does that mean you’ve already tried the fix in the article?
If you still need help you will need to post the full error messages you’re seeing.
I tried doing the second option and besides it taking 3x longer to run a full backup, the log files still did not truncate.
Did you make sure you’re doing a VSS Full backup and not just a Copy backup? The default Exchange database backup type in Windows Server Backup is Copy, which doesn’t truncate logs.
Pingback: Exchange 2010 Mailbox Database Backup and Restore with Windows Server Backup - Exchange Server Pro
Thanks I need this solutions.
I have one question.
If I use The second solution that uses registry value, did the passive database backed up complete??
And how can I recover the passive database?