SQL Agent Hangs With Wait Type PREEMPTIVE_OS_GETPROCADDRESS

MattS

Hi,

We are using SQL Backup 6.5.1.9, SQL Server 2008 R2.

We have log shipping running between a production and DR server (it has been running for several months), and twice in the past week the restore job has hung with a wait type of PREEMPTIVE_OS_GETPROCADDRESS.

The following error messages are recorded in the event logs:

SQLVDI: Loc=SVDS::Open. Desc=Bad State. ErrorCode=(-1). Process=1492. Thread=3140. Server. Instance=MSSQLSERVER. VD=Global\SQLBACKUP_B3CE6C55-9A7C-4B6B-A435-B048FBF7960A_SQLVDIMemoryName_0.

Cannot open backup device 'SQLBACKUP_B3CE6C55-9A7C-4B6B-A435-B048FBF7960A'. Operating system error 0x80770006(failed to retrieve text for this error. Reason: 15105).

Are you able to shed some light on what these error messages mean and what may be the cause? Because the SQL Agent job is hanging, rather than failing, there are no error notifications, and we are therefore in a situation where our DR site is out of sync and we are unaware of it!

Thanks in advance for your help.

Chris.Allen

This appears similar to an issue that some other users experienced. In those cases, we provided a patch, which resolved the issue.

The patch is available using this link:
ftp://support.red-gate.com/patches/sql_ ... 5_2_18.zip

The zip contains two files:

SQBCoreService.exe
SQLBackupC.exe

To apply the patch please follow the steps below:

1. Close all GUI connections to the SQL Server.
2. Ensure that no SQL Backup jobs are in progress.
3. Stop the SQL Backup Agent Service, or disable the cluster resource if on a cluster.
4. On the SQL Server where the SQL Backup server components are installed, navigate to the installation directory. Defaults are:

32-bit Machines - C:\Program Files\Red Gate\SQL Backup 6\(local) or <SQL Instance Name>
64-bit Machines - C:\Program Files (x86)\Red Gate\SQL Backup 6\(local) or <SQL Instance Name>

5. In the folder you will find an existing SQBCoreService.exe and SQL BackupC.exe. Rename the existing files, for example to OLDSQBCoreService.exe and OLDSQLBackupC.exe.
6. Now copy in the files extracted from the downloaded zip.
7. Restart the SQL Backup agent service/cluster resource, and ensure that it starts successfully.
8. Perform a test backup to ensure that all is OK. Then test to see if the original error occurs. SQL Backup should now error out correctly and not become unresponsive.

Let me know if the error still occurs after applying the patch.

MattS

Hi Chris,

Thanks for the response. I'll deploy the patch and let you know.

ChrisAVWood

Chris,

Are all the fixes in 6.5.2.18 included in the new V 7?

Thanks

Chris

MattS

We deployed the patch on Tuesday and our log shipping backup/restore processes have been running ok.

Last night there was a failure in the restore job, but this was handled and subsequent restore jobs have completed successfully. The same SQLVDI error messages are recorded in the windows event logs, so it seems likely this is the same error that was causing the jobs to hang previously.

Deploying the patch therefore seems to have resolved the issue with the jobs hanging.

For reference, the errors recorded in the SQL backup log file are:

10/05/2012 03:05:08: Thread 0 error: 
Process terminated unexpectedly. Error code: -2139684861 (The api was waiting and the timeout interval had elapsed.)
10/05/2012 03:05:08: 
10/05/2012 03:05:08: SQL error 3013: SQL error 3013: RESTORE LOG is terminating abnormally.
10/05/2012 03:05:08: SQL error 3201: SQL error 3201: Cannot open backup device 'SQLBACKUP_2022D7BB-9FE4-4987-AF55-9B1A8AFA6512'. Operating system error 0x80770006(failed to retrieve text for this error. Reason: 15100).

Should the errors persist I will raise a new forum topic.

Thanks for your help.

Chris.Allen

Are all the fixes in 6.5.2.18 included in the new V 7?

Yes. As a general rule, Red Gate has a 'linear' approach to developing software- we hardly ever create branches.

Hope this helps