Archive

Archive for March, 2009

Ops Manager: Office Communication Server R2 MP Releases

March 26, 2009 Leave a comment
Have you been waiting for this management pack?  Wait no more it is here, but for Ops Manager 07 SP1 only, not MOM 2005 (and may not be released for MOM 2005 due to the 64 bit requirement only):
 
 

Overview

The Office Communications Server 2007 R2 Management Pack monitors Standard and Enterprise Edition of Office Communications Server 2007 R2. This release also incorporates the Quality of Experience (QoE) MP which was previously a separate Management Pack. Monitored types are event log entries, performance counters, as well as stateful monitoring of QoE. Note that this version of the Management Pack only monitors Office Communications Server 2007 R2, and cannot be used to monitor Office Communications Server 2007.

Office Communications Server 2007 R2 monitored objects are:

  • Enterprise Edition
  • Standard Edition
  • A/V Authentication service
  • Application Sharing Server
  • A/V Conferencing Server
  • Conference Announcement Service
  • Conference Attendant Service
  • Web Components
  • Archiving Server
  • Communicator Web Access
  • Mediation Server
  • Monitoring Server
  • CDR Server
  • SIP Proxy Server
  • A/V Conferencing Edge service
  • Access Edge service
  • Web Conferencing Edge service
  • Monitoring Server
  • QoE of A/V Conferencing Server
  • QoE of Mediation Server
  • QoE of network locations

Feature Bullet Summary:

  • Alerts indicating service impacting events
  • Alerts indicating configuration, and other non-service impacting issues
  • State monitoring of Office Communications Server services
  • Product knowledge: Cause and resolution of identified problems

Release History

  • 3/26/2009 – Version 6.0.6907.21, Original release of the English version.

System Requirements

  • Supported Operating Systems: Windows Server 2003 Service Pack 2 x64 Edition; Windows Server 2008
    Note: Windows Server 2008 64-bit OS is required for this MP.

  • Other Software:
    • Office Communications Server 2007 R2 Standard or Enterprise Edition
    • System Center Operations Manager 2007 Service Pack 1

    Note: This Management Pack does NOT support the RTM version of Operations Manager 2007 or Microsoft Operations Manager (MOM) 2005

 
Categories: Operations Manager

Ops Manager: Reporting on Logical Disk % Free

March 18, 2009 2 comments

I saw a few posts about this in the Operations Manager news groups and decided to outline the steps to achieve a report for Logical Disk % free space.  The steps I outline will show you how to set up the Generic report for “Performance” to report on:

· Particular instances (or all instances) on a server or multiple servers.

· Particular instances (or all instances) on servers in a group (maybe in several groups, but you would have to play around with the settings)

· High level initial views and charts of performance with several drill-down reports that can be subscribed to or saved to favorites (several detail reports nested in this report).

· Known issues (OpsManager Reporting) with large numbers, but this is a percentile report so you shouldn’t run into the funny looking numbers in the details for samples.

· Be mindful of the date range. All too often you might be running a report with a time range of from 9 PM today to 12PM tomorrow, but then the day changes as you are configuring the report or running the report, therefore it’s best to switch to advanced in the time range and select -1 hour, day, week etc). A messed up date range may return no results and no errors, thus causing you to think there is no data.

Here are the steps (if you have verified or know that the “% Logical Disk Free Space 200X” collection rule is enabled and not overridden, then you can skip these verification steps: 1 -3):

1) In the authoring pane select rules and change the scope to Windows Server 200X Logical Disk. Searching for Logical should limit the scope as seen below:

clip_image002[6]

2) Once you are scoped to those targets look for performance collection rules starting with “% Logical.” Once you do that you should see the rules listed:

clip_image004[6]

3) Ensure that they are enabled (should be by default) and that they are no over rides applied to the servers you wish to report on.

4) Open the reporting pane.

5) Select “Microsoft Generic Report Library” from the reporting library and then select the report called “Performance”

clip_image006[6]

6) Open the report “Performance” and you should see the following:

clip_image008[6]

7) Change the “From” field to something in the past.

clip_image010[6]

8) After you change the time range click on the button labeled “Change…” and you should see the following:

clip_image012[6]

9) Click the button “New Chart” and give the chart a label. In my example I have two charts and they are described based on what they are collecting. Make your Objects box look similar to this:

clip_image014[6]

As you can see from above I have customized each chart name to represent what is being reported on (to some degree). I could have put a company name in there, application name, etc. Label them however you want, but make sure it will make sense to whoever is viewing them. Below are detailed images of the criteria you will need to select to build the chart. For the first chart I only selected two servers and their instance of C:\. I could have added all C:\ objects (on all servers), but for the example I show just selected two. Be sure you have the collection rule specified and that the specific instances match (I am not 100% certain if you need to specify the instance when you are selecting the object).

· Criteria for % Logical Disk Free Space 2003 for C:\ object on SERVERA and SERVERB:

clip_image016[6]

· Criteria for % Logical Disk Free Space 2003 for D:\ object on SERVERA and SERVERB:

clip_image018[6]

· Criteria for % Logical Disk Space 2003 for all logical disk instances on all servers in the Management Server Computer Group and Root Management Server Computer Group (you can choose other groups, but keep them to computer groups):

clip_image020[6]

· Run the report and you should get a report that looks something similar to this:

clip_image022[6]

clip_image024[6]

Now you can drill deep down to get full details on all of your logical disks. If you need these reports for groups of servers, or just certain instances for your entire agent managed servers, then be smart about setting these reports up. Consider the amount of data you are requesting (instances/objects/time range) before clicking “Run Report”.

Categories: Operations Manager

Ops Manger: Notes on hotfixes and patching (Windows Server 2008: Agent Role)

March 13, 2009 Leave a comment

Last night I installed a Ops Manager SP1 agent on a Windows 2008 64x Server.  After the install I saw some odd errors and decided to investigate.  From what I saw, I decided to uninstall the agent and look for solutions.  There were several hits returned in my search for advice and one seemed to hold the answer I was looking for.  I brought this up to our internal teams, but there was some discussion around if the patches had to be applied to agents or not.  They wanted clarification.  Luckily Kevin Holman was there to save the day.  He informed us of the following:

MSFT will not support a SCOM SP1 agent running on Windows 2008 unless you have previously applied the OS hotfixes outlined in KB953141.

Which two are those? They are these:

952664 (http://support.microsoft.com/kb/952664/ ) The Event Log service may stop responding because of a deadlock on a Windows Server 2008-based computer

953290 (http://support.microsoft.com/kb/953290/ ) An application may crash when it uses legacy methods to query performance counter values in Windows Vista or in Windows Server 2008

So if you are going to deploy Ops Manager agents to a Windows 2008 server, then you will need to have these two OS hotfixes installed.  If you have deployed agents already and your Windows 2008 servers don’t have the hotfixes, then apply the OS hotfixes (reboot may be required) and you will be good to go and fully supported by MSFT for this agent.

My blog post is intended to help you ONLY in regards to issues with Ops Manager AGENTS (not other Ops Manager roles – just for clarification).  Also, ensure that you have previously applied Ops Manager hotfixes for agent related issues: KB954903 and KB957511

For more information on Ops Manager related hotfixes please visit Kevin’s blog.

Categories: Operations Manager

Ops Manager: Microsoft.EnterpriseManagement.Common.UnknownServiceException: The service threw an unknown exception.

March 12, 2009 Leave a comment

I work in a shared Operations Manager environment so today I created a management pack that would contain all of my custom computer groups (I.E. Exchange: ASIA, Exchange: EMEA, etc).  The reason I did this is so that I can seal the MP (which would contain groups only) and apply any over rides from other unsealed MP’s to my computer groups.

I created the management pack and then started to create my first computer group.  While I was trying to complete the creation of the group in the console I got this exception:

Microsoft.EnterpriseManagement.Common.UnknownServiceException: The service threw an unknown exception. See inner exception for details. —> System.ServiceModel.FaultException`1[System.ServiceModel.ExceptionDetail]: parsing "^.*(mcs|MCS).*)$" – Too many )’s. (Fault Detail is equal to An ExceptionDetail, likely created by IncludeExceptionDetailInFaults=true, whose value is:
Microsoft.EnterpriseManagement.Mom.DatabaseQueryModules.DatabaseQueryModuleException: parsing "^.*(mcs|MCS).*)$" – Too many )’s. —-> System.ArgumentException: parsing "^.*(mcs|MCS).*)$" – Too many )’s.
   at System.Text.RegularExpressions.RegexParser.ScanRegex()
   at System.Text.RegularExpressions.RegexParser.Parse(String re, RegexOptions op)
   at System.Text.RegularExpressions.Regex..ctor(String pattern, RegexOptions options, Boolean useCache)

…etc…etc…etc…

The output was similar over and over again, despite the fact that I changed the criteria for the regular expression.  I found a post on the news groups about something similar (from Anders) and I tried the solution (bounce the SDK service on the RMS and run the console with the /clearcache switch).  This did not fix anything and I got the same error.

I decided to delete the management pack, to see if that was what might have been causing the issue.  I deleted the management pack and created a new one.  I created a group and saved it to the new management pack and it worked.  I am not 100% sure that the management pack was the issue, but once I deleted that I was able to create groups and save them in the new management pack with no issues.

If this happens to you, then I hope you will find this helpful.

Categories: Operations Manager

Ops Manager Patching Issues: Great Blog Article!

March 11, 2009 Leave a comment
Today I was checking my feeds from other outstanding MOM/SCOM Bloggers and I saw this from Daniele Grandini on Quaue Nocent Docent.  If you are concerned about hf’s for Ops Manager, then you should really check this out.
Categories: Operations Manager

SNMP Traps and MobileMonitor for BlackBerry

March 1, 2009 Leave a comment

So I documented about this in the past, but had to revisit it and noticed I might have made a few mistakes <g>.  I will document this again and post it shortly.

Categories: Uncategorized

HP StorageWorks MP 2.3 for MOM 2005

March 1, 2009 Leave a comment

In the past a lot of people wondered about this management pack.  How good it was for monitoring their HP EVA storage solution.

We have been dealing with this MP since it initially came out.  HP released the 2.3 version of this MP several months ago and it has fixed a lot of the issues we were seeing.

One issue we discovered with an older version of the MP is that a patch to WEBES on our EVA management servers broke feeds to our MOM servers with the WEBES Listener service.  I believe this is a known issue and well documented.

Now we are running the right version of WEBES on our EVA servers and the feeds are working as expected.  In the past you would have to add the "ghost" machines to a computer group, but that is no longer the case, as HP has enabled the MP to now display the server the alert was forwarded from.  In the past you would see the alert on your Management server and the details would show you which EVA system was broken, but now the host and domain appear in the console (remember the management server is raising these alerts via the XML feed similar to SNMP Traps that might be fed in to MOM).

So now when you import this MP you just have to add the management servers that have the webes listener service in to the WEBES listener computer group.  If you are not using SNMP to get alerts from the various hardware components on the EVA’s, then I suggest you disable this RG as it might cause noise.

Outside of the default rules we have added a few performance collection and threshold rules.  I have also added a rule that looks for the DESTA service crashing, if it does crash, then it raises an alert and tries to restart.  Ensure this rule is running against the management server hosting the listener service.

There are a few other customizations I have done per the request of our EVA engineers and the MP has been doing very well.  The next step is testing it with a new version of WEBES that will be rolled out to our EVA boxes as well as testing this MP in SCOM.  From what I can tell this maybe a converted MP, so don’t expect many changes in SCOM.  I will document all of this in more detail as I work through the deployment.

If you are having issues with the EVA MP, then let me know and I will do my best to assist you.  From what I have seen over the past year or so, you shouldn’t be having any new issues.  The only issue you might have is if your WEBES service on your EVA’s is not compatible with the listener service on your management server.

Categories: Uncategorized