Pete Zerger

Coming Soon: Configuration Manager 2007 Unleashed (with coverage of R2! )

Sat, 04 Jul 2009 03:12:14 GMT

The Configuration Manager 2007 Unleashed authoring team has announced their book complete and off to the printer. We hosted the lab environment used by the authors, so I can attest that the book must have been a MASSIVE undertaking. And from what I previewed, it’s going to be an essential reference, just like the Operations Manager 2007 Unleashed book.

Congratulations to the Kerrie, Byron, Greg and the entire Unleashed authoring team!

Pre-order your copy at Amazon.com by clicking on the picture below

OpsMgr 2007 Web Console Series Part 3: Publishing the Web Console through ISA 2004

Sat, 04 Jul 2009 01:08:17 GMT

Previous Installments:

In part 1 we configured our Web Console for forms authentication and installed an SSL certificate to support authentication across the Internet. In part 2, we branded the console with our corporate logo. In part 3, we're going to publish our Web Console through ISA 2004 so it is fully accessible from the Internet.

Steps below assume:

SSL Certificate: An SSL certificate with a publicly routable DNS name is installed on the Web Console web site.
Public DNS Entry: A DNS entry on your public DNS server for the FQDN on the certificate and a public IP address has been created.

If you've not completed either of these steps, you should complete these tasks before proceeding.

To publish the URL:

Open ISA 2004 Console. Right click Firewall Policy and select New --> Web Server Publishing Rule.

Settings on your Web Server Publishing Rule should appear as follows:

Tab \ Setting

General: Enabled should be selected.
Action: Action to take = Allow
From: Anywhere
To: FQDN of your Web Console. (When you ping this name from the ISA Server, it should resolve to the INTERNAL IP assigned to the Web Console web site.). Select the 'Forward original header' checkbox. Select the radio button 'Requests appear to come from original'.
Traffic: Rule applies to the following protocols: HTTP, HTTPS
Link Translation: Select 'Replace absolute links in web pages'.
Users: All Users group should be in the list.
Listener: Listener should contain the public IP in your DNS record, and be configured to listen on the same ports designated on the Bridging tab.
Bridging: The Web Server radio button should be selected. The Redirect Request to SSL Portcheckbox should be selected, and the appropriate listening TCP port on the Web Console web site (443 by default) listed in the text box provided.
Paths: /*
Public Name: This rule applies to requests for the following web sites: Click Add and enter the FQDN on the SSL certificate you installed on the Web Console web site (if you have an equivalent internal DNS entry) or the internal IP address of the Web Console.

NOTE: Since we only want SSL to be used, and with forms authentication enabled you really do not need to publish HTTP.

You’re done. Now you should be able to access your Web Console

In Part 4, we'll look at how to modify the Alert view link in your notifications to reflect the publicly reachable FQDN in the link to the alert.

OpsMgr 2007 Web Console Series Part 2: Branding the Web Console in Operations Manager 2007 R2

Sat, 04 Jul 2009 01:00:13 GMT

Now that we've made our Web Console accessible on or off the domain via a login page (See Part 1), let's consider branding. What if I want to put my corporate logo on the generic login graphic?

This is easily accomplished actually.

Simply drill down to the %programfiles%\System Center Operations Manager 2007\Web Console\Images directory.
Find the file named Log_in.png.
Make a copy of this file so you can revert back to the original image if necessary.
Add your logo to the upper corner of the file, and you're all set!

In the next installment, we'll publish our Web Console securely through ISA 2004/2006.

Part 1: Secure Publishing of the OpsMgr 2007 Web Console to the Internet

Thu, 02 Jul 2009 21:18:05 GMT

The web console has come a long way in the last release, so no doubt more administrators are using this tool. And the whole point is to bring some flexibility to our lives as OpsMgr administrators, so accessing the Web Console securely from anywhere is a fairly common need.

There are a couple of things to be aware of where the Web Console is concerned:

If your Web Console is not hosted on the Root Management Server, be aware of the Kerberos double hop issue. In short, this means that Windows will not forward impersonated credentials from the remote Web Console to the RMS. In short, impersonated credentials cannot be passed from the server hosting the Web Console to the Root Mgmt Server.
To see this in action, setup a remote Web Console and then access first remotely from your desktop, and then again locally from the console of the server hosting the Web Console. (More on the double-hop issue HERE).

Securing Credentials Traversing the Internet

Secondly, remember when accessing from the Internet, we don't have domain credentials that can be passed, so this will be a problem when exposing the Web Console to administrators on the Internet as well.

On the LAN, we'd use Kerberos delegation (or constrained delegation ) to work around this limitations, essentially configuring computers to be "trusted for delegation" in the Active Directory, allowing the forwarding of impersonated credentials. But in our case, we need to actually present a login UI to capture credentials from the user.

In the .NET and OpsMgr world, the means to this end is to configure the Web Console to use Forms-based authentication. Forms-based authentication is an ASP.NET authentication service that enables applications to provide their own logon UI and do their own credential verification. ASP.NET authenticates users, redirecting unauthenticated users to the logon page, and performing all the necessary cookie management. This sort of authentication is a popular technique used by many Web sites.

Like any .NET web application, this setting is actually controlled in the web.config file for the application. For the OpsMgr Web Console, you'll find this in the %programfiles%\System Center Operations Manager 2007\Web Console directory on the server hosting the Web Console.

You can easily uninstall and reinstall the Web Console, selecting Forms authentication on the reinstall, or you can simply edit your web.config file to enable Forms authentication. If Forms-based authentication is enabled, or if you wish to enable it, the section in your web.config file should be configured to look like this

You'll note here that requireSSL is set to true, which is the default when using Forms-based authentication with the Web Console. I would recommend against using Forms-based authentication without SSL, as it will result in Active Directory credentials traversing the Internet in clear text. So for the solution to be functional and secure, you'll need to secure the Web Console site in IIS with an SSL certificate from your certificate authority or purchase one from one of the many companies who issue SSL certificates (at a cost) on the Internet.

The result is a login UI rather than the Web Console being launched directly without a prompt for credentials (see image below). And when deployed with SSL, you can enjoy secure, anywhere-access to your Operations Manager 2007 deployment.

In a future post on this topic, we'll cover how to publish the Web Console through ISA 2004/2006.

Instant Software Deployment in Essentials 2007 - When wuauclt /detectnow just is not fast enough

Thu, 02 Jul 2009 21:05:33 GMT

I had a customer in need of truly instant gratification related to a small software component. This trick will work in OpsMgr 2007 as well, but is not something I encourage.

Scenario:

Support engineer is required to uninstall and reinstall a small software component prone to failure on demand. In the past, this required the engineer to login as local administrator (users do not have local admin rights), uninstall, reinstall and logoff with the user waiting on the phone.

Given this has to be performed on demand, several times a week, using the normal software deployment functionality with the wuauclt /detectnow task really just did not meet support the need for instant gratification.

Workaround:

Create uninstall and installation tasks using the Windows Installer (msiexec), which provides well known command line parameters to silently install uninstall. This can can easily be accomplished using the msiexec -x (uninstall) and /qn (silent) flags.

As long as the package does not run beyond the task timeout limit specified when creating the task, it works well, and reports success or failure in the task window.

The Uninstall Task

Here's how you'd configure the task. Notice the software package is located on a network share with Everyone READ permissions. FYI- This was not the point from which the package was originally installed.

The Install Task

This looks much like the Uninstall Task, just minus the -x flag.

Hope you find this tip useful! Please post feedback via comments on this post

SCOM 2007 R2: How to test Cross-Platform Monitoring without purchasing a RHEL license

Tue, 30 Jun 2009 17:13:47 GMT

I’ve had a couple of questions on this in the last week, and wanted to post the how-to here:

Q: How can I demonstrate or test cross-platform monitoring in SCOM 2007 R2 without purchasing a RHEL (RedHat), Suse or other license?

A: In truth, both of these distributions have a trial download (60 days in each case I think), but there is also a way you can setup a long-term demo with no license purchase at all using CentOS. We’ll simply change the OS caption…Read on for details on the process and why CentOS looks so much like RHEL.

This scenario is very real to me, as we do have a couple of VM hosts running CentOS 5 and VMware Server in my organization. So while it may not be supported, by one method or another,monitoring our CentOS systems is something we have to do

While unsealing and modifying a copy of the RHEL 5 MP is possible as has been alluded to in the blogosphere, I wanted to point out a much faster method for testing monitoring of CentOS 5.x when you are learning and experimenting in your non-production scenarios. Truly I hope MS is someday able to provide support for CentOS and OpenSolaris in particular, as I think these platforms are widely used by developers and somewhat more predictable in their similarities to their commercial counterparts.

How Similar are Red Hat Enterprise Linux (RHEL) 5 and CentOS?

For anyone that may not be aware, Red Hat releases all source code for the product publicly under the terms of the GNU General Public License and other licenses. CentOS developers use that source code to create a final product that is very similar to RHEL; the logos must be changed, because Red Hat does not allow them to be used for redistribution. If you look at the release dates, you will notice CentOS releases come very quickly behind their Red Hat counterparts, generally within 4-5 weeks, as shown HERE. In short, while there is no guarantee of their ultimate similarity, Red Hat and CentOS are in the end very similar at present.

Discovering and Monitoring CentOS

As mentioned, I've seen posts on unsealing management packs, etc., and wanted to point out the "easy button" method for this. If you look at the Microsoft.Linux.RHEL.5 MP, you'll see the discovery for RHEL systems involves a simple query to the CIM Namespace root/scx to find the "Caption" of the operating system (basically a display name), as shown in figure 1.

Figure 1 - OS Discovery for RHEL Systems

Clearly this will be very different on a CentOS server versus a Red Hat server. However, this is easily changed for testing purposes, with no ill effects that we have yet been able to identify after several days.

Changing the OS Caption on a CentOS Server.

This value is stored in the /etc/redhat-release file. In fact, this Caption is the only value stored in this file. So, first make a backup of this file. The value present on a CentOS system will be something similar toCentOS release 5 (final). If you change this value to reflect the value on a standard RHEL 5 or 5.1 system. I used Red Hat Enterprise Linux Server release 5.1 (Tikanga), which I pulled straight from the redhat-release file on an RHEL 5.1 system. This changes the value returned for the OS Caption when the root/scx namespace is queried by the object discovery, as well as how the Unix and Linux Computer Management Wizard identifies the target computer.

Results

After making this change, I ran the Unix and Linux Computer Management Wizard and successfully pushed an agent to the server. Within a couple of hours, the same discovery, monitoring and performance collection data present for my RHEL 5.1 server also seemed to be working for my CentOS server - another testament to the similarity of the two. As for production scenarios, I am taking a wait-and-see attitude in hopes that we'll see some unofficial guidance from the product team at some point down the road.

OpsMgr 2007 Health Check – Finding rules and monitors targeted to groups

Tue, 30 Jun 2009 06:16:13 GMT

We’ve been talking about the Requirements for a SCOM 2007 Health Check script that is freely available, and amongst the list is identifying rules and monitors targeted to group. I’ll put a bulleted list together based on the results of this thread, so if you have any thoughts on requirements for determination of a healthy SCOM 2007 installation, add them to the list HERE. This could be a really useful tool for periodic checks, and perhaps the only option if you do not have a Premier support contract. (Premier customers can get the full exam as part of a professional services engagement)

Here is some T-SQL for identifying Rules and Unit Monitors targeted to a GROUP in error. This doesn't get us to an absolute list of culprits, but does return a very small result set (including name, target and MP) very quickly in which we can easily identify mistakes. You will find a couple of singleton classes targeted by rules which are not groups in some internal MPs, but these queries get us very very close in tests I conducted in the lab.

If we were to query a level deeper to identify base class, we could be 100% accurate (assuming we correctly identified all base classes for a group), but stopped here due to time constraints.

I tried to identify the same information with Powershell and successfully did so, but found it took much longer to execute, thus the move to T-SQL to minimize impact to MG resources. (PoSh took a couple minutes to execute, T-SQL took only a couple seconds). T-SQL seems a bit easier to put together for these types of checks than PoSh in some cases.

For Unit Monitors

SELECT * from dbo.ManagedTypeView.DisplayName AS Target, dbo.MonitorView.DisplayName AS MonitorName,
dbo.ManagementPackView.DisplayName AS MP, dbo.ManagedTypeView.Singleton
FROM dbo.MonitorView INNER JOIN
dbo.ManagementPackView ON dbo.MonitorView.ManagementPackId = dbo.ManagementPackView.Id INNER JOIN
dbo.ManagedTypeView ON dbo.MonitorView.TargetMonitoringClassId = dbo.ManagedTypeView.Id
WHERE (dbo.ManagedTypeView.Singleton = 1) and dbo.MonitorView.IsUnitMonitor = 1
ORDER BY Target

For Rules

SELECT TOP (100) PERCENT dbo.ManagedTypeView.DisplayName AS Target, dbo.RuleView.DisplayName AS RuleName,
dbo.ManagementPackView.DisplayName AS MP, dbo.ManagedTypeView.Singleton
FROM dbo.RuleView INNER JOIN
dbo.ManagementPackView ON dbo.RuleView.ManagementPackId = dbo.ManagementPackView.Id INNER JOIN
dbo.ManagedTypeView ON dbo.RuleView.TargetMonitoringClassId = dbo.ManagedTypeView.Id
WHERE(dbo.ManagedTypeView.Singleton = 1)
ORDER BY Target

Let me know if you have anything to add to the discussion as a comment to this post or by joining the forum discussion mentioned above.

SCOM: Report Models for Operations Manager 2007 SP1 (and how to install and use)

Thu, 25 Jun 2009 15:25:16 GMT

Using Report Builder with Operations Manager 2007 SP1 has not been as widespread as it might have been otherwise because there are no report models included with this version of SCOM. Do you want an answer to this issue? Simply look on the OpsMgr 2007 R2 media

On the R2 product media in the “Report Models” folder, you will find two report models that can be used with the OpsMgr 2007 SP1 release as well! Look in the Report Models –> Other and you’ll the following:

The bottom line – if there are no schema changes in the DW (or you take care to use Views in your reporting components), it’s definitely possible

To Install These Models in SQL Reporting Services for use with SCOM 2007 SP1

With these models, you will need to upload the model and then update the datasource it is pointed to so it will function.

Browse to the http:///reports URL on your SCOM 2007 SP1 Reporting Server.
Click the View Details button, and then the Upload File link.
Click the Browse button, and then browse to the folder containing the files. Pick either one and click Open, and then OK.
On the Home page of the Report Manager page, click the Edit button

Click the Data Sources link, followed by the Browse button to update the data source.
Browse to “Data Warehouse Main” following by OK and then Apply.

To Use the Models

In the Operations Console, you can access these models from the Reporting space by clicking the “Design a Report” link to launch Report Builder.

If I’m able to design some compelling reports with the models, I’ll post them here.

Powershell Tip: Working with TRUE and FALSE Comparisons in Powershell

Wed, 24 Jun 2009 20:29:34 GMT

I posted an updated Powershell script for a client to resolve rule-generated alerts in SCOM 2007, but also to illustrate an example of TRUE and FALSE evaluation in Powershell.

The fact is, it’s really about $TRUE and $FALSE. Powershell has a more accurate notion of true / false than one might imagine. I just wanted to add a pointer to some clarifying information for the benefit of all who may not be familiar.

My client had written a script that attempted TRUE / FALSE evaluation using the –eq comparison operator as though “TRUE” and “FALSE” evaluation where strings, and they were getting an unexpected results, getting a true match when a false was expected, and vice versa. (Code snippet below)

get-alert | where-object { ($_.LastModified -lt $targetdate) -and ($_.ResolutionState -eq 0) -and ($_.IsMonitorAlert -eq“false”)}

This is actually a Boolean comparison, and using –eq $TRUE or –eq $FALSE will deliver the expected results.

So what are the potential consequences of using “TRUE” and “FALSE” strings in such comparisons?

Notes from “The Architect”

Powershell architect Jeffrey Snover has a great explanation if you’re not familiar. Jeffrey also explains why the unexpected results are returned as well.

http://blogs.msdn.com/powershell/archive/2006/12/24/boolean-values-and-operators.aspx

Sample Script

I modified one of my old scripts to address rule-generated alerts only. For a “Close Rule-Generated Reports” that treats these as boolean values, click HERE.

SCOM Console Task: Finding currently logged on user from the Operations Console

Tue, 23 Jun 2009 15:18:33 GMT

Someone asked how to retrieve the currently logged-on user on agent-managed computers from the SCE Admin Console. I created a quick Console Task to do this, which could be done with equal success on the Operations Manager 2007 platform.

Start by downloading the PSLoggedOn command line utility from Sysinternals.com (now Microsoft). It can be downloaded HERE.
In the Authoring Pane select New Task --> Console Tasks --> Command line

Settings were configured as follows:

Application: %systemroot%\psloggedon.exe

Parameters: \$Target Property[Type="Windows!Microsoft.Windows.Computer"]/NetbiosComputerName$

I should note here that when I preceded this parameter with the -l (local only) and/or -x (dont show times) I got an InnerException Failure message. SCE didn't like something about those parameters.

Working Directory: c:\windows\temp

And DO check the box for capturing output.

For this example, I dropped the tool in the Windows directory. If you added a \tools directory of some sort to the Path environment variable for these types of tasks, it will make life easier in the long run.

SCOM: Dynamic Group to detect SQL 2005 with (or without) specific Service Pack

Mon, 22 Jun 2009 14:10:00 GMT

Recently, a user was asking about how to create a custom group to report all SQL servers without SP3 in their environment, so we new which servers needed the update.

Here is an example of how to do that using a dynamically-populated group in Operations Manager 2007.

Launch the Operations Console. In the Authoring space, right click on Groups and click "Create a new group".
In the Name field, enter the desired name of the Group and select the destination management pack. Click Next.
Skip the "Explicit Group Membership" by clicking Next.
On the Dynamic Inclusion step, click Create/Edit Rules.
Change the class from the default (Windows Computer) to "SQL 2005 DB Engine". Click Add.
In the Property box, select “Version”
In the Operator box, select “Not Equal”
In the Value box, you can input the following version: 2005.90.4035 (which is SQL 2005 SP3)
Click OK to close the Query Builder window and then click Next to move on to the next step.
Skip the "Add Subgroups" step by clicking Next.
Skip the "Exclude Objects" step by clicking Create.

Now you have a group that only includes SQL 2005 without SP3 and is dynamic so it will always include all SQL 2005 installations without SP3 found by Ops Mgr 2007.

For a list of SQL version numbers based on SP installed, see http://support.microsoft.com/kb/321185

How to convert management packs for OpsMgr 2007 (without a live MOM 2005 management group)

Sun, 21 Jun 2009 20:19:44 GMT

Q: Can I convert MOM 2005 management packs without having an active MOM 2005 Management Group?

A: Yes, this can be done with the MOM 2005 to OpsMgr 2007 Migration Wizard.

NOTE: The wizard does require that the MOM 2005 UI be loaded on the Ops Mgr management server where you run the tool.

The steps would be as follows:

Load the MOM 2005 UI (when prompted for a MOM 2005 Mgmt Server during install, simply leave that field blank).
Make sure the MOM 2005 MP akm files you wish to convert are available (converting MPs from file will be your only option with no MOM 2005 MG)
Launch the Migration Wizard and select Migrate MPs from file and click Next.
Select one or multiple MP akm files you wish to convert in the space provided, then click Next
On the next screen, select the Migrate to file radio button and select a target directory, then click Next.
On the next screen, click the Migrate button. MPs are converted to Operations Manager .xml format and ready for import.

Then, simply import into Operations Manager 2007 per the normal procedure.

One additional note here. Leave conversion of MS management packs to MS whenever possible. The need to convert MOM 2005 MPs should only be occasional at this point in the product cycle.

ResearchThis! KB - The Internet Information Service NNTP Virtual Server named NNTPSVC/1 is unavailable as the virtual server has been stopped

Sun, 21 Jun 2009 16:12:10 GMT

Alert: The Internet Information Service NNTP Virtual Server named NNTPSVC/1 is unavailable as the virtual server has been stopped.

Issue: On Exchange 2003 servers this service is required to install but it is not required after it is installed.

Resolution: If this service is disabled/not in use you can remove it. To remove the service, log into the server and use sc delete NNTPSvc. Or you can create an override to ignore this on Microsoft Exchange 2003 Server Group, as NNTP was required for the installation but can be disabled after the installation has been completed.

Submitted By: Cameron Fuller (MVP)

ResearchThis! KB - Auto Close Flag

Sun, 21 Jun 2009 16:08:12 GMT

Alert: Auto Close Flag

Issue: The auto close flag for database MSCUPTDB in SQL instance MSSQL SERVER on computer 123.abc.com is not set according to best practice.

Resolution: As this is a standard Microsoft application (patch Management for SMS and Configuration Manager) and a default configuration, created an override to exclude this database.

Submitted By: Cameron Fuller (MVP)

ResearchThis! KB - Auto Shrink Flag

Sun, 21 Jun 2009 16:05:28 GMT

Alert: Auto Shrink Flag

Issue: The auto shrink flag for database (DBNAME) in SQL instance MSSQL SERVER on computer 123.abc.com is not set according to best practice.

Resolution: This was found on a series of standard Microsoft applications including SUSDB, WSUS and MSCUPTDB. Additional databases found with the Auto Shrink Flag: Backup Exec, ItAssist, SOE, DSPre, XRXDBDiscovery (Xerox), XRXDBCWW (Xerox). Options available include contacting the vendor to determine if this flag can be changed safely (and changing the flag if it can) or to create an override to exclude this database from monitoring this configuration.

This is sometimes configured by design to minimize maintenance no 3rd party databases used for utility-type purposes.

Submitted By: Cameron Fuller (MVP)

Windows Server Hyper-V MP for SCOM 2007 now available (ver 6.0.6633.0)

Sat, 20 Jun 2009 18:16:50 GMT

The Hyper-V management pack for Operations Manager 2007 (SCOM) has been released:

Overview

This management pack supports monitoring of Windows Server Hyper-V systems. This includes monitoring coverage of Hyper-V host servers, including critical services and disks, and Hyper-V virtual machines, including virtual components and virtual hardware.

Feature Summary

This management pack provides the following functionality:

Management of critical Hyper-V services that affect virtual machines and host server functionality
Management of host server logical disks that affect virtual machine health
Full representation of virtualization in a single Hyper-V host server, including virtual networks, virtual machines, and guest computers
Monitoring of virtual machine hardware components that affect availability’

Where to get it:

Download the MP from the Microsoft site at the URL below

http://www.microsoft.com/downloads/details.aspx?FamilyID=502e7a26-2fea-4052-89fd-8f75142de4f2&displaylang=en&displaylang=en

Command Shell: Dumping all overrides in Operations Manager 2007

Fri, 19 Jun 2009 00:20:07 GMT

Want a quick way to view ALL overrides in your Operations Manager 2007 installation?

Here's a one-liner to dump all overrides in descending order by LastModified date to a spreadsheet, which should tip you off to recent changes enacted via override.

get-ManagementPack | where {$_.Name -like "*" } | get-Override |sort-object LastModified -descending | select-object name, displayname, xmltag, value, timeadded, lastmodified | export-Csv -Path "c:\overrides.csv"

While not perfect, the Name attribute provides some insight into the purpose of the override. Unfortuntely, not all overrides have a friendly name (DisplayName), and I didnt see any trace of who made the changes in the verbose output. The xmltag attribute does reveal if the override is applied to a rule, monitor, etc (values I encountered were RuleConfigurationOverride, MonitorConfigurationOverride, CategoryOverride). The value parameter the current value data for the given override.

In the future I'll look further into the potential of connecting overrides to their associated rules / monitors for inclusion in the output.

Command Shell: Retrieving Agent Count by Management Server

Fri, 19 Jun 2009 00:06:35 GMT

A recent request came for a Powershell script that would return a list of management servers along with a count of the number of agents reporting to them. This would be handy in large environments for validating that you are properly balancing the agent load across management servers.

Here's a sample script that will do just that.

$rootMS = "NOCMS01"
#Initialize the OpsMgr Provider
add-pssnapin "Microsoft.EnterpriseManagement.OperationsManager.Client";
set-location "OperationsManagerMonitoring::";

#set Management Group context to the provided RMS
new-managementGroupConnection -ConnectionString:$rootMS;
set-location $rootMS;

$agent = get-agent | sort Name

$agent | Group PrimaryManagementServerName -Noelement | sort Name | select Name, Count | export-Csv -noTypeInformation -path "c:\agents.csv"

Output

The output is a csv file with Management Server name and count, as shown below.

Name,Count
NOCMS01.noc.systemcenterforum.org,81
NOCMS02.noc.systemcenterforum.org,34
NOCMS03.noc.systemcenterforum.org,12
NOCMS04.noc.systemcenterforum.org,45

OpsMgr 2007: Updating Primary and Failover Mgmt Server settings for multiple agents

Thu, 18 Jun 2009 23:43:43 GMT

I had a question today regarding how one could update the primary and failover management server settings for a number of Operations Manager 2007 agent-managed computers. Powershell comes to mind as the easiest way to make this happen.

This sample script illustrates how to retrieve a list of agents based on a common search element, and then loop through these agents, updating the settings on each. Update setting as instructed in comments, then save with a .ps1 file. The script can be run on any computer with the Operations Console / Command Shell interfaces loaded.

#-------------Begin Sample Script--------------

#Update $rootMS variable with the name of your RMS

$rootMS= "opsmgr.contoso.com"

#Initializing the Ops Mgr 2007 Powershell provider and Connecting to Mgmt Group
     add-pssnapin "Microsoft.EnterpriseManagement.OperationsManager.Client" -ErrorVariable errSnapin;
     set-location "OperationsManagerMonitoring::" -ErrorVariable errSnapin;
     new-managementGroupConnection -ConnectionString:$rootMS -ErrorVariable errSnapin;
     set-location $rootMS -ErrorVariable errSnapin

# Retrieve a list of agents and assign to variable $agent
# In this example, all servers with server name starting with 'FS'

$agents = get-agent | where-object {$_.Name -like 'FS*'}

# set variables for primary and secondary management servers.
# make sure the WHERE clause in each one-liner below matches only 1 MS!

$PriMS = get-managementserver | where-object {$_.Name -eq 'ms1.contoso.com'}
$SecMS = get-managementserver | where-object {$_.Name -eq 'ms2.contoso.com'}

#Loop through list of agents and update primary and failover MS settings

ForEach ($agent in $agents) {
Set-ManagementServer -PrimaryManagementServer $PriMS -AgentManagedComputer $agent `
-FailoverServer $SecMS #| Out-Null
}

#-------------End Sample Script--------------

Hope this helps. Post any feedback as comments on this post.

OpsMgr 2007 (SCOM) v10 Wish List Wrap-up

Wed, 17 Jun 2009 18:36:35 GMT

For those of you that commented on the “v10 Wish List” for Opsmgr 2007, I just wanted to touch base and let you know your feedback has been channeled back to the product team.

Thanks to everyone who commented and special thanks to Brian D. for kicking off the discussion!

Provision/Deprovision entire Virtual Lab Environments with Visual Studio Lab Management 2010 [beta]

Tue, 16 Jun 2009 02:42:58 GMT

VMware has touted the abilities of their vCenter Lab Manager for provisioning, deprovisioning and taking snapshots entire lab scenarios including multiple virtual machines in a matter of seconds. The few vendors busy in the virtual lab automation space (which include VMware, Surgient, VMLogix, Skytap and the almost dead StackSafe) will soon have some fresh competition…Microsoft.

The product just entered the beta 1 phase and has the potential to become a huge hit in the .NET world. I was really wondering if MS was expecting us to duplicate Lab Manager using Powershell…I guess not!

Read the full story at the source (virtualization.info)

Tuning Items for New Essentials 2007 SP1 Installations

Mon, 15 Jun 2009 16:42:56 GMT

While Essentials 2007 monitoring settings are based on "smart defaults", there are some rules, monitors and discoveries not enabled by default deprive administrators of some information I believe most would be interested in. Some of these items are in several cases especially pertinent to MOM 2005 administrators, who are accustomed to receiving certain types of alerts by default. I'm going to point to a few items that are of particular interest ( at least in my opinion) for acceptable alert behavior in Essentials 2007. Of course most of the MP tuning advice you see for Operations Manager 2007 also applies to Essentials. Just bear in mind that with OpsMgr 2007 R2, the Essentials and OpsMgr monitoring functionality sets diverge a bit, most notably with cross-platform not being part of Essentials 2007 SP1.

Global Windows Service Monitoring

Some of the more generic service monitoring rules are disabled by default. I find I generally want to know anytime a service terminates abnormally or a service fails to start. To fix that, use an override to enable the following rules for the Windows Server 2000 and 2003 Operating System object types. You may also want to enable this for XP Client Operating Systems where an agent is deployed.

"The service or driver failed to start. and The service terminated unexpectedly."

Windows Update Roaming

This rule runs every 60 minutes on managed computers to verify that the computer can contact its WSUS server. The rule runs a script to verify that the computer is able to contact its WSUS server. If it cannot contact the WSUS server for 6 hours, then the script will configure it to use the Microsoft Updates site. Once the computer is able to contact the WSUS server again, the script configures it to use the WSUS server for updates.

If you require tight control of what updates are applied to a system, consider disabling this rule.

AD Topology Discovery

Find in: Object Discoveries

While the Active Directory MP requires some configuration, I found this setting particularly strange and potentially confusing for some users. You may find that the Active Directory Topology Diagram and distributed application are basically blank. This may be due to the AD Topology Object Discovery not being enabled. To get to this discovery, go to the Authoring space, and scope your view to Active Directory Domains (this is only visible when the View All Targets radio button is selected ) and in the Navigation (left) pane, select Object Discoveries. You see in the image below, Enabled by default = Yes

I have found that even though this discovery showed "Yes" under enabled by default, that the Effective Value = false, as shown below.

SQL 2000 / 2005

Object Discoveries

Discover SQL 2000 Agent Jobs - without this, you will not be able to override the “Job Duration” monitor for long running jobs on a per-job basis. However, you will receive alerts for failed agent jobs even without this.

Rules

Free Space (MB) counters - these are optional in my book, as they are of varying importance from one environment to the next for trend reporting. All the equivalent % free monitors are enabled, so just decide which works for you. If you have autogrow enabled on databases, then MB free can come in handy.

Windows 2000

The Object Discovery for Windows Physical Disks is not enabled by default (although the discovery of relationships between physical disks and the partitions is).

Network Monitoring

There are several rules disabled by default that definitely may be of interest in most environments. Data collection of inbound and outbound bandwidth utilization on routers and switches would be important, and critical on WAN interfaces for sure, especially to provide a view into the last few hours for troubleshooting purposes. Scope your view under Authoring --> Rules to show only network and SNMP components. Use overrides to enable any of these rules (listed below) for target devices.

Are there other settings disabled by default in Essentials 2007 you find critical in your environment? Leave a comment with your list.

Monitoring Database Blocking Through Rules and Alerts [Article and MP]

Sun, 14 Jun 2009 17:55:48 GMT

Blocking in SQL is handled by separate monitors for the SQL 2000 and 2005 engines. Blocking is watched in SQL 2000 by a monitor called:

Microsoft.SQLServer.2000.DBEngine.BlockedSPIDMonitor

For SQL 2005 Servers, this is handled by a monitor called Blocked SPIDs.

The alert descriptions on these monitors are NULL (and not overridable), so the resulting alerts of these monitors are not always straightforward. For customers concerned about blocking, I've heard some feedback about these monitors. They worry about 1) the accuracy of the monitors and 2) the lack of detail in the alert description. I suspect the lack of description in an alert can lead to concerns about accuracy of the condition identified, as no doubt the OpsMgr Test Team has tested this many times at this point.

Beefing up monitoring of blocking processes

It was past customer experience that peaked my interest when I saw this article from a SQL DBA community site of which I am a member - sqlservercentral.com. The Custom SQL Blocks MP, created by Marios Philippopoulos. Marios has created a workaround that has flexible configuration options - You can tune the frequency and logging verbosity. You definitely see the details in the event and alert descriptions, making identification of the affected processes easy, as shown below:

Alert: Custom - Persistent Blocking Detected in SQL Server Database
Source: compName.domainName.com
Path:
Last modified by: System
Last modified time: 8/24/2008 4:40:53 PM
Alert description:

Server\Instance: servr\instanceName

BLOCKING SPID(S)
----------------

Blocking SPID: 80
Database: TESTDB
Host: compName Program: Microsoft SQL Server Management Studio - Query
Cmd: UPDATE Login: loginName

This is a nice improvement I think.

Implementation and Targeting

He also did a good job with targeting (to the DB Engine classes themselves) to ensure default and named instances alike are automatically picked up. He did use rules instead of monitors, so the health state of the SQL instances is not affected. However, if one wanted to convert these rules to monitors, it would be a relatively easy task.

Where to get it

You'll actually need to register (free) at sqlservercentral.com in order to read and download. You'll find Marios article and MP at the following location:

http://www.sqlservercentral.com/articles/Blocking/64038/

While you're at it, there is a nice TechNet article on Minimizing Block in SQL Server by Cherie Warren. Get it at

http://technet.microsoft.com/en-us/magazine/cc434694.asp

MS Exchange Server 2007 MP for OpsMgr (SCOM) 2007 R2 now available

Sun, 14 Jun 2009 07:32:29 GMT

The R2-only version of the Exchange 2007 MP is out. The marquee feature: a wizard-like interface for configuring synthetic transactions against Outlook Web Access (OWA), Exchange ActiveSync, Web Services, POP3, and IMAP.

DOWNLOAD URL: http://www.microsoft.com/downloads/details.aspx?displaylang=en&FamilyID=e9f3cd3f-9bc0-45cd-b10f-120e937ee4c4

The Exchange Server 2007 Management Pack is designed to monitor Exchange 2007 key health indicators, collect Exchange component-specific performance counters in one central location, and raise alerts for operator intervention as necessary. By detecting, sending alerts, and automatically responding to critical events, this Management Pack helps indicate, correct, and prevent possible service outages or configuration problems, allowing you to proactively manage Exchange servers and identify issues before they become critical. The Management Pack monitors and provides alerts for automatic notification of events indicating service outages, performance degradation, health monitoring, and centralized management.

Feature Summary

A number of synthetic transactions ensure the Exchange servers are available and responding in a timely manner. The synthetic transactions are maintenance-mode aware, so that if the target of a transaction is in maintenance mode, the source will not run the transaction, and not alert unnecessarily.
This Management Pack includes a Management Pack template that provides a wizard-like interface for configuring synthetic transactions against Outlook Web Access (OWA), Exchange ActiveSync, Web Services, POP3, and IMAP.
This Management Pack includes a Management Pack template that provides a wizard-like interface for configuring mail flow synthetic transactions between agent-managed Exchange 2007 Mailbox servers.
This Management Pack provides 30+ reports specific to Exchange 2007 that track availability and performance compared to service level objectives. For the list of reports and for more information about the reports, see the Management Pack Guide.
All the synthetic transactions in this Management Pack use an Operations Manager 2007 R2 hosting feature for Windows PowerShell technology that provides a performance improvement when running synthetic transactions.
A significant number of rules and monitors that are not actionable or may be noisy are disabled. Note that many of these rules are still in the Management Pack so that you can enable them if necessary.
Support for monitoring any number of Exchange organizations using a single Operations Manager 2007 management group.
Full support for Microsoft clustered configurations. For more details, see the Management Pack Guide.
Discovery of Exchange 2007 server roles is disabled by default, and no Exchange 2007 monitoring is applied by default. This allows you to discover and monitor your servers gradually, as well as tune the Management Pack as you bring more agent-managed Exchange 2007 servers into the Operations Manager environment.

ResearchThis! KB - Core Service File Writing

Sun, 14 Jun 2009 01:41:54 GMT

Alert: Core Service File Writing

Issue: Alert created when a new reverse zone was added.

Resolution: This error will occur once after the new reverse zone is added. Logged into the servers reporting the error and verified that the new zone had been created and populated correctly. It can be closed out, and is not an issue unless it recurs.

Submitted By: Cameron Fuller (MVP)

Cross Platform Management Pack Authoring Guide for OpsMgr 2007 R2 (SCOM) Now Available

Fri, 12 Jun 2009 10:13:47 GMT

MS have released the Cross Platform MP Authoring Guide for R2

System administrators of Windows-based, UNIX-based, and Linux-based platforms can use this document as a guide to how to create and use management packs to manage their computer systems. Because Operations Manager 2007 R2 extends monitoring support to UNIX-based and Linux-based servers, organizations can monitor and manage their cross-platform environments from a unified management console. The key to how to configure and run Operations Manager is the management pack, an XML-based file that defines the objects that Operations Manager will discover and monitor and the information that will be collected about these objects.

By using Operations Manager 2007 R2, system administrators of UNIX-based and Linux-based servers can now monitor the health of key system attributes, including the following:

File systems and network interfaces.
Critical processes (for example, syslog, cron, and others).
Key configurations (for example, resolution of host name and correct configuration of Web Services Management components).
Core system attributes (for example, the health of system memory and processors).

Intended Audience for the Cross Platform Management Pack Authoring Guide: Describes the expectations and the minimum knowledge that are required to implement a cross platform management pack.

Getting Started with the Cross Platform Management Pack Authoring Guide: Provides information on the structure and workflow of a management pack.

Management Pack Tasks: Describes typical administrative tasks and how to implement those tasks by using a management pack.

Reference

Online Training

TechNet Virtual Lab: System Center Operations Manager 2007- Introduction (http://go.microsoft.com/fwlink/?linkid=152883)
TechNet Virtual Lab: System Center Operations Manager 2007- Advanced Topics (http://go.microsoft.com/fwlink/?linkid=152884)
HOW TO: Use XML to Create a Basic Management Pack with Discovery for a Windows-based Computer (http://go.microsoft.com/fwlink/?linkid=150433)

Published Documentation

Operations Manager 2007 Management Pack Authoring Guide (http://go.microsoft.com/fwlink/?linkid=150417)
System Center Operations Manager 2007 Unleashed. (includes CD-ROM), Meyler, Kerrie, et al., Sams, 2008, p. 1385

Online Resources

Operations Manager Cross Platform and Interop Solutions (http://go.microsoft.com/fwlink/?linkid=144278)
Microsoft.Unix.Computer: Team blog for System Center Cross Platform and Interop (http://go.microsoft.com/fwlink/?linkid=152885)
Operations Manager Web site (http://go.microsoft.com/fwlink/?linkid=108041)
System Center Web site (http://go.microsoft.com/fwlink/?linkid=118294)
TechNet Webcast: Successfully Monitor UNIX and Linux Alongside Your Windows Infrastructure with Operations Manager 2007 R2 (Level 300)(http://go.microsoft.com/fwlink/?linkid=150435)

Powershell: Updating Proxy Agent for Multiple SNMP-enabled Devices (part 2)

Thu, 11 Jun 2009 23:23:37 GMT

In part 1, we demonstrated how to update the proxy agent for a group of managed SNMP-enabled devices. The difficulty in using the solution presented is part 1 is in using the get-agent cmdlet to retrieve the potential proxy agent to pass to set-proxyagent, which only returns agents; no management servers are presented. Using the get-managementserver cmdlet in it's place results in a failure because it does not return the right object to pass to set-proxyagent.

If I want to use a management server as the proxy for multiple network devices, the script below will allow you to submit a wildcard filter based on device IP address to update the proxy agent for all SNMP-enabled devices matching the filter, and assign a managment server as the proxy agent.

Thanks to Marco Shaw for a critical assist in working out the code to allow passing wildcard for the -deviceName parameter.

INSTRUCTIONS FOR USE

Cut-and-paste the script below into Notepad, and save as updateproxyagentMS.ps1. Script should run from any computer with OpsMgr Console and Powershell installed.This script will load the OpsMgr Powershell snap-in and connect to your Management Group.

SYNTAX

The script accepts the following 3 required parameters. Run with no parameters to view syntax help.

-rootMS: FQDN of the root management server for the target management group.

Ex: 'RootManagementServer.fqdn.local'

-proxyAgent: Management Server that will serve as the proxy agent

Ex: ‘ms1.contoso.com'

-deviceName: IP address or IP wildcard range.

Ex: ‘192.168.1*’

SAMPLE SCRIPT

Here is the sample script. You should be able to cut-and-paste this one with no worries about line wrap.

#-----------Begin Sample Script--------------

param($rootMS,$proxyAgent,$deviceName)

#Echo syntax if -rootMS parameter is NULL
     if ($rootMS -eq $null) {

        Write-Host ""
        Write-Host "-rootMS: FQDN of the root management server for the target management group.";
        Write-Host "Ex: RootManagementServer.fqdn.local";
        Write-Host ""
        Write-Host "-proxyAgent: Agent-managed computer that will serve as the proxy agent";
        Write-Host "Ex: ‘dc.contoso.com";
        Write-Host ""
        Write-Host "-deviceName: IP address or IP wildcard range.”;
        Write-Host "Ex: ‘192.168.1*’”;
        Write-Host ""
 exit;
     }

#connect to mgmt group
$ServerName=$rootMS
add-pssnapin  Microsoft.EnterpriseManagement.OperationsManager.Client ;
set-location  OperationsManagerMonitoring:: ;
new-managementGroupConnection -ConnectionString:$serverName;
set-location $ServerName;

$mg=new-managementgroupconnection $rootMS
$mgname=$mg.managementgroup
$admin=$mgname.getadministration()

#Retrieve management server that will serve as proxy agent
$crit=new-object microsoft.enterprisemanagement.administration.managementservercriteria("Name = '$proxyAgent'")
$server=$admin.getmanagementservers($crit)[0]

#Retrieve all our monitored network devices
$netdevices=$admin.getallremotelymanageddevices()

#Sets the proxy of all network devices to the specified proxy server
$coll= new-object "system.collections.generic.list``1[[Microsoft.EnterpriseManagement.Administration.RemotelyManagedDevice, Microsoft.EnterpriseManagement.OperationsManager, Version=6.0.4900.0, Culture=neutral, PublicKeyToken=31bf3856ad364e35]]"

for($i=0;$i -lt $netdevices.count;$i++){
  if($netdevices[$i].ipaddress -like "$deviceName"){$coll.add($netdevices[$i])}
}

$new_netdevices= new-object "System.Collections.ObjectModel.ReadOnlyCollection``1[[Microsoft.EnterpriseManagement.Administration.RemotelyManagedDevice, Microsoft.EnterpriseManagement.OperationsManager, Version=6.0.4900.0, Culture=neutral, PublicKeyToken=31bf3856ad364e35]]" (,$coll)

$admin.setproxyagent($new_netdevices,$server)

#-----------End Sample Script--------------

Conclusion

Hope you find this one useful. Please submit feedback as a comment on this post.

Powershell: Updating Proxy Agent for Multiple SNMP-enabled Devices (part 1)

Thu, 11 Jun 2009 23:17:38 GMT

As explained in part 3 of the SNMP in OpsMgr / Essentials series (still in progress), we explained that SNMP-enabled devices can be monitored from any management server OR agent-managed computer. The question we answer today is:

"How do I change the proxy agent for some or all of my network enabled devices?"

If I want to use an agent-managed computer as the proxy for multiple network devices, the script below will allow you to submit a wildcard filter based on device IP address to update the proxy agent for all SNMP-enabled devices matching the filter. Check back for the next installment in which we demonstrate how to specify a management server as the proxy agent for multiple devices.

INSTRUCTIONS FOR USE

Cut-and-paste the script below into Notepad, and save as updateproxyagent.ps1. Script should run from any computer with OpsMgr Console and Powershell installed.

SYNTAX

The script accepts the following 3 required parameters. Run with no parameters to view syntax help.

-rootMS: FQDN of the root management server for the target management group.

Ex: 'RootManagementServer.fqdn.local'

-proxyAgent: Agent that will serve as the proxy agent

Ex: ‘dc.contoso.com'

-deviceName: IP address or IP wildcard range.

Ex: ‘192.168.1*’

SAMPLE SCRIPT

Here is the sample script. You should be able to cut-and-paste this one with no worries about line wrap.

#-----------Begin Sample Script--------------

param($rootMS,$proxyAgent, $deviceName)

#Echo syntax if -rootMS parameter is NULL
     if ($rootMS -eq $null) {

        Write-Host ""
        Write-Host "-rootMS: FQDN of the root management server for the target management group.";
        Write-Host "Ex: RootManagementServer.fqdn.local";
        Write-Host ""
        Write-Host "-proxyAgent: Agent-managed computer that will serve as the proxy agent";
        Write-Host "Ex: ‘dc.contoso.com";
        Write-Host ""
        Write-Host "-deviceName: IP address or IP wildcard range.”;
        Write-Host "Ex: ‘192.168.1*’”;
        Write-Host ""
 exit;
     }

#connect to mgmt group
$ServerName=$rootMS
add-pssnapin  Microsoft.EnterpriseManagement.OperationsManager.Client ;
set-location  OperationsManagerMonitoring:: ;
new-managementGroupConnection -ConnectionString:$serverName;
set-location $ServerName;

#Retrieve all our monitored network devices
$netDevices = get-remotelymanageddevice | where-object {$_.Name -like $deviceName}

#Retrieve agent that will serve as proxy agent
$proxy = get-agent | Where-Object {$_.PrincipalName -like $proxyAgent }

#Sets the proxy of all network devices to the specified proxy server
Set-ProxyAgent -ProxyAgent $proxy -Device $netDevices

#-----------End Sample Script--------------

Conclusion

Hopefully some of you will find this one useful. Check back soon when we'll look at how to assign a management server as a proxy agent for multiple SNMP-enabled devices using Powershell. Please submit feedback as a comment on this post.

Powershell Tip: Finding the recipients on all enabled notification subscriptions in OpsMgr 2007

Thu, 11 Jun 2009 23:02:33 GMT

If you want to very quickly see the list of recipients configured to receive alerts for each of the enabled notification subscriptions in your environment, look no further than the get-notificationsubscription cmdlet in the Operations Manager 2007 Command Shell.

Here are a couple versions of a handy one-liner to dump this information to the console. I like the list version for easy reading on the screen.

Get-NotificationSubscription |where-object {$_.Enabled -eq $true}| Format-List DisplayName, ToRecipients OR

Get-NotificationSubscription |where-object {$_.Enabled -eq $true}| Format-Table DisplayName, ToRecipients

Windows Media Services (Server 2003 and 2008) Management Pack now available

Wed, 10 Jun 2009 09:01:46 GMT

Raphael Burri has written an OpsMgr 2007 management pack that discovers and monitors Windows Media Services 9.0 (Server 2003) and 9.5 (Server 2008). This is a native OpsMgr 2007 management pack.

It’s now available for download from the Community Management Pack Catalog at System Center Central.

Windows Media Services (Server 2003 and 2008)

Great work Raphael!

Troubleshooting Console Connectivity Problems in Essentials 2007

Tue, 09 Jun 2009 02:57:15 GMT

Good info from one of the group threads on troubleshooting SCE Console connectivity:

The SCE console actually attaches in three ways to the SCE server:

The first is SCE channel and that renders the main console data.
The second is the reporting data in the console which comes from an HTTP port 80 connection to the SQL 2005 Reporting Services component of SCE.
The third connection is to the WSUS 3.0 Server processes on the SCE server, using HTTP on port 8530 (HTTPS is on port 8531). To test this channel, browse to the following URL and verify it is available.https://SCEServer:8531/SimpleAuthWebService/simpleauth.asmx

This URL should render a SimpleAuth web page error-free. If you can't get to that web page you need to check out the IE/proxy/Windows firewall settings on your PC. If your PC produces a certificate error, it could mean that the "System Center Essentials All Computers Policy" (GPO) has not been applied to your PC. That GPO inludes the SCE Server certificate that is used on port 8531 of the WSUS portion on the SCE server. You need to apply that GPO to your Windows XP box or manually import the SCE certificates.

If your Windows XP box is already a managed computer of the SCE server that certificate should be already in the Trusted Root Certification Store on your XP box from the GPOs. Another route would be to manually install the WSUS 3.0 MMC (management MMC only) on your XP box and point the WSUS 3.0 MMC to the SCE server, see if you can attach to it.

Operations Manager 2007 R2 Management Pack Guide now available for download

Mon, 08 Jun 2009 20:54:18 GMT

System Center Operations Manager 2007 R2 includes an updated management pack and associated MP guide for monitoring Operations Manager 2007 R2.

The Operations Manager Management Pack helps you manage your System Center Operations Manager 2007 R2 infrastructure by monitoring the health of the Operations Manager components and services. The Operations Manager Management Pack alerts you to problems with components such as the management servers, the Operations Manager database, agents, modules used by workflows, and services so that you can continue to monitor the servers and clients that your business depends on. The management pack includes tasks that you can automate to get easy access to common diagnostic tools, such as restarting a health service or reloading an agent configuration.

This guide was written based on version 6.1.7221.0 of the Operations Manager Management Pack for Operations Manager 2007 R2.

Where to get it

You can download the guide from the System Center Pack Catalog with the link below

Operations Manager 2007 R2 Management Pack Guide

Source: System Center Pack Catalog

Powershell Tip: Retrieving a list of Virtual Machines from OpsMgr 2007

Mon, 08 Jun 2009 03:04:38 GMT

How about a PowerShell snippet to determine if a particular system is a virtual machine?

get-monitoringclass|where{$_.displayname -eq "Windows Server"}| `

get-monitoringobject|where{$_.displayname -eq "Oxford.contoso.com"}| `

get-monitoringobjectproperty|where{$_.displayname -eq "Virtual Machine"}|foreach{$_.value}

The only issue with this info out of the box is that OpsMgr does not recognize VMware guests as virtual machines. You can change this with the Virtual Machine Discovery MP. This MP extends existing discovery of virtual machines by Operations Manager and Essentials 2007 to include VMware guests.

Powershell Tip: Parsing the Operations Manager Event Log [sample script]

Mon, 08 Jun 2009 02:35:43 GMT

One of the nice features of OpsMgr 2007 is that it logs virtually everything to the Operations Manager Event Log. Things like script failures, latency in writing data to the Operational and Data Warehouse, script failures, etc. What's not always so easy is determining when and with what frequency these events are occurring on an ad-hoc basis.

This is the problem we will tackle in this post.

An Example from the Real-world

What is not so easy is evaluating this data to determine the frequency of these events, many of which are not raised as alerts in Operations Manager. Let's consider event 2115 in the Operations Manager Event Log, which is frequently an indicator of Operational database latency. If I see 2 or 3 of these an hour, it's not a big deal. However, if I see 20 or 30 of these an hour, then I almost certainly have an issue worth investigating.

What if I could easily parse my Operations Manager Event Log and get an hourly count of these events?

I've posted the script syntax below. However, if you run the script with no parameters, syntax will be echoed to the screen.

SYNTAX (Example):

EventsByHour.ps1 -Log 'Operations Manager' -EventID 2115 -Hours 10

Sample Script

Cut-and-paste the script below into Notepad and save as events.ps1

#---------Start Sample Script----------------

param($log,$id,$hours)

#Returning Syntax on null $log parameter
     if ($log -eq $null) {
        Write-Host ""
Write-Host "Syntax:";
        Write-Host ""
Write-Host "-Log: Name of the NT Event Log to query. `n Ex: 'Operations Manager'";
        Write-Host ""
Write-Host "-id: The Event ID to report on. `n Ex: 2115";
        Write-Host ""
Write-Host "-hours: The number of hours of log history to parse. `n Ex: 12";
        Write-Host ""
exit;
     }

$current=get-date

$events=get-eventlog $log | `
where {
($_.timegenerated -lt $current) -and ($_.timegenerated -gt $current.addhours(-$hours)) -and ($_.eventiD -eq $id)
}

for ($a = 0; $a -lt $hours; $a++) {
$begin=$current.addhours(-$hours+$a)
$end = $begin.AddHours(1)
$cnt=@($events | `
    where {
      ($_.timegenerated -lt $end) -and ($_.timegenerated -gt $begin) -and ($_.eventiD -eq $id)
    }).count
write-host "Event count for event between $begin and $end - $cnt"
}

#---------End Sample Script----------------

Output

The output is an hourly count of occurrences of this event, allowing me to make a quick judgement of the problem. In this example, I was looking for event 2115 to identify the frequency the event was being logged. In this case, my problem started about 7 hours ago, so I am going to look back about 8 hours.

PS C:\temp\Powershell> .\events.ps1 -log 'Operations Manager' -id 2115 -hours 8

And the output is a tidy list showing the number of occurrences over each of the last 8 hours. In this case, I can see by the event count that something must have come up between 12:30pm and 1:30pm

Event count for event between 11/19/2008 08:29:23 and 11/19/2008 09:29:23 - 30
Event count for event between 11/19/2008 09:29:23 and 11/19/2008 10:29:23 - 29
Event count for event between 11/19/2008 10:29:23 and 11/19/2008 11:29:23 - 30
Event count for event between 11/19/2008 11:29:23 and 11/19/2008 12:29:23 - 26
Event count for event between 11/19/2008 12:29:23 and 11/19/2008 13:29:23 - 24
Event count for event between 11/19/2008 13:29:23 and 11/19/2008 14:29:23 - 4
Event count for event between 11/19/2008 14:29:23 and 11/19/2008 15:29:23 - 5
Event count for event between 11/19/2008 15:29:23 and 11/19/2008 16:29:23 - 2

Hopefully you'll find this useful in one or more scenarios in your environment. Let me know if there are any examples / scenarios with OpsMgr where you'd like some Powershell assistance.

------

Kudos to Powershell MVP Marco Shaw for his pointers on formatting the count output a bit more nicely.

Community Day.be is back!

Sun, 07 Jun 2009 15:19:35 GMT

Community Day is back again, are you ready for our Third Edition of Microsoft Community Day?

Eleven Microsoft User Groups combine their efforts to organize this unique networking and knowledge sharing event.

A unique opportunity to learn about Microsoft’s latest developments and technologies like Exchange
2010, Silverlight 3, Visual Studio 2010, SQL Server 2008, Windows 2008R2, Powershell V2, Forefront
Identity Manager 2010, XNA 3.1. and OCS R2.

Microsoft Community Day will take place on Thursday 25th June 2009 in Utopolis, Mechelen, where we will
bring together 300 IT Pro’s and developers.

The Microsoft Community Day 2009 is supported by BESUG, BIWUG, IT-Talks, MVUG, Pro-Exchange, SCUG, SQLUG, VBIB, Visug, WinSec, XNA-BUG.
Register now for free for this all day event!

Agenda: http://www.communityday.be/cd/tabid/130/Default.aspx

Source: Alexandre Verkinderen, MVP-OpsMgr

MP Tuning Checklist: Windows 2008 Operating System MP (6.0.6321.5) for SCOM / SCE 2007

Fri, 05 Jun 2009 09:43:42 GMT

June 2009 - As with most MPs, there are few areas that may require some tuning, or at least deserve a second look. Here’s a quick checklist of items of interest in the Windows Base OS MP. If you use Boris MP Viewer, you’ll find you can spot items like these quickly and easily in most MPs before you deploy them to your environment.

Objective Discoveries

Name

Description

Discover Windows CPUs

By default, you’ll only get CPU utilization data for _Total of all CPUs. In cases where per CPU utilization is of interest, such as when using Processor Affinity in SQL 200x Server, you should consider enabling this discovery. Once enabled and the individual CPUs are discovered, the CPU monitoring targeting the “Windows Server 2008 Process” class

Discover Windows Physical Disks

This Discovery Rule discovers and populates the Windows Server class named “Windows Server 2008 Physical Disk” with instances of Physical Disks that are discovered on Windows 2008 Servers.

If monitoring physical disk I/O is of interest (as it often is in physical servers with local storage), you may consider enabling this discovery.

Once instances of the class are discovered, three unit monitors (for I/O monitoring) and 8 performance collection rules will begin monitoring discovered instances of the object class. Check the collection rules to be sure you actually want that performance data. If not, disable the rules for the unwanted counters.

Rules

There are a handful of alert-generating rules which are disabled by default. You may want to consider enabling depending on your monitoring priorities.

Name	Description
A Service or Driver Failed to Start A Service Terminated Unexpectedly	These two rules will alert you to unexpected failures of Windows services. While Microsoft MPs have monitors targeted to their services, these rules can catch 3rd party service failures that may otherwise not raise alerts.
A Software Update Installation Failed	This will alert you to failed Windows update installation. You’ll see these in the WSUS Admin Console if you’re watching. Notification may be a desirable for mission-critical servers or in high-security environments.
A Share Configuration in Invalid	This rule generates an alert when a file share experiences a configuration issue that prevents it from being made available on the network.This alert will occur when a file share points to a folder that has been deleted – could be handy on your corporate file servers

Monitors

The following monitors are disabled by default, and may be worth consideration.

Name

Description

Network Adapter Connection Health

This monitor generates an alert when Windows detects that the network adapter has been disconnected from the network and no longer has network connectivity.If the computer has only a single network adapter, this alert arrives only after network connectivity has been reestablished.

This monitor could help you determine if unexplained heartbeat failures are due to network connectivity issues by using this check from the host itself.

Average Disk Seconds Per Read

Average Disk Seconds Per Write

Average Disk Seconds Per Read

Average Disk Seconds Per Write

These repeated threshold performance monitoring rules can identify spikes in disks activity due to bursts of OS or application activity. Not something I enable as a general rule because they can be noisy, especially during nightly backups.

The following unit monitors are enabled, but have a GenerateAlerts value of false.

Name	Description
All monitors targeted to the Windows Server 2008 Processors class	If you enabled the object discovery for individual CPUs, you should be aware that all the monitors targeted to this class are not configured to generate alerts.

Settings for the following unit monitors should be be customized for the needs of your environment.

Name

Description

Logical Disk Free Space

This monitor needs to be adjusted from the default in many environments.

The Logical Disk Free Space monitoring routine is a high configurable solution that enables Operators to set varying threshold values for system and non-system logical disk volumes. Errors are generated only when BOTH MB and Percent thresholds breach their configured thresholds

System Drive Free Space Thresholds (Defaults)

Non-System Drive Free Space Thresholds (Defaults)

Conclusion

What changes do you make to the default settings of the Windows OS MP? I’d be interested to hear your feedback. Leave your thoughts as a comment on this post.

Changes effective July 1st to System Center Server Management Suite Enterprise Licensing

Thu, 04 Jun 2009 23:20:52 GMT

Two System Center Server Management Suite licensing changes will become effective 1 July, 2009, with the release of Microsoft System Center Operations Manager 2007 R2*:

System Center Server Management Suite Enterprise will be switched from an unlimited operating system environment to a four operating system environment, limited licence with a corresponding 20 per cent price decrease.
A new suite offering, Microsoft System Center Server Management Suite Datacenter will be introduced and will include the same products as System Center Server Management Suite Enterprise, but, it will be licensed per processor and will provide for the management of an unlimited number of operating system environments.

(continue at source)

Source: Lady Licensing

OpsMgr: How to verify SNMP support in your network devices

Wed, 03 Jun 2009 23:31:29 GMT

Here are some tools, tips and resources to assist when SNMP device discovery or monitoring is not going as expected in Operations Manager (SCOM) and Essentials 2007.

Question: Discovery of my SNMP devices fails. How can I verify my device supports SNMP, and test device response before using in Operations Manager 2007 or Essentials 2007?

Answer: There are a number of resources you can use to determine what version of SNMP your device supports. You'll need device IP and SNMP community string.

Checking SNMP version support

You can verify verify a device supports SNMP version 2c with the open source snmpget utility, which is part of the net-snmp tools. Download snmpget.exe and the accompanying mibs.zip file from the following URL and unzip to a local directory. Download HERE.

If you are not successful discovering your target network device, verify your device supports SNMP version 1 or 2c. To check, you'll need the SNMP community string for your device, and the IP address with that of your target. Remember to replace IP and community string in the command below with those applicable to your environment.

snmpget -v2c -M D:\utilities\mibs -c public 192.168.240.250 1.3.6.1.2.1.1.1.0

Assuming the device has the required SNMP support, you should see output similar to the following:

SNMPv2-MIB::sysDescr.0 = STRING: 24-Port 10/100/1000 Gigabit Switch w/WebView

Note that the output will very by the device you are querying, change -v2c to -v1. If you still get no response, verify the criteria above and check with the device manufacturer to ensure your network device has the required SNMP support.

Reading the MIB

I get the mib from the hardware vendor. I then use MIB Browser 1.0 to view the OID tree. MIB Browser can be found HERE.

Another good MIB browser is the iReasoning browser, which is available HERE.

If the vendor cannot give me a mib, I check the MIB database at oidview.com to find the MIB for my device.

http://www.oidview.com/mibs/detail.html

Polling a device for expected response

I then use iReasoning browser to perform a get on the OID to make sure it returns values as expected. I occasionally also use the snmpwalk utilities, which is part of the same tool set as snmpget.

Links: OpsMgr 2007 Scripting Articles at System Center Central

Wed, 03 Jun 2009 10:17:33 GMT

There are a few articles and a lot of sample scripts we have written over the last couple of years since OpsMgr 2007 was released. By request, here is a list of many of the instructional articles along with their related sample script downloads ( in one place for reference):

Scripting-related Articles and Posts

OpsMgr: Writing and debugging OpsMgr 2007 scripts outside of a live Mgmt Group

SCF Scripting Series Part 1 - Converting MOM 2005 Scripts to OpsMgr / SCE 2007

SCF Scripting Series Mgmt Pack - Part 2

Scripting Series (Part 3): Debugging Runtime Scripts for OpsMgr and SCE

Using the Property Bag with Custom Scripting in Operations Manager 2007 <—This is the big article on the PropertyBag and custom Performance data'

How to create a Script-based 2-State Monitor <—This pdf illustrates how to get your OpsMgr scripts into a monitor

Scripting Tip: Valid Event ID ranges when using the MOMScriptAPI.LogScriptEvent method

Sample Downloads (to go with the articles above)

Remember to login to get to the Downloads (registration and download is free)

SCF Scripting Series Mgmt Pack - Part 1

SCF Scripting Series Mgmt Pack - Part 2

Script: Registry Value Check for use in a 2-state monitor in OpsMgr 2007

Script: Windows Service Configuration Script (Service Startup) for 2-State Monitor for OpsMgr / Essentials 2007

Script: File Version Check Script for 2-State Monitor

Script: Windows Process Monitor for use in a 2-state monitor in OpsMgr 2007

Script: File Count Monitoring for OpsMgr 2007

Scripting Tip: Valid Event ID ranges when using the MOMScriptAPI.LogScriptEvent method

Wed, 03 Jun 2009 09:52:13 GMT

I was working with a custom runtime script the other day and received an ambiguous error message when I executed the script on my development workstation.

Error: Microsoft VBScript runtime error: Subscript out of range

Issue: The Event ID I was using was actually an invalid event number!

Workaround: The .LogScriptEvent method only accepts event IDs from 1 to 20000. 20001 and higher result in the error above.

Additional Reading

Incidentally, if you're developing custom runtime scripts for Operations? Manager and Essentials 2007 check out our Scripting Series, which includes several articles on writing monitoring scripts for OpsMgr. (I still need to put a discovery scripting sample together). Search the site for “Scripting Series” which should reveal all the posts and related downloads.

Job Aid: Is System Center Essentials 2007 right for my organization? (Self Survey)

Mon, 01 Jun 2009 21:56:04 GMT

This is a question many run into when comparing SCE, SCCM and SCOM 2007, and it deserves a straight answer. Below are some questions and considerations to help your organization honestly assess your circumstances and choose the right path for IT operations and systems management for your organization.

Essentials 2007 (aka SCE) is an exciting product in that it draws together many features of multiple MS enterprise systems and operations management platform in a package designed (and priced)for the mid-market. But let's face it, IT pros are busy in the enterprise too - that a one stop shop? concept, delivering all aspects of day-to-day management in a unified console is pretty attractive. I get that, and I feel your pain - that's my life too. But the fact is SCE was designed for the mid-market IT generalist - the guy without the time, the staff, the budget and the specialized skills of the IT shop in the large enterprise and without some of the more advanced management requirements and environmental complexity seen in that space.

Essentials delivers a lot of functionality for a single product, and I see some enterprises out their trying to fit a square peg into a round hole, largely because the product is so convenient and easy to use, delivering so much functionality:

Monitoring (including network device monitoring for FREE!)
Update Management
Hardware / Software Inventory
Software Deployment
Reporting (capacity, performance, availability, asset tracking) and
Packed in a way that makes setup easy for the less-than-guru class of administrator, all in a single console

But let's get to the bottom line. Is Essentials 2007 right for your organization?

Let's find out. Ask yourself the following questions to make a 'self assessment' of the appropriate path for your organization. Below are the relevant questions, broken down into some rough categories.

If you can honestly answer “no” to all the following, then you organization is a good candidate for Essentials 2007 SP1. Get that Essentials deployment rolling!

If you answer ”yes” to one or more of the following, you should consider taking the enterprise path for operations and systems management. That means SCOM, SCCM and WSUS 3.0

Update Management

Do you currently have or require a distributed WSUS deployment?

If the answer to either of these is yes, then you'll need to opt for the stand-alone WSUS 3.0 platform. Essentials does not support or participate in a distributed WSUS topology as an upstream / downstream server.

Software Deployment

Do you have advanced deployment needs, such as Vista OS deployment?

If your answer is yes, then Configuration Manager (SMS) is the route for you. Essentials is great for application deployments such as Office 2007, but lacks the robust feature sets of the enterprise platform when it comes to OS deployment.

Configuration Management

Do you have a need for monitoring / auditing of specific desktop or server configuration settings?

If so, you should opt for the enterprise Configuration Manager (SMS) platform, which contains advanced functionality for monitoring of desired configuration items (Desired Configuration Manager). Essentials has great base inventory reporting, but is not customizable for monitoring a user-defined attribute set.

Software / Hardware Inventory

Do you have the need to customize your software and hardware inventory?

SCE reports on approximately 60 aspects of software and hardware in managed systems, but cannot be customized. If you have substantial need to customize, chances are you need to go the enterprise route of Configuration Manager.

Capacity / Complexity

Do you have more than 30 servers / 500 clients?

Those are the supported limits. If you're at these limits, or think you will be approaching them in the next 18-24 months, you should probably opt for the enterprise platforms SCOM, SCCM, and WSUS.

Do you require multiple management groups within a single domain, such as for autonomous IT units in delegation scenarios (resource OUs)?

Only 1 Essentials Server is allowed per domain. If you require multiple management groups, then you are a candidate for Ops Mgr 2007.

Do you have multiple branch offices with substantial client and server counts?

Gateway scenarios, tiering and multiple management server deployments are not available with SCE. If you have need to distribute workloads to conserve bandwidth, then the enterprise route will be the path for you.

Untrusted domains?

Essentials does certificate-based authorization scenarios of the enterprise Operations Manager platform. However, you cannot monitor untrusted forests. Such scenarios will necessitate going the enterprise route.

Security

Do you require role-based security for IT specialists in specific support roles?

SCE doesn't support the robust role-based options you see in SCOM. If you need to drill down on role-based security, SCOM is the tool for the job.

Conclusion

That's it! Count up your answers. If you have even a single yes in here, weigh your options carefully. You can make some concessions, but don't back yourself into a corner to save a buck. Essentials is a awesome tool, but it's not right for every environment. Choose wisely. Questions? Comments? Leave your comments here or ping me through the contact page on this site.

TIP: How to reset and re-run the Feature Configuration Wizard in Essentials 2007 SP1

Mon, 01 Jun 2009 03:56:25 GMT

Here is a question posed recently that may be useful to administrators or consultants working with System Center Essentials 2007 SP1.

Question: Once I've run the Feature Configuration Wizard in Essentials and selected Local Policy, how can I reverse my decision and switch to Active Directory Group Policies applied at the domain level? The policy type can be selected only when the Feature Configuration wizard is run for the first time, so this poses a bit of a problem. Once you've made the decision, it's done. After this, the selection will be unavailable next time the wizard is run. However, there is a workaround:

Answer: This can be changed by running SCECertPolicyConfigUtil.exe /uninstall /ManagementGroup on the Essentials 2007 server.

After you run this command, check the following:

That the "SCE All Computers" and "SCE Managed Computers" GPOs were deleted from AD
That the "SCE Managed Computers" security group was deleted from AD
That the following registry key is set to 0:

HKLM\SOFTWARE\Microsoft\System Center Essentials\1.0\Console\FeatureWizardSettings\FeatureConfigurationCompleted

Certificates were deleted from the \Certificates folder.

After running the command and checking that all the above are true, open SCE console, re-run the "Feature Configuration Wizard".

R2 FAQ: Can an agent report to an OpsMgr 2007 SP1 and R2 systems at the same time?

Sat, 30 May 2009 14:33:38 GMT

Question: Can a SCOM agent report to both Operations Manager 2007 SP1 and R2 management groups at the same time?

This is a question we’ve seen a few times already, and may be issue if you want to run Operations Manager 2007 SP1 & R2 side-by-side for POC or multiple management group scenarios.

Answer: The answer is yes, but it requires using the R2 version of the agent on the agent-managed systems. The R2 version of agent can communicate to SP1 management groups as well. To do this, you will need to discover the SP1 agent-managed computer from the R2 management group to install the R2 version of the agent which will allow it to communicate with both a R2 and a SP1 management group.

When you are ready, you can then uninstall the agents from the SP1 management group, which remove the settings for the SP1 management group from the registry, leaving the agent reporting to the R2 group only.

Just a point of interest: While the service display names were changed in SCOM 2007 R2, the service names themselves have not changed. They are still omsdk, omcfg and healthservice.

There is a upgrade guide on the Microsoft site at http://technet.microsoft.com/en-us/opsmgr/bb498235.aspx

Part 5: Exploring Discovery Data with the OpsMgr Command Shell - Virtual Machine Inventory

Fri, 29 May 2009 23:32:51 GMT

Previous Installments

What we illustrated in part 4 with Dell Server inventory, we can easily explore and present the properties of any object class using the OpsMgr Command Shell. In this case, we want to retrieve a list of all computers for which the IsVirtualMachine property equals true.

But the IsVirtualMachine property only reveals VMs running on Microsoft virtualization platforms! You can actually solve this problem using our Virtual Machine Discovery MP for Operations Manager 2007. This MP will update the default discovery to identify both Microsoft and VMware VMs.

But I digress…back to our Virtual Machine Inventory report.

I started by finding the name of the Windows Computer object class by running using the Get-MonitoringClass cmdlet like so:

get-monitoringclass | where {$_.DisplayName -match 'Windows Computer'} | select Name, DisplayName

This returned a list that made it easy to pick out the class name for Windows Computer, which is Microsoft.Windows.Computer.

To create a computer inventory, we will first retrieve at least one instance of the object class so we can enumerate all the properties of the class. We will assign it to a variable ($computer) for easy reuse, as shown below:

$computer = Get-MonitoringClass | where {$_.Name -eq 'Microsoft.Windows.Computer'} | get-monitoringobject

Now, let's return all the properties of the class. That's another simple one-liner

$computer | get-member

As with the Dell inventory in part 4, you can also retrieve just the Noteproperty membertype. This will be important, as it takes a slight change in how we retrieve those properties. So, retrieve these so you know which they are.

$computer | get-member -Membertype Noteproperty

Figure 1 - Portion of output of $computer | get-member -Membertype Noteproperty

PathName

UniquePathName

[Microsoft.SystemCenter.ManagedComputer].InstallDirectory

[Microsoft.Windows.Computer].ActiveDirectoryObjectSid

[Microsoft.Windows.Computer].ActiveDirectorySite

[Microsoft.Windows.Computer].DNSName

[Microsoft.Windows.Computer].DomainDnsName

[Microsoft.Windows.Computer].ForestDnsName

[Microsoft.Windows.Computer].IPAddress

[Microsoft.Windows.Computer].IsVirtualMachine

[Microsoft.Windows.Computer].NetbiosComputerName

[Microsoft.Windows.Computer].NetbiosDomainName

[Microsoft.Windows.Computer].NetworkName

[Microsoft.Windows.Computer].OrganizationalUnit

[Microsoft.Windows.Computer].PrincipalName

[Microsoft.Windows.Server.Computer].IsVirtualNode

[System.Entity].DisplayName

[System.Mom.BackwardCompatibility.Computer].Bridgehead_Server

[System.Mom.BackwardCompatibility.Computer].BusinessCriticality

[System.Mom.BackwardCompatibility.Computer].ComputerType

[System.Mom.BackwardCompatibility.Computer].Computer_Manufacturer

[System.Mom.BackwardCompatibility.Computer].Computer_Model

[System.Mom.BackwardCompatibility.Computer].DomainControllerService

[System.Mom.BackwardCompatibility.Computer].ExchangeService

[System.Mom.BackwardCompatibility.Computer].Front_end_Server

Retrieving Data for the Report

At this point, I wanted to simply retrieve the Name, IsVirtualMachine, for each server. Again, the NoteProperty type is covered in Part 4 – Dell Server Inventory Report. So, with a quick modification to replace the bracketed string with a wildcard, the output came as desired.

$computer|select *.NetbiosComputerName, *.IsVirtualMachine

You can add properties from above as you like and, as illustrated in previous posts, you can export to an Excel-friendly csv file.

Part 4: Exploring Discovery Data with the OpsMgr Command Shell - Dell Server Inventory Report

Fri, 29 May 2009 22:53:01 GMT

Previous Installments

I was recently in need of a server inventory of Dell Servers in a customer environment. More importantly, I wanted an easily repeatable method for my customer to run their own Dell Server inventory report with a minimum of effort. After looking briefly at available reports, it was quickly apparent that the Command Shell was going to be the right tool for the job. While I am illustrating a report for Dell hardware, this should be achievable for IBM and HP hardware with a bit of exploration of the hardware classes provided through those management packs.

Finding the Right Object Class

What I really wanted were the Dell-specific properties. like Service Tag, Model, OMSA Version, etc. I had imported the latest version of the Dell MP (3a.01 at the time of this writing), so the first step was to look at my available Dell object classes. This is easily done using the get-monitoringclass cmdlet with a where clause to retrieve a list of classes containing the word 'Dell'.

get-monitoringclass | where {$_.Name -like 'Dell.*' } | ft Name, DisplayName, Description –auto

From this, I found the class representing Dell Servers, called Dell.Connections.DellServer (DiplayName: "Dell Server Instances"). The display name and description made it immediately apparent this was the class that represented Dell Servers.

Retrieving Available Class Properties

Next, I wanted to find out what properties of Dell Servers are discovered for my inventory report. So, I retrieved the instances of this object class by piping the output to the get-monitoringobject cmdlet, and assigned it to the $server variable for later use, as shown below

$server = get-monitoringclass | where {$_.Name -eq 'Dell.Connections.DellServer' } | get-monitoringobject

Then, I enumerated the properties of the Dell Server Instance class using get-member cmdlet like so

$server | get-member

In the very long list of output, I found that most of the properties I was interested in were of the NoteProperty type. NoteProperty is a special type added by Powershell's Extended Type System (ETS) and not actually defined in the OpsMgr SDK, as described by Marco in a previous guest post. This caused some difficulties in creating our output as you'll see momentarily. So, the Dell-specific properties of the NoteProperty type were retrieved with the following one-liner.

$server | Get-Member -MemberType noteProperty

Figure 1 - Portion of output of $server | get-member -Membertype NoteProperty

Retrieving Data for the Report

At this point, I wanted to simply retrieve the IPAddress, Model, OMSA version and Service Tag for each server. However, I found when I tried to retrieve any member of the NoteProperty type by name, like so:

$server|select [Dell.Connections.DellServer].ServiceTag

...my output was either null or resorted in an error. (I seem to have this problem whenever Noteproperty and square brackets are involved). So I needed a workaround. Conveniently, this was ground I found Marco had covered in his guest post as well. So, with a quick modification to replace the bracketed string with a wildcard, the output came as desired.

$server|select Name, *.BiosVersion,*.ServiceTag, *.AssetTag, *.OMSAVersion, *.Model

Now, we can send the output to a spreadsheet with the export-csv cmdlet, taking care to pass the -noTypeInformation flag to remove the type output at the top of the report. Here's your one-liner to retrieve Dell Server inventory:

$server|select Name, *.BiosVersion,*.ServiceTag, *.AssetTag, *.OMSAVersion, *.Model | export-csv -path 'c:\DellServers.csv' –noTypeInformation

Part 3: Exploring discovered inventory with the Command Shell - Groups

Fri, 29 May 2009 10:32:56 GMT

In the previous two posts on the topic, we illustrated first how to export a list of object classes and their descriptions to a spreadsheet for reference, and then we did some exploration into the discovered instances of an existing class. In this post, a quick note on how to enumerate the membership of a group in the OpsMgr Command Shell.

Previous Installments

Part 1 - Powershell Tip: Retrieving an inventory of object classes and descriptions from OpsMgr 2007
Part 2 - OpsMgr 2007: Exploring your discovered inventory with the Command Shell

However, you may have noticed that there were groups listed in your spreadsheet of class names and descriptions. And you may have wondered, why is that?

You may be surprised to learn that each group actually represents a unique object class. However, groups represent a special type of class known as singleton class. Singleton classes are used when there is only one instance of a class (as is usually the case with groups). The class is the instance and always exists. For example, there will only ever be one instance of the group called Exchange 2003 Servers in your management group, so it is a singleton class. Because there is only one instance of this class, you never need to write discovery for the class. The single instance of the class is automatically created.

But I digress. What I wanted to illustrate that you can view the membership of a group through the command shell quite easily. For example, to view the members of the ?SQL 2005 Computers? group, perform the following steps:

Launch the Operations Manager Command Shell.
Retrieve the Name property of the "SQL 2005 Computers" group from your inventory spreadsheet we created in part 2 (FYI - the name is Microsoft.SQLServer.2005.ComputerGroup)
Execute the following commands from the Command Shell prompt

cd Microsoft.SQLServer.2005.ComputerGroup

This actually changes our location to the group, just as though it is a folder. Now, to view group members, you can simply enter the following:

Get-childitem

And of course the same formatting and filtering cmdlets we illustrated previously are available, so you can retrieve only the fields you wish in a format that suits your purpose. So a simple example of a prettier version might be

get-childitem | format-table DisplayName

For example, this could be great for retrieving a list of Exchange databases and the server on which they are hosted.

That’s all for now. Next, we'll look at a simple server inventory report from data collected by OpsMgr

Part 2: Exploring your discovered inventory with the OpsMgr Command Shell

Fri, 29 May 2009 02:32:39 GMT

As illustrated in part 1 of the series, you can see that it's quite easy to enumerate the object classes present in your Operations Manager 2007 environment. You may be surprised to learn it's equally simple to take this example one step further to enumerate the discovered instances of a given object class.

Using a similar one-liner to the previous tip, we can use get-monitoringclass cmdlet to retrieve a list of the discovered object classes, like so:

get-monitoringclass | select-object DisplayName, Name, Description | export-csv -path c:\classes.csv

We can then find the Name column in this output (not to be confused with DisplayName) to retrieve a list of discovered instances of the object class of our choice by passing output of the get-monitoringclass cmdlet for a single class to the get-monitoringobject cmdlet, as illustrated for the 'SQL 2005 DB' object class below:

get-monitoringclass -name "Microsoft.SQLServer.2005.Database" | get-monitoringobject

And using the format-table cmdlet demonstrated in several past examples, we can filter and format results in a neatly organized tabular format, so display only the database DisplayName (e.g. OperationsManager) and PathName, which represents the server and SQL instance that hosts the database.

get-monitoringclass -name "Microsoft.SQLServer.2005.Database" | get-monitoringobject | ft DisplayName, PathName

The resulting output of the PathName is a bit messy, but you can spot the host server and SQL instance name (MSSQLSERVER is the default SQL instance). DisplayName/PathName/msdb/OPSMGR.contoso.com/MSSQLSERVER

In the next installment, we'll take a short look at a special object class...groups

Part 1: Retrieving an inventory of object classes and descriptions from OpsMgr Command Shell

Fri, 29 May 2009 02:30:37 GMT

Have you ever wondered what the object classes represent when you're trying to scope your console view or target a rule?or monitor? It? can sometimes be confusing because the description is not included in all areas of the Operations Console.

Here's a short one-liner to retrieve a list of all object classes present in your Operations Manager 2007 management group, including a description of the class:

Get-MonitoringClass | ft DisplayName, Description

Remember, you can send this to a csv file for easy reading in Excel as well.

Get-MonitoringClass | select-object DisplayName, Description | export-csv -path c:\classes.csv

This is really just the beginning. Once you’ve got the class names in hand, we can begin to do some exploration of discovered inventory, which we'll begin in part 2.

OpsMgr and SCE: 4 Tips for Proactive Management Pack Tuning

Thu, 28 May 2009 04:28:23 GMT

Most of us think of management pack tuning as a reactive process in response to alerts raised in our production environment. Management pack tuning is certainly an iterative process that we as Operations Manager and Essentials administrators must engage in every day. But the fact is, there are several proactive steps you can take to ensure a minimum of alert noise and unnecessary impact to your infrastructure and applications.

1. Create a Management Pack Report

A management pack report provides instant visibility into the contents of a management pack, rules, monitors, discoveries, tasks, object classes...everything you see in the Monitoring and Authoring workspaces. This is a great way to quickly check out the types of monitoring present, default performance thresholds, etc. This can bring to light monitoring thresholds you know will need tuning immediately. There are two tools available for creating a Management Pack Report:

MP Viewer - by Boris Yanuspolsky, which provides export to HTML and Excel options. I prefer the Excel option as it provides the ability to sort report contents. This tool also exposes rule type and performance threshold data, so it's a great way to get a look at MP internals quickly.

You can download MP Viewer HERE.

Silect MP Studio Lite - which provides a "document this management pack" feature, providing sortable columns without the need to export to XML, along with some additional key info for Essentials admins as you'll see in a second.

You can obtain a copy of MP Studio Lite HERE.

The Manual Method - In the Authoring space of the Operations or Essentials Consoles, you can look at the properties of rules, monitors or object discoveries, but it is a much slower process to learn the details in this way.

TIP: You'll find the columns exposed in MP Studio Lite vary a bit from those of MP Viewer. I keep both loaded on my workstation for ad-hoc MP analysis.

2. Look at what is Enabled and DISABLED by default

As important as understanding what will be monitored by a given management pack is understanding what will NOT be monitored by default. What you will find varies with every management pack, so be sure to check each time you prepare to deploy a new MP. For example, in the Windows Base OS management packs, you will find that the rules that provide global monitoring of all Windows Services are disabled by default. If you deploy to production and do not have 3rd party management packs for all your applications, you may not be notified of unexpected service stoppages for your important 3rd party applications. For more information on where to find and how to enable these rules, see the "Global Windows Service Monitoring" of this blog post. The methodology described within is applicable to both Operations Manager and Essentials 2007.

So Is it REALLY Enabled? (Category Overrides)

For Essentials 2007 users, this can be a difficult question to answer outside the Console UI. The MP Viewer utility provides accurate reporting of the enabled / disabled status of rules for Operations Manager 2007 only. For Essentials users, you will need to use Silect MP Studio Lite for accurate reporting. This is because Essentials handles category overrides differently than Operations Manager 2007. Here's a quick explanation. In the management pack XML, you will see there are 5 category overrides that apply to all categories of rules and monitors. The Enabled= attribute will contain one of these values"

True
OnEssentialMonitoring
OnStandardMonitoring
OnAdvancedMonitoring
False

In Operations Manager 2007- Everything except those marked as Enabled=FALSE will be ENABLED. In Essentials 2007 - Everything marked as TRUE or OnEssentialMonitoring will be ENABLED. Everything marked as OnStandardMonitoring, OnAdvancedMonitoring and False will have an effective Enabled value of FALSE. NOTE: You can find this data in the Operations and Essentials Consoles by looking at the "Enabled by Default" column for rules, monitors and discoveries in the Authoring space.

3. Read the Management Pack Guide BEFORE You Import!

Included in very management pack Microsoft releases is an MP guide that describes any required configuration steps in detail, as well as release notes with any known issues. You can find an online copy of the many Microsoft Management Pack Guides HERE.

4. Check for Bug Reports

Check the Microsoft newsgroups, Operations Manager and Essentials blogs and the MS Connect Site for any user reports of MP bugs that may affect your environment and/or success in monitoring for specific conditions. The Microsoft Newsgroups in particular are full of this type of information.

Conclusion

Hopefully I've helped illustrate the value of a little up front effort to ensure good results for all your management pack deployments. A few minutes of your time can save a few hours of effort down the road!

Powershell Tip: Formatting trick to make your search criteria stand out

Thu, 28 May 2009 04:23:50 GMT

As I mentioned in a post earlier this week, I really like the many formatting options in Powershell. Here's another formatting tip not well-documented in my experience. Let's take this example in which we are looking at whether agent proxy is enabled on my Operations Manager 2007 agents. I want to print all agents and their agent proxy status to the screen, but I want to be able to quickly identify agents where agent proxy is NOT enabled.

This is quite easily done with a foreach loop and the Powershell equivalent of the if-then statement to change the color of the text based on the value of the ProxyEnabled status for each agent.

In this case, if ProxyEnabled = false, text output will be RED.

If ProxyEnabled = true, text output will be GREEN.

I have commented the sample script below to clarify the meaning of each statement.

#-----------------Begin Sample Script-----------------------------

#Retrieve all agents and assign to a variable

$agents = get-agent

#For each agent in the $agents variable, execute the IF statement

foreach ($agent in $agents)

{

#If ProxyEnabled = TRUE, write agent Name and ProxyEnabled status to the console in RED.

If ($agent.ProxyingEnabled -match $true) {write-host $agent.name, $agent.proxyingenabled -foregroundcolor "red"}

else

#If ProxyEnabled = FALSE write agent Name and ProxyEnabled status to the console in GREEN.

{

write-host $agent.name, $agent.proxyingenabled -foregroundcolor "darkgreen"}

}

#-----------------End Sample Script-----------------------------

Have something you'd like to do with the Command Shell but can't quite figure out how? Drop me a line via a comment on this post.

Pete Zerger

Coming Soon: Configuration Manager 2007 Unleashed (with coverage of R2! )

OpsMgr 2007 Web Console Series Part 3: Publishing the Web Console through ISA 2004

OpsMgr 2007 Web Console Series Part 2: Branding the Web Console in Operations Manager 2007 R2

Part 1: Secure Publishing of the OpsMgr 2007 Web Console to the Internet

Securing Credentials Traversing the Internet

Instant Software Deployment in Essentials 2007 - When wuauclt /detectnow just is not fast enough

Scenario:

Workaround:

The Uninstall Task

The Install Task

SCOM 2007 R2: How to test Cross-Platform Monitoring without purchasing a RHEL license

How Similar are Red Hat Enterprise Linux (RHEL) 5 and CentOS?

Discovering and Monitoring CentOS

Changing the OS Caption on a CentOS Server.

Results

OpsMgr 2007 Health Check – Finding rules and monitors targeted to groups

SCOM: Report Models for Operations Manager 2007 SP1 (and how to install and use)

To Install These Models in SQL Reporting Services for use with SCOM 2007 SP1

To Use the Models

Powershell Tip: Working with TRUE and FALSE Comparisons in Powershell

Notes from “The Architect”

Sample Script

SCOM Console Task: Finding currently logged on user from the Operations Console

SCOM: Dynamic Group to detect SQL 2005 with (or without) specific Service Pack

How to convert management packs for OpsMgr 2007 (without a live MOM 2005 management group)

ResearchThis! KB - The Internet Information Service NNTP Virtual Server named NNTPSVC/1 is unavailable as the virtual server has been stopped

ResearchThis! KB - Auto Close Flag

ResearchThis! KB - Auto Shrink Flag

Windows Server Hyper-V MP for SCOM 2007 now available (ver 6.0.6633.0)

Overview

Feature Summary

Where to get it:

Command Shell: Dumping all overrides in Operations Manager 2007

Command Shell: Retrieving Agent Count by Management Server

OpsMgr 2007: Updating Primary and Failover Mgmt Server settings for multiple agents

OpsMgr 2007 (SCOM) v10 Wish List Wrap-up

Provision/Deprovision entire Virtual Lab Environments with Visual Studio Lab Management 2010 [beta]

Tuning Items for New Essentials 2007 SP1 Installations

Monitoring Database Blocking Through Rules and Alerts [Article and MP]

Beefing up monitoring of blocking processes

Implementation and Targeting

Where to get it

Related Articles

MS Exchange Server 2007 MP for OpsMgr (SCOM) 2007 R2 now available

Feature Summary

ResearchThis! KB - Core Service File Writing

Cross Platform Management Pack Authoring Guide for OpsMgr 2007 R2 (SCOM) Now Available

Reference

Powershell: Updating Proxy Agent for Multiple SNMP-enabled Devices (part 2)

INSTRUCTIONS FOR USE

SYNTAX

SAMPLE SCRIPT

Conclusion

Powershell: Updating Proxy Agent for Multiple SNMP-enabled Devices (part 1)

INSTRUCTIONS FOR USE

SYNTAX

SAMPLE SCRIPT

Conclusion

Powershell Tip: Finding the recipients on all enabled notification subscriptions in OpsMgr 2007

Windows Media Services (Server 2003 and 2008) Management Pack now available

Troubleshooting Console Connectivity Problems in Essentials 2007

Operations Manager 2007 R2 Management Pack Guide now available for download

Where to get it

Powershell Tip: Retrieving a list of Virtual Machines from OpsMgr 2007

Powershell Tip: Parsing the Operations Manager Event Log [sample script]

An Example from the Real-world

SYNTAX (Example):

Sample Script

Output

------

Community Day.be is back!

MP Tuning Checklist: Windows 2008 Operating System MP (6.0.6321.5) for SCOM / SCE 2007

Objective Discoveries

Rules

Monitors

Conclusion

Changes effective July 1st to System Center Server Management Suite Enterprise Licensing

OpsMgr: How to verify SNMP support in your network devices

Checking SNMP version support