Thursday, November 1, 2012

IBM Storage MP updated (XIV special)

IBM had updated the Storage MP a while ago. Now I had my first touch with it in a productive environment and I’d like to share my findings.

First, glad to see that, in general, my guide is still valid: IBM Storage MP for XIV

But there are some things I’d like to mention, especially regarding XIV again:

  1. The MP supports SCOM 2007 (R2) and SCOM 2012 as well
  2. If you update from 1.1 to 1.2 you have to update all server that should monitor the XIV. Other nodes running SCOM-control utility would not work anymore
  3. Finally IBM changed the discovery. So the workflows will only fired on systems with the SCOM-control utility and not on all Windows Computers as before (if you did not override it by your own
  4. You can use any agent based system as monitoring host and don’t have to use a MS (and RMS in 2007 environments)
  5. It might be a good advice to put the SCOM-control utility on a system with no volume on the monitored storage boxes
  6. Configure discovery intervals (usually 84600sec should be fine to discover XIV Systems and Monitoring Computers)
  7. Installing and configuring the integration on more than one server will result in multiple discovered instances

Get the recent MP here.

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

Wednesday, October 17, 2012

Antivirus exclusions for Operations Manager 2012

There is a knowledge base entry regarding what virus scan exclusions should be made when using SCOM: http://support.microsoft.com/kb/975931

Unfortunately there is a documentation bug regarding agent exclusions for SCOM 2012:

The KB says:

C:\Program Files\System Center 2012\Operations Manager\<Component>\Health Service State\

what means

C:\Program Files\System Center 2012\Operations Manager\Agent\Health Service State\

but the default path is

C:\Program Files\System Center Operations Manager\Agent\Health Service State\

Further take note that you can (should) exclude the “Health Service State” folder and all it’s subfolders.

In general it is recommended to verify your exclusions by creating a text file in the folder and paste in the EICAR test definition:

X5O!P%@AP[4\PZX54(P^)7CC)7}$EICAR-STANDARD-ANTIVIRUS-TEST-FILE!$H+H*

More information about the EICAR test “virus” can be found here:

http://eicar.org/86-0-Intended-use.html

Annotation for Symantec Endpoint Protection customers: be aware that the prefix-variable targets to “Program Files (x86)” on 64bit machines. So you have to type in the whole path by your own.

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

Wednesday, October 10, 2012

Update Rollup 3 for System Center available

Update Rollup 3 for System Center has been released. And finally it is available through Microsoft Update and WSUS for SCOM as well.

More information and the manual download can be found here:

http://support.microsoft.com/kb/2756127

As always I strongly recommend to wait with the update for the blog post of Kevin Holman:

http://blogs.technet.com/b/kevinholman/

He always provides good information and step-by-step guides for the most recent update.

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

Tuesday, October 9, 2012

MMS is back in Vegas!

Yesterday Microsoft announced that MMS is back in Las Vegas and not in New Orleans.

The summit will happen from April, 8th to 12th, 2013. So save the date!

In order to be one of the first to know about the opening of MMS 2013 registration, send mail to mmsnotfy@microsoft.com to subscribe to MMS News & Updates.

Read Full Post...

Wednesday, August 8, 2012

SCOM 2012 Network Discovery with primary IP

In a customers environment I recently had the behavior that multiple network devices have not or have been discovered with other than the primary IP address.

After some deep dives into SCOM and it’s brand new network monitoring I found out that there is a config file that can be modified.

Within the file you find two paragraphs were one has a variable called “AccessAddressFormat” with a default value “AUTOADDRESS”. After changing the value to “LOOPBACKADDRESS” as shown in the screenshot below, restarting the Health Service and running the discovery rule again, the devices have been discovered as expected.

discovery.conf

Hint: don’t forget to delete the devices with the unwanted IPs and remove them from the exclusion list of your discovery rule!

Further there is another variable called “DisplayNameFormat” with a default value “AUTOASSIGNED” as shown in the screenshot below. If your network devices have a proper MIB-II sysName configured and you do not have DNS entries for those devices, you can change the value to show the sysName value as Display Name.

discovery.conf

Here is the default path for the config file:

C:\Program Files\System Center 2012\Operations Manager\Server\NetworkMonitoring\conf\discovery\discovery.conf

Don’t forget to copy it to all of your Management Servers in your Network Devices Resource Pool.

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

Tuesday, July 31, 2012

IBM Hardware MP for IBM BladeCenter and System x

IBM has released (very silently) the new Management Pack for SCOM 2012.

Features include:

  • Support Microsoft System Center Operations Manager 2012
  • Extensive monitoring of the health of hardware components for IBM System x servers and BladeCenter x86/x64 blades running Windows
  • Rich monitoring of the health of BladeCenter chassis and BladeCenter modules via the SNMP protocol
  • Comprehensive monitoring of the health of software stacks for managing IBM hardware
  • Easy determination
  • Power Monitoring of UEFI/IMM System X Servers and Blades running Windows 2008 & R2. Offers the ability to monitor overall system power usage, and generates alerts when power consumption rises above predefined consumption thresholds
  • BladeCenter and Blade hardware health correlation and event propagation providing BladeCenter specific hardware health condition monitoring under the Windows health explorer view.
  • Remote power on and off of Blades Servers via the Operations Manager console.
  • Hardware Management Software Configuration Advisor for IBM Systems detects the presence of IBM Hardware MP software dependencies in order to make appropriate configuration recommendations.
  • Set custom power consumption thresholds for Power Monitoring alerts via Operations Manager Agent task
  • Enable Power Capping and set the maximum power consumption wattage via Operation Manage Agent task

You can download the Installation Files, Release Notes and Guide from the following URL:

http://www-947.ibm.com/support/entry/portal/docdisplay?lndocid=migr-5082204

The release notes and the guide as well are well documented. So I recommend to read that stuff first!

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

Tuesday, July 24, 2012

PowerShell: show Property Bag data

Of course this is not rocket science but  thought this might still be useful when developing or troubleshooting a Management Pack with PowerShell modules:

Your script body usually looks like this:

<ScriptBody>
  <![CDATA[
  # Always use the param statement to access parameters
  param($myArg1)
             
  # Get access to the scripting API
  $API = new-object -comObject "MOM.ScriptAPI"
  # Create the property bag
  $BAG = $API.CreatePropertyBag()
  # Do your magic script stuff here
  # Populate the property bag with data
  $BAG.AddValue("State","Healthy")
  $BAG
  ]]>
</ScriptBody>

The “problem” here is that if you run your script directly in PowerShell for troubleshooting you’d always get “System.__ComObject”.


After adding the following line to your script you’ll get the XML data that has been added to the property bag:

$API.Return($BAG)
Not a big thing but helpful at all…

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

PowerShell Modules & Parameters

Last week I created an advanced Management Pack for a customer to monitor their VMware environment [1]. To achieve their needs the VMware admin created the PowerCLI (PowerShell with VMware Snap Ins) scripts and provided them to me. My part of the task was to implement the scripts into SCOM PowerShell workflows.

There is a real good guide for creating a PowerShell workflow by our friend Stefan Koell so it is not necessary to invent the wheel again. You’ll find all parts of the blog here, here, here and here.

So I created the MP and imported it into my MG but unfortunately it did not run as expected. After some troubleshooting (that was hard, even in the traces I was not able to find the issue) I found out that the parameters from the MP itself were not given to the scripts.

After reviewing the scripts I realized that there was a very simple part missing: The script author used $args variable instead of declaring the parameters.

So after declaring them by adding

Param(
  [string]$firstargument,
  [int]$secondargument,
  [string]$thirdargument
)

and replacing the $args[…] by the variables declared before everything worked as expected.


That was easy but still tricky to troubleshoot…


[1] There are 3rd Party solutions from our partners available. You’ll find them here.


All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

Thursday, March 1, 2012

Invalid Discovery Source

Today I run into an issue when I tried to remove discovered objects from SCOM 2012 RC using the Remove-SCOMDisabledClassInstance CMDLET.

You can use this command to remove previous discovered instances. Before that you have to disable the discovery for the objects you like to remove by using the specific override. Be careful, you should really know what you do!

The issue mentioned above occurred when I fired the CMDLET in the Operations Manager (Power) Shell. Here is the output, I highlighted the important part regarding this problem:

PS C:\> Remove-SCOMDisabledClassInstance
**WARNING** This operation WILL result in class instances and relationships being permanently deleted.  This operation i
s irreversible and will likely result in a significant amount of activity in the operational database.  Are you sure you
want to do this? [Y/N]
y
Started.  This operation make take a very long time to complete.
Remove-SCOMDisabledClassInstance : Discovery data generated by invalid discovery source. Id:8F02035B-FDFE-690F-9E62-FA8D4EF0F73E.
At line:1 char:33
+ Remove-SCOMDisabledClassInstance <<<<
    + CategoryInfo          : InvalidOperation: (Microsoft.Syste...nstancesCommand:RemoveSCDisabledClassInstancesComma
   nd) [Remove-SCOMDisabledClassInstance], DiscoveryDataIn...SourceException
    + FullyQualifiedErrorId : ExecutionError,Microsoft.SystemCenter.OperationsManagerV10.Commands.RemoveSCDisabledClas
   sInstancesCommand

Here is my workaround for that. Please be aware that this is completely unsupported and should not be used when you don’t know what you are doing!

I fired the query below against my Operations Manager DB to find the Discovery source:

USE OperationsManager
SELECT *
FROM [OperationsManager].[dbo].[DiscoverySource]
where DiscoveryRuleId = '8F02035B-FDFE-690F-9E62-FA8D4EF0F73E'
GO

Check if the TimeGeneratedOfLastSnapshot is pretty old. Get the DiscoverySourceId for the next step. Then I verified if there is an existing relationship for this particular DiscoverySource:

USE OperationsManager
SELECT *
FROM [OperationsManager].[dbo].[DiscoverySourceToRelationship]
where DiscoverySourceId = '<the DiscoverySourceId from the query above>'
GO

Because there is no relationship it seems that the DiscoverySource has not been removed successfully.

So the next step is to update the DiscoverySource table and set the IsDeleted column for the DiscoveryRuleId from the first query to “1”. Because this is really not supported I will not provide the statement here. I’m sure you find it by your own and if not: kids, don’t try that at home (or at work)!

Done that the CMDLET should run successful (as long as there are no problems with other discovery sources):

PS C:\> Remove-SCOMDisabledClassInstance
**WARNING** This operation WILL result in class instances and relationships being permanently deleted.  This operation i
s irreversible and will likely result in a significant amount of activity in the operational database.  Are you sure you
want to do this? [Y/N]
y
Started.  This operation make take a very long time to complete.
Completed.  The operations took 0 hours and 1 minutes to complete.

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

Wednesday, February 29, 2012

OM12 Dashboards fail after upgrade to RC

You upgraded (or think about to upgrade) your SCOM 2007 R2 environment to SCOM 2012 RC?

Well, if you installed the SCOM databases using the DBCreateWizard tool (for instance because based on SQL 2008 R2) you might run into the issue, where every access to the DWH will fail. Unfortunately the new OM12 dashboards provided in the Network Montoring views (and others as well) will be concerned.

Reports are still available (or at least should be) because accessing the DWH through SRS.

At the moment there is no workaround available but PG is working on that issue.

Update: there is a workaround available now. You can open an issue in the Connect portal and refer to the following issue:

RC- Network Node Dashboard contains no data

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

Tuesday, February 28, 2012

Network Devices not discovered?

Today a customer of mine run into an issue when he wanted to discover network devices with SCOM 2012. So he added the devices to the explicit discovery rule and started the discovery process immediately.

After a while he checked for the new devices in the Network Devices view (Administration, Network Management, Network Devices) and was not able to see them in this view.

So he double checked the devices in the discovery rule and started again – same problem.

As it seems, the devices are excluded from the discovery. To check that we had to export the Network Discovery Internal MP (unsealed). After that we opened it with a XML or Text editor and went <ExcludeIP> tag and usually the IPs are listed there. (Be aware that there might be several rules and each has its own exclude- list!)

We simply removed the whole line with the device IP, increment the MP version number and import the MP again.

Now we ran the discovery rule again and after a while the devices occur in the view as mentioned above.

At this time we where not able to reproduce the matter for this behavior but I assume that the devices had been deleted manually in SCOM during the testing phase what will put them into the exclude list (to be sure that they won’t be discovered again by accident).

Note: for easier following the discovery process you can filter the Operations Manager log for source OpsMgr Network Discovery. Event ID 12008 includes a summary and there you can check for the number of discovered devices as well as for the amount of excluded ones.

Update: when using Recursive Discovery (= auto-discover) you can remove those devices in the wizard of the discovery rule as well. Find them in the Exclude Filters page within the discovery rule properties.

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

Thursday, February 2, 2012

SCVMM 2012 RC: high CPU load

Today I run in an issue with SCVMM 2012 RC: I experienced high CPU load on my VMM machine. The task manager showed me that the vmmservice.exe consumed all CPU time.

After disabling the CEIP (VMM Console, Settings, General, Customer Experience Improvement Program Settings) the problem was solved.

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

Monday, January 16, 2012

Agent not working after flushing the Health Service State

After running the "Flush Health Service State and Cache" task from the Operations Console you may notice that the agent is (maybe) healthy but neither expected state changes nor alerts appear.

You should check the "Operations Manager" event log on the agents machine for reoccuring Event ID 103 from Source "HealthService".

In that case you have to flush both, the (Root) Management Server's "Health Service State" and "Config Service State". Doing that by stopping the "System Center Management" and "System Center Management Configuration" service and delete (or rename) the folders.

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.
Read Full Post...