Category: SCOM 2007

The first public beta release of OpsMgr 2007 R2 is available for testing now.  Features include:

• Manage UNIX and Linux seamlessly.  I’ve seen this in action.  It is not a bolt on.  Operationally, it looks very nice.  They decided to do it only in R2 because it needed some changes in how to use the role user accounts.
• More VMM integration
• Improved web application monitoring
• The SLA stuff appears to be integrated rather than being a bolt on as it currently is
• A faster console
• Better management pack … management (stuck for a better word there!)
• Simplified notification  – it was needed because it’s a maze to figure out for the first time
• Improved and simplified authoring – I really hope so because discovery is a nightmare

The SQL management pack for Operations Manager 2007 now has support for discovering and monitoring SQL 2008.  Congratulations to the SQL team for getting this update out so quickly.  I only wish some of the Windows teams were half as efficient.

Microsoft has released a new Windows Server 2003/R2 Distributed File System management pack for Operations Manager 2007.

I just deployed the HP Blade System Management Pack for Microsoft System Center Operations Manager 2007 (what a mouth full!).  It was relatively easy.  I’m assuming you’ve already added the Proliant Server management pack.

• Get the IP address of each of the blade chassis’ you have.
• Set up a "user" account for OpsMgr and note the password.
• Configure SNMP alerts to your OpsMgr server and set up a read string for the OpsMgr Server.
• Download and install the exe from HP.
• Start the BladeSystem monitor and add your chassis in there using your chassis IP address and OpsMgr user account/password.
• Add the blade management pack from the HP Blade Management Pack folder in Program Files.
• Discover a network device using the SNMP read string and the chassis IP address.
• Deploy agents to your blade servers.

Wait a while, maybe 10-15 minutes.

Come back and you’ll find lots of newly discovered objects and a really cool BladeSystem diagram that drills down to all of the components.

I went about doing this today and to be honest, the instructions are non-existent.  You’ll need a bunch of updates.  Note that the hotfix rollup includes X64 and X86 updates and there are individual downloads for the x86 and x64 versions on the other KB’s.  Here’s how you can get an Operations Manager 2007 agent working on W2008:

• Download and "install" KB954049 (a hotfix rollup).  This "install" is a messy extraction.  Note where it installs and copy the contents to somewhere safe.  You can "uninstall" the hotfix now.
• Upgrade all of your OpsMgr server components to service pack 1.
• Install KB951116, KB952664, KB953290 onto the W2008 server that you want to install an agent on to.
• Reboot the destination agent server.
• Install the OpsMgr Agent.  Make sure it is the SP1 agent!
• Stop the OpsMgr Health Service.
• Browse to where you saved the contents of KB954049.  Install the update.
• Restart the server.

Your agent should now be able to communicate with the management server assuming that all the security side of things is OK (the same rules apply).

You’ll also find that you need to install the Windows Server 2008 Discovery Management Pack in order to be able to manage the OS on these new servers.  This management pack also appears to be a pre-requisite for role/feature management packs for Windows 2008, e.g. you’ll be unable to import those other management packs.  This is probably quite logical, i.e. there is a discovery dependancy for Windows Server 2008 objects.

Note that there’s a 4th fix (KB951327) required for any machine running a OpsMgr console on Windows Server 2008.

This is all a very manual thing.  You can probably script this pretty easily.  If you have System Center Configuration Manager 2007 (SCCM / ConfigMgr) or SMS then you can create a sequence of package programs with prerequisites to take care of this for you.  I’ve an example of this in my SCCM 2007 beta whitepaper on software management.

EDIT:

Remember that Windows Server 2008 has its firewall turned on by default and that it blocks both inbound and outbound traffic.  You’ll need to configure rules to allow your required traffic (e.g. TCP 5723) either manually, by script or by Group Policy.

I had a problem with a faulty health status showing up on my System Center Operations Manager 2007 console.  An agent (the SCOM server itself) was telling me the "Local Application Health Rollup" under "Performance" was unhealthy.  There health in the underneath objects was healthy.  There were also no associated alerts on the server itself and performance was OK.  Therefore there was a glitch in the health status rollup.  I tried to reset and refresh the health but that did nothing for me.

I did two things to resolve this glitch and get everything back to green:

1. I navigated to the Computers view in Monitoring and double-clicked on "Health Service" on the affected agent.  This opened a new window showing the state of the health service on this agent.  On the Actions menu, I selected "Health Service Tasks -> Flush Health Service and Cache".
2. I restarted the OpsMgr Health Service.

A few minutes later I the health status in Health Explorer was reset to green.

BTW, it’s strange that the MS spell checker in Windows Live Writer does not recognise the word "Rollup".  It’s valid in MS Word.

As techies, we tend to look at business applications as components, e.g. disks, CPU’s, servers, services, etc.  However, the owners and consumers of those applications see them very, very differently.  My first experience of this was when an old boss wanted to bring in some of the concepts of ITIL into the organisation.  ITIL is a British government standard for guaranteeing the quality of IT services.  I used the Wikipedia link instead of the official site because the official site does a bad job in describing the standard – a common fault of web sites!  It has gained acceptance from organisations globally and Microsoft has tailored it for MS centric networks in the form of the Microsoft Operations Framework or MOF.

ITIL gets us to view everything as services.  A CRM application is a service.  It consists of web applications running on web servers, databases on servers, networks and storage.  If one component fails or performs poorly then the service fails or performs poorly.  That’s all that the customer or consumer of the service cares about!  They don’t care about a CPU running at 100% because it means nothing to them because they are not techies.  They only know that they are losing productivity or profit because a service is performing poorly.

OpsMgr 2007 was designed to allow the ITIL/MOF view of services to be modelled in the form of distributed applications.  You can either use one of a number of templates or create your own distributed application to drop in monitored components to model your service(s).  This is great for operators or more tech savvy people to monitor.  However, how about the business or service owners?

Microsoft recently released the Service Level Dashboard for SCCM 2007.  This allows you to define Service Level Agreements (SLA’s) for your applications, measure and report on them.  The dashboard allows you to use historical data on your defined distributed applications to see how those applications measure up against your agreed SLA’s.  There’s more information on TechNet about the service.

I manage a growing number of OpsMgr agents on un-trusted networks.  This means I have to assign these agents the CA cert and an OpsMgr cert instead of using Kerberos authentication.  It’s a painful manual process but once it’s done I don’t have to go back to it for several years to replace the certs.

Someone in MS just came up with a wizard to "simplify" the process.  To be honest, reading the page made it sound like things just got more complicated.  Feel free to have a look but I think I’ll stick to the CertSrv web page.

Microsoft has released a document detailing common scenarios for System Center Configuration Manager 2007.  As you’ll soon see, ConfigMgr is very scalable.  They are rating a single site server with dual CPU and 4GB RAM for up to 10,000 manage clients.  That might be a small site by MS/USA standards but that’s a pretty large deployment by mine!

This Solution Accelerator provides a set of System Center Ops Manager 2007 Management Packs for Windows SharePoint Services 3.0 and Microsoft Office SharePoint Server 2007 (including SP1). The Toolkit also includes user documentation for each of the Management Packs.