Thursday, July 25, 2013

SCVMM UR3 causes VMM Service to crash on Placement

I am sharing this because it burned me, and why should it burn anyone else.

This should have the sub-title of: “you need to read the release KB articles in deep detail” or “just don’t install Update Rollup 1 in the first place and you will be happier”.

I recently jumped straight into Update Rollup 3 for SCVMM 2012 SP1.  Because I had reported some bugs and the fixes were in there.

So, I downloaded the update rollup and I applied it to my VMM Console and to my VMM Server.

I then pushed out new agents to my remote Library Server and my Hyper-V Servers.  All was good.  Or so I thought.

I then deployed a Service Template.

During the time when SCVMM was sorting out where to place the VMs in my Service the VMM Service crashed.  Okay, so I tried again.  (not to expect a different result but to pay attention to figure out what was going wrong).

In the event log of the SCVMM Server I found the following message:

Log Name:      Application

Source:        Windows Error Reporting

Date:          7/24/2013 12:15:47 PM

Event ID:      1001

Task Category: None

Level:         Information

Keywords:      Classic

User:          N/A

Computer:      beSCVMM.brianeh.local

Description:

Fault bucket , type 0

Event Name: VMM20

Response: Not available

Cab Id: 0

Problem signature:

P1: vmmservice

P2: 3.1.6018.0

P3: Engine.Placement

P4: 3.1.6027.0

P5: M.V.E.P.C.VMDCConversionHelper.GetVMDCPrecheckResources

P6: System.MissingMethodException

P7: bf35

P8:

P9:

P10:

These files may be available here:

Analysis symbol:

Rechecking for solution: 0

Report Id: 711cf4e8-f495-11e2-9404-00155d289b00

Report Status: 262144

Hashed bucket:

Event Xml:

<Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">

  <System>

    <Provider Name="Windows Error Reporting" />

    <EventID Qualifiers="0">1001</EventID>

    <Level>4</Level>

    <Task>0</Task>

    <Keywords>0x80000000000000</Keywords>

    <TimeCreated SystemTime="2013-07-24T19:15:47.000000000Z" />

    <EventRecordID>5184</EventRecordID>

    <Channel>Application</Channel>

    <Computer>beSCVMM.brianeh.local</Computer>

    <Security />

  </System>

  <EventData>

    <Data>

    </Data>

    <Data>0</Data>

    <Data>VMM20</Data>

    <Data>Not available</Data>

    <Data>0</Data>

    <Data>vmmservice</Data>

    <Data>3.1.6018.0</Data>

    <Data>Engine.Placement</Data>

    <Data>3.1.6027.0</Data>

    <Data>M.V.E.P.C.VMDCConversionHelper.GetVMDCPrecheckResources</Data>

    <Data>System.MissingMethodException</Data>

    <Data>bf35</Data>

  </EventData>

</Event>

So, I removed UR3 (the VMM Server KB…510).  I removed the hosts and remote library server from vmm management and added them back.  I fixed up all of my templates (Export them and Import them do fix them quickly).

I opened a case.

In the end, what was the root problem?   I didn’t uninstall Update Rollup 1.

Seriously?

Come to find out, buried in the details of this KB: http://support.microsoft.com/kb/2802159 is the installation information that Update Rollup 1 for SCVMM should be uninstalled manually prior to installing Update Rollup 2 (or any following Update Rollup).

Personally, I never read that particular article prior to installing UR2.  So I never knew to install UR1.  And, it is highly unusual to have to manually uninstall an Update Rollup prior to adding a second.

So, I then uninstalled UR1.  I re-applied UR2.  Then I installed UR3.  And all was happy.

Now, one easy way to check for this: 

If you select About from the ribbon of the SCVMM console you will see a version.  If you have UR3 installed it should be 3.1.6027.0 

If you have UR2 installed it should be 3.1.6020.0 

If you never manually uninstalled UR1 it will remain at 3.1.6018.0 after applying UR2 or UR3.

No comments: