Applying Exchange 2007 SP1 Update Rollup 6

Perhaps I should have picked another day for applying Exchange 2007 SP1 Update Rollup 6. I ran into the services not starting issue documented in KB 944752 Exchange Server 2007 managed code services do not start after you install an update rollup for Exchange Server 2007.

It was a single server topology - Exchange 2007 SP1 sitting behind ISA Server 2006 (SP2), with no previous post-SP1 rollups installed. The System Attendant, Information Store, and Active Directory Topology services started without any issues. The other Exchange services did not. Manually starting the stopped services resulted in the familiar 1053 error - service failed to start in a timely fashion.


Figure 1: Manually starting Exchange managed services resulted in error 1053

A helpful colleague informed me about the nature of these managed services, and pointed me to the above KBA and the related downloads. Installing .Net Framework 2.0 SP2 fixed it, although according to the KBA .Net Framework 2.0 SP1 would work just as well.

The managed code services not starting issue is mentioned clearly in KBA 959241 Description of Update Rollup 6 for Microsoft Exchange Server 2007 Service Pack 1:

Note Certain Exchange Server 2007 managed code services may not start after you install this update rollup. This problem occurs if all the following conditions are true:

• The services cannot access the following Microsoft Web site:
  http://crl.microsoft.com/pki/crl/products/CodeSigPCA.crl
  
• You do not have a .NET Framework common language runtime (CLR) build that supports the generatePublisherEvidence configuration setting.

For more information about how to resolve or work around this issue, click the following article number to view the article in the Microsoft Knowledge Base:
944752 Exchange 2007 managed code services do not start after you install an update rollup for Exchange 2007

Rudimentary best practices
When working in mid to large deployments, we normally tend to follow change control procedures in place. Service packs, hotfixes, and updates are tested adequately. Change control also translates into being able to articulate a suitable fallback plan. Backups need to be performed before applying any updates. These practices, even in a scaled down version, are routinely ignored or avoided in smaller environments.

Installing updates without testing and without reading the accompanying documentation could turn into your little adventure on days you can ill-afford to indulge in such adventures. The consequences could range from hours wasted (read "unbillable hours" if you're a consultant) troubleshooting or restoring service, to potentially more serious consequences caused by the downtime.

The good news is this can be avoided easily!

Have you been in a situation before where not following these rudimentary best practices has resulted in undesirable consequences?