v8 Update 3 & SQL Log Truncation

[millardjk]

Ref: Case 01080444

VMs with SQL Server 2014 were backing up w/o issue prior to applying U3 now throw "unable to truncate" warnings.

Have seen the same change in behavior in a completely different environment; in that one, there are SQL 2008R2, SQL 2012 and SQL 2014. SQL 2014 is common element; no change in behavior with SQL 2012 or SQL 2008R2 VMs.

[cdlane]

Us too, no issues prior to upgrading today.

Failed to finalize guest processing. Details: Failed to process 'TruncateSQLLog' command. Failed to truncate transaction logs for SQL instances: MICROSOFT##WID. Possible reasons: lack of permissions, or transaction log corruption.

All server 2012R2.

[Gostev]

Also SQL Server 2014?

[cdlane]

SQL 2012 on both.

[Gostev]

OP states he has no issues with SQL 2012 though, everything is working fine with those after update.
Could be a different or unrelated issue in your case. By the way, please also post your support case ID here.

[cdlane]

Hi Anton,

Well I guess your going to be removing my posts as I don't have a support ID. After paying all the money for the Veeam product, I'm probably going to lose my job when my bosses find out it's an unsupported product without paying more money which they will refuse point blank to do. It was hard enough to get them to part with the original cost in the first place...ouch never thought of checking whether it was supported or not .

I shall continue to lurk and pray to God nothing really bad happens and see if any tips/hints are posted by anybody else.

Best regards,

Chris

[btmaus]

Which version of SQL is installed on your VBR server?

[cdlane]

SQL 2012 on VBR also.

[cdlane]

Hi Anton,

I believe I owe Veeam and my reseller an apology. I was unaware that the paid support was included as I didn't set up the Veeam install so was unaware of the existence of the Support ID etc.

Case ID = 01080564 and include on http://forums.veeam.com/veeam-backup-re ... 30-75.html

[millardjk]

My interaction with support shows that a Windows Internal Database is causing the truncation warnings to be thrown. And while the support person says that the logs show the WID being added in the last couple of days (ie, coincidental to updating VBR), I can see in the host that the WID instance has been there--and undetected--since late 2014. New code in U3 is now detecting the instance, and its inability to perform the truncation is creating warnings where it blithely ignored the entire instance in the past.
Still no good reason why the other 2014 instance is throwing the same warning, but it's also an unsupported lab box, so I'm pretty much on my own to figure it out as it's not got a WID causing the issue.

[btmaus]

There was a recent post (before U3 was released), saying that the SQL version on the VBR server needs to be >= whatever version(s) of SQL that you are backing up ... so is that the issue?

[Shestakov]

Did you install any private fixes before the update 3 by any chance?
Thanks!

[foggy]

Chris & others, the "unable to truncate" warnings do not actually mean that something is unsupported, they just tell that logs for the particular internal database instance were not truncated. Everything else is working fine and logs are truncated where they should be. It is just some changes in instance detection logic introduced by Update 3 (to fix other issues around this functionality) that result in improper processing of internal databases (where there are no logs to truncate, anyway). R&D is currently looking into this and I believe the solution will be found soon.

[foggy]

Hotfix for this issue is already available through support. Please contact them directly if you observe similar behavior after upgrading to Veeam B&R v8 Update 3.

[Gostev]

Can you all please confirm that impacted VMs are running WSUS, or are there some other applications which are impacted as well?

[RGlintmeijer]

I have a customer who has WSUS on a 2012R2 VM and encounters this issue (after installing U3) on "MICROSOFT##WID". There's also an SQL Server 2014 Express instance installed on the same VM.

The customer was told the current hotfix wouldn't fix the warning on "WID". Case # 01080855

[foggy]

Robert, please ask the customer to escalate the case (you may refer to this thread) to get the new hotfix for Update 3.

[millardjk]

I figured out what was going on with the Lab server, too: while the final error message in the console is related to permissions, the root cause seems to be the lack of a full database backup:
Scenario: new VM; new Database
Having never backed-up the Database in Full Recovery mode on SQL itself, the Truncate Logs command would fail with "no full backup exists." Then VBR would attempt to issue the same Truncate Logs command using the NETWORK SERVICE account, which then failed due to lack of permissions.
After manually running a backup on the SQL Server itself, the Truncate Logs command succeeded when issued manually; I'll see later tonight if it succeeds within the context of the Backup Job.

This does lead to a new question, however: is that expected behavior? Isn't the VSS backup happening to the image itself supposed to indicate to SQL Server that a Full backup of the target database exists, thus permitting log backups to be performed?

[cmillian]

Hi all,

Since update 3 my SQL backups ends with a warning.

Failed to finalize guest processing. Details: Failed to process 'TruncateSQLLog' command.
Failed to truncate transaction logs for SQL instances: MSSQLSERVER. Possible reasons: lack of permissions, or transaction log corruption.

When I check the log file “VeeamGuestHelper_Date” I see the following.

Code: Select all

10/14/2015 3:08:43 PM  10436 Instance: MSSQLSERVER
10/14/2015 3:08:43 PM  10436           Database: master skipped.
10/14/2015 3:08:43 PM  10436           Database: tempdb skipped.
10/14/2015 3:08:43 PM  10436           Database: model failed to truncate. Error: Code = 0x80040e14
10/14/2015 3:08:43 PM  10436           Code meaning = IDispatch error #3092
10/14/2015 3:08:43 PM  10436           Source = 
10/14/2015 3:08:43 PM  10436           Description = 
10/14/2015 3:08:43 PM  10436           Database: msdb skipped.

The service account that managed the application aware has the rights to backup those databases.

Any ideas?

[Shestakov]

Hi Chris and welcome to the forums.
Please read the thread where your post has been merged and contact Veeam technical support to get the issue fix.
Thanks!

[cmillian]

Hi all,

As I read the post I deed a manual backup with the SQL management studio and then I restarted the job in VEEAM and it is now successful.
I’m going to follow this for a few days now

Regards

[theviking28]

I updated a couple days ago to the "update 3" version of Veeam B&R 8.0

Since then, I've had one VM throw a warning on each backup. The VM runs Microsoft WSUS as well as Sophos Safeguard encryption server.

It is running Windows Server 2012 R2 as its OS.

The warning is "unable to truncate sql server transaction logs. Details: Failed to process TruncateSQLLog" command. Failed to truncate transaction logs for SQL instances. MICROSOFT##WID. Possible reasons are lack of permissions or transaction log corruption.

Now, I do see in the release notes of update 3 the following...

Log truncation is automatically skipped for stopped and unsupported SQL Server instances to prevent errors from being logged.

So, what should I do in order to fix the problem? It worked fine with success for this VM with SQL truncation enabled. I've not changed the jobs at all.

[jed-hyper]

I have installed "update 3" version of Veeam B&R 8.0
Since then getting this error.

Failed to finalize guest processing. Details: Failed to process 'TruncateSQLLog' command.
Failed to truncate transaction logs for SQL instances: MICROSOFT##WID. Possible reasons: lack of permissions, or transaction log corruption.

I do not have WSUS installed on this server.

I have booked a case, CASE#01083675

[RJI]

Windows System Resource Manager on Windows 2008R2 installed the Windows Internal Database for us. After update v8 update 3 we started to receive this warning.

[veremin]

Hi, Richard, have you already reached the support team to get the hot fix?

[enoch]

I talked to Veeam support about this last week.. Currently no fix for this other then wait for 9.0 (should be release end this month)

[Shestakov]

Peter,

As far as I know, there is a hotfix.
Could you post your support case number?

Thanks!

[enoch]

That would be great, the closed case number is: 01079837

[Brad.Morris]

So should I install this update or hold out for a 3a update?

[enoch]

Yes please advise if we should wait for minor update / v9 or contact support?