Archive | SCOM

The new management pack has a bunch of new rules and most importantly doesn’t cause the WMI errors like previous versions.

Download: My Custom Dell MP without SNMP

After importing the new Management Pack I immediately started getting errors on my Root Management Server and Management Server.  Not the good kind of errors you would expect, (like errors telling you there is a hardware issue).  The kind of errors that tell you that your SCOM infrastructure is now hosed!

Here were some the errors I received
*****************************************************************************************************************************
Data Warehouse failed to enumerate database components to be deployed. Failed to enumerate Data Warehouse components for deployment. The operation will be retried.
Exception ‘SqlException’: Timeout expired.  The timeout period elapsed prior to completion of the operation or the server is not responding.
One or more workflows were affected by this.
Workflow name: Microsoft.SystemCenter.DataWarehouse.Deployment.Component
*****************************************************************************************************************************
Data Warehouse managed object type synchronization process failed to write data to the Data Warehouse database. Failed to store data in the Data Warehouse. The operation will be retried.
Exception ‘SqlException’: Timeout expired.  The timeout period elapsed prior to completion of the operation or the server is not responding.
One or more workflows were affected by this.
Workflow name: Microsoft.SystemCenter.DataWarehouse.Synchronization.TypedManagedEntity
*****************************************************************************************************************************
Event ID: 31569

Description:
Report deployment process failed to request management pack list from Data Warehouse. The operation will be retried.
Exception ‘SqlException’: Timeout expired.  The timeout period elapsed prior to completion of the operation or the server is not responding.

Workflow name: Microsoft.SystemCenter.DataWarehouse.Deployment.Report

****************************************************************************************************************************
Health Service Unloaded System Rule(s)
Alert raised by monitor when system rules have been unloaded by the Health Service.
Event ID:      4000
A monitoring host is unresponsive or has crashed.  The status code for the host failure was 2164195371.
*****************************************************************************************************************************

After doing some digging I found there was a fix for the Data Warehouse errors.  Installing KB954643 immediately fixed the Data Warehouse errors.

The only thing left was the Event ID 4000 – A monitoring host is unresponsive or has crashed.  The status code for the host failure was 2164195371.  So how do we fix these errors?  I found KB951526.  Problem was I already installed this fix and was still having the issues.

So now how do we fix the problem?  The old management pack didn’t cause these issues so why is the new one? 
So I decided to export both the old and new management pack using Borris’s powershell script to see what changed is this new version of the Dell MP.

After evaluating the two management packs in the Authoring Console it was apparent that the new management pack contains tons of new SNMP monitors that I don’t need.  Also my experience after talking to people in the community is that the SNMP provider is not very scalable or robust.   So I set all of SNMP related discoveries to disabled and removed all rules related to SNMP.  After doing that and re-importing the management pack the 4000 errors are gone and everything seems to be working much better.  I have posted my customized management pack at the beginning of this post.

Microsoft began a major virtualisation push late yesterday, with the introduction of new virtualisation tools and by making its core hypervisor product free of charge. The new Hyper-V Server 2008 would be available via the web at no charge, Microsoft said in a statement. The download, which will go live in 30 days, supports virtual machines running Windows or Linux. The software comprises Hyper-V itself, plus a cut-down Server Core installation of Microsoft’s Windows Server 2008, and as such is a stand-alone version of the virtualisation solution.

Microsoft had previously said it would charge US$28 for the standalone Hyper-V product. As expected, Microsoft announced its new System Center Virtual Machine Manager (SVCMM) 2008, which will also be released within 30 days.

The software allows companies to centrally manage their virtual infrastructure, whether their virtual networks run on Microsoft’s Windows Server 2008 Hyper-V, Server 2008 Hyper-V or Virtual Server 2005 R2; or on VMware Virtual Infrastructure 3. Also yesterday, Microsoft said it had demonstrated that for the first time it could run live migration in a virtualised environment.

Full story: ZDNet News

PDF Version

Something that has always puzzled me with MOM 2005 and now System Center Operations Manager 2007 is that when I create a dynamic group of computers and one of the computers goes offline I don’t get a heartbeat.  This seems like something that should work “out of the box “as all other monitors are dependent upon the server being up.  Last year I posted a tool called the “watchanator” that addressed this issue.  The “watchanator” worked well but was complicated to setup and needed to be run every time the dynamic group changed.  Well now I have a better solution.  This new solution may seem a little complicated as well first but is really very easy if you follow my simple steps.

1.     Create the dynamic group, using the group wizard. 

a.  Give your dynamic group a name and select the unsealed management pack you want to store the group in.

b.  On the Explicit Members tab click next.

c.  On the Dynamic Members tab click  and create the formula you want to use to create you dynamic group.  I used a simple one that dynamically includes all of my exchange servers.

d.  From the groups window, right click to verify that your dynamic group includes the computers you want.

As you can see the dynamic group only contains windows servers and does not contain the heartbeat object(or agent watcher as it is know in scom)

2.     Export the Management pack

3.   Open up the Management pack in any xml editor.  I am using visual studio

4.  Search for <MembershipRules>.  The membership rules make up the dynamic group.  As you can see the first membership rule between <MembershipRule> and </MembershipRule> contains the formula that creates my dynamic group.

5.  Now we need to add the code to include the health watchers.  Open up watchers.xml available at https://www.scom2k7.com/downloads/Watchers.xml and copy the xml code. 

Here is the code.

<MembershipRule>

 <MonitoringClass>$MPElement[Name=”SystemCenter!Microsoft.SystemCenter.HealthServiceWatcher”]$</MonitoringClass>

<RelationshipClass>$MPElement[Name=”MicrosoftSystemCenterInstanceGroupLibrary6062780!Microsoft.SystemCenter.InstanceGroupContainsEntities”]$</RelationshipClass>

<Expression>

<Contains>              

<MonitoringClass>$MPElement[Name=”SystemCenter!Microsoft.SystemCenter.HealthService”]$</MonitoringClass>

      <Expression>

<Contained>         

<MonitoringClass>$MPElement[Name=”MicrosoftWindowsLibrary6062780!Microsoft.Windows.Computer”]$</MonitoringClass>

          <Expression>

            <Contained>

              <MonitoringClass>$Target/Id$</MonitoringClass>

            </Contained>

          </Expression>

        </Contained>

      </Expression>

    </Contains>

  </Expression>

</MembershipRule>

6.     Paste this code after the first </MembershipRule> and before </MembershipRules> then save the file.

*Note if you choose another class other than Microsoft.Windows.Computer to create you dynamic group you will have to replace the line from watchers.xml with the matching line in your dynamic group membership rule.

<MonitoringClass>$MPElement[Name=”MicrosoftWindowsLibrary6062780!Microsoft.Windows.Computer”]$</MonitoringClass>

7.   Save the MP and re-import the Management Pack into SCOM

8.   Go back to your group and right click View group members

9.  The health watchers may up to 20 seconds to populate.  Push F5 a couple of times to refresh the screen.

Now when we create a subscription to this group and a server in the group goes offline we will now get the heartbeat alert.

You download the sample management pack at https://www.scom2k7.com/downloads/Exchange.Servers.xml

Feature Summary

• DHCP 2003/2008
  • Native Systems Center Operations Manager 2007 MP
  • Windows Server 2008 Support
  • Scope & Superscope Address Availability Health
  • Updated Health Logic
  • New Reports
  • Updated Knowledge Base Information
  • Coexists with the included DHCP 2000 MP
• DHCP 2000
  • Converted Systems Center Operations Manager 2007 MP
  • Coexists with new native DHCP 2003 / 2008 MP

Note: This package includes library MP, 2000 and 2003 MPs, and DHCP 2008 MP. DHCP 2008 MP can only be installed on System Center OpsMgr 2007 SP1.

If you have ever worked with SNMP devices inside SCOM you probably have noticed that when you discover a SNMP network device and it goes down no alert is created.  The SNMP device will show up in the monitoring console as down but no alert goes out.  The reason for this is that when the SCOM team created the “Device Status Check monitor” they either forgot or intentionally left the alert disabled.

So how do we enable the alert?  You can easily override the alert, but the problem is the alert you get will only include the IP address and not the DNS Name.  In a small environment this is fine but in a larger environment the DNS names are a must.  There is no way to override the description so working with Marius Sutara from Microsoft he was able to create a unsealed management pack that disables the built in monitor and then duplicates the exact functionality of the “Device Status Check monitor” .

null

You can download the unsealed managment pack here.

Microsoft.SystemCenter.NetworkDevice.Custom.xml

Also worth checking out is the entire post on how he created the management pack.

http://blogs.msdn.com/mariussutara/archive/2008/07/18/customizing-monitor-from-sealed-mp-device-status-check.aspx

Big thanks to Marius for this.

Jonathan Hambrook complied this excellent guide a couple of months ago but it looks like the links are broken and he no longer maintains his site so I am going to host the guide here.

PDF File
 

Also this post is worth checking out if you want to monitor SNMP devices

https://www.scom2k7.com/how-to-create-a-basic-probe-based-snmp-esx-console-cpu-monitor-in-operations-manager-2007/

Dell has released an updated version of there Management Pack for System Center Operations Manager 2007 and System Center Essentials 2007 Version 3.1, A01.

Fixes and Enhancements

Version A01 changes:
– Dell Server and Printer Management Pack (Dell.Connections.001.mp) code was modified with fixes for supporting upgrade from Dell Management Pack v3.0 to Dell Management Pack v3.1 release.

Supported on System Center Operations Manager 2007 SP1 and System Center Essentials 2007 SP1 only. It does not function with prior releases of the MOM product.

Download

http://support.us.dell.com/support/downloads/download.aspx?c=us&l=en&s=gen&releaseid=R194919&formatcnt=1&libid=0&fileid=269302

Overview

Windows Deployment Services Management Pack for Operations Manager 2007

Microsoft released SQL Server 2008 on Wednesday. The new database software is a continuation of Microsoft’s efforts to bolster SQL Server’s credibility as a highly scalable database and the cornerstone to Microsoft’s evolving “data platform.”

“If you added up all of our investments in SQL Server 2008, it was data warehouse scale that got the highest level of investment,” Ted Kummert, the VP of Microsoft’s data platform and storage division, said on a conference call with reporters and analysts. Microsoft already claims numerous deployments of multiterabyte, heavily accessed databases and has largely gotten over its reputation as a system that couldn’t scale, but the company continued that focus in this release.

http://www.informationweek.com/news/software/database/showArticle.jhtml?articleID=209903873

It’s available on MSDN

While installing System Center Operations Manager 2007 on a new Server 2008 box with all the patches my health service was failing.  After researching I noticed that that problem was because the OpsMgr Health Service and the OpsMgr SDK Service was trying to start before the Netlogon service was running.  This is due to spanning tree running my network switches.  Here were the errors I was getting.

Event ID:  20056
Description: Failed to initialize the IPC channels for other processes to communicate. The HealthService might fail to start.

Event ID:  2000

Description: The Management Group MYGROUP failed to start.  The error message is Unspecified error
(0x80004005).  A previous message with more detail may have been logged.
Event ID:  2001

Description: No Management Group could be started.  See previous errors for more specific details.  If no errors are present, verify the machine has sufficient memory and has at least one Management Group configured.  Because Active Directory Integration is disabled, the Health Service will not remain running.

To quickly fix the problem in Server 2008 I set my services to delayed start.  (The Service Control Manager starts services that are configured for delayed automatic start after all of the automatic-start threads have finished starting.)  http://blogs.technet.com/askperf/archive/2008/02/02/ws2008-startup-processes-and-delayed-automatic-start.aspx

To fix the problem in Server 2003 set the OpsMgr Health Service and the OpsMgr SDK Service to be dependent on the Netlogon Service. Here are in instructions http://support.microsoft.com/kb/193888