Video Screencast Help
Symantec Appoints Michael A. Brown CEO. Learn more.

BE 12.5 - Database could not be found

Created: 22 Aug 2012 | 7 comments

Dear All:

I have been experiencing this issue for a week.

I have BE 12.5, patched with the latest service packs. As my autoloader died about a year ago, it was decided to use a NAS (exactly, a Lacie 5Big Network 2, 10TB, RAID5, cheap and sturdy). 

So, my boss and I were happy for about 6 months...but a week ago it started to behave weird. All jobs, except for Exchange 2003 Info Store are not working. It actually does the job, but when it finishes the verify stage it says "Database could not be found..". There are no errors on the Exchange 2003 server (on the event logs). As the time of this post, I had to create a media set locally using NT Backup.

Job is configured as follows:

- GRT is ON

- Selection list is only for IS and System State for that server.

------------------

Some info about my environment:

- Media Server has lots of free space.

- Media server has /3GB /PAE as boot parameters. (just added the 3GB after reading a TECHNOTE, no joy).

- Media Server commit load is 4GB (there are many services in there).

- Agents are updated to the latest version.

- All servers are 2003 Standard x86, 4GB RAM, RAID 5.

- Exchange Physical Size is 76GB (limit is at 75GB...I just noticed it after a reboot). However, logical size is 64GB and has been for about 3 months.

- NAS has 2TB free.

- I can do backup jobs to other servers.

I am kind of desperate because today marks 1 week without email backup. That's why NT Backup is being used right now. I have searched all the forum and tried to apply any eligible workaround with no success.

Any help will be appreciated.

Thanks,

Marco Tafur

 

 

 

Comments 7 CommentsJump to latest comment

CraigV's picture

Hi Marco,

 

Check below:

http://www.symantec.com/business/support/index?page=content&id=TECH53031

Thanks!

Alternative ways to access Backup Exec Technical Support:

https://www-secure.symantec.com/connect/blogs/alte...

Marco Tafur - Qualitas Technology's picture

 

Craig:

 

I checked the page (i forgot to mention I had checked it already).  

 
Proposed Solution Result
1: This issue will occur if the IMG folder used to store GRT related data is being scanned by anti-virus software while the backup is running. To resolve this issue, exclude Backup to disk folder  where GRT backups of exchange is run from Avti- Virus Scan on  local media server. NOTE: IMG folders are located in the root of the target Backup to Disk folder. DONE. I use SEP 11 and have excluded everything. I even disabled all the services to test. No results.
2: The steps described in Microsoft KB article http://support.microsoft.com/kb/304101 for Registry setting 1 will need to be applied to the Backup Exec media server.  A reboot on the media server is required after making these changes. Haven't tried that yet. I have only added the 3GB switch. No result. Will try today afternoon the registry setting (it is the file server).
3. This issue can also occur if the drive which hosts the Backup to Disk folder is fragmented. A disk defragmentation will need to be preformed for the backups to complete successfully.  Cannot defragment as it is a NAS. Will test if it has the same problem if I backup to a DAS.
4. Specifically exclude Backup Exec services files from the antivirus scan. The backup exec services files are located in the location e.g. 'C:\Program files\Symantec\Backup Exec\". The backup exec services are beremote.exe, pvlsvr.exe, beserver.exe, bengine.exe and benetns.exe. Done, no result. Whatsmore, those files were within the exception list when I installed it.
5.) This issue can occur when backing up multiple Exchange databases with GRT enabled at the same time. If possible, seperate the backup jobs so that each database backs up one at a time. I have only one database.
NOTE: If the error in the debug is followed by: "StopVirt FAILED!  RC = 1450 (0x5aa) - Insufficient system resources exist to complete the requested service", the following suggestions should be tried: Will try de Debug today afternoon. Will report.
 For 32-Bit media servers with at least 1GB of RAM, the /3GB switch can be set in the boot.ini file on the Backup Exec media server (see acknowledgements) Done. /PAE was there before, so I left it. I just added the /3GB switch. No result.
 Due to how Backup Exec handles mapping the Exchange database in memory, some databases (typically large databases, or databases with a large amount of logs) require more virtual memory than a 32bit server can provide. These limitations can be avoided when using a x64 bit server, and as such may require Backup Exec to be installed onto a x64bit Windows server.  Can't try. I don't have my new servers yet (they'll be on October I guess...oh shiny new toys...).
I will try a debug tomorrow to see if the errors are related to this Tech note. Will keep you posted.
 
Do you have any other insights?
 
Marco
 
 

 

Marco Tafur - Qualitas Technology's picture

Praveen:

I have tried the posible solutions given in the KB you sent me. 

 

This issue can occur because Exchange Maintenance is scheduled to run at the same the backup is, Exchange is being backed up by another process (i.e.: NTBackup), or there is actually corruption within the Exchange Database. 

We don't have any scheduled task running NT Backup. We are using it now because I don't have BE working properly.
Confirm that Exchange Maintenance is not scheduled to run when a backup does.  For more information on this, review the following Microsoft Article: OK, I Changed it. However, I have tried doing it during the day and no success.
 If NT Backup is scheduled to backup an Exchange Information Store at the same time Backup Exec does, Backup Exec may not successfully backup all the data via the Exchange Agent, and a corrupt file exception will show in the job log.  An entry, similar to the following, will also show in the job log, if the file system on the Exchange Server is selected for backup:

Unable to open the item \\Exchange Server\E:\backups\exchange.bkf - skipped.

A skipped bkf file indicates that NT Backup Exec is running an active backup to that file. To resolve the issue with the Exchange Backup not completing successfully, make sure NT Backup is not backing up Exchange at the same time the Backup Exec Backup Job is running.  

NT Backup is not running on server.
   

 I can confirm I can do Backups using NT Backup on the same server. So, the problem must be the media server. I'll start debugging it right away.

Regards,

 

Marco Tafur.

Marco Tafur - Qualitas Technology's picture

Just to give an update. I am running the job to DAS. It is really slow (60MB/min).

Will keep you posted!

Praveen_Shetty's picture

Can you please check in the failed job log if the backup shows the error "Database could not be found" for every database (in case you have more than 1 mailbox database) or only for some specific databases.

Also make sure beremote.exe file version on exchange server matches the version of  beremote.exe on the Backup sever,If not then push install the remote agent on the exchange server again.

Which Antivirus application is installed on the Exchange server? Does it have mail security?

Thanks

Marco Tafur - Qualitas Technology's picture

Pravin,

 

Thanks for the reply...sorry for the long absence.

I have two databases, my main database (aprox. 75GB) and the public database. Backup is set for main only.

I haven't updated BE in a while. However, I will push install RAWS again. 

Finally, no mail security is installed in the Exchange box. We use SEP 11 RU6.

I will report back when I reinstall RAWS.

Regards,

 

Marco