Category: OpsMgr

Infrastructure Planning & Design Guide For OpsMgr 2012

Microsoft has released an IPD for System Center 2012 – Operations Manager.

This guide outlines the infrastructure design elements that are crucial to a successful implementation of Operations Manager. It guides you through the process of designing components, layout, and connectivity in a logical, sequential order. You’ll find easy-to-follow steps on identification and design of the required management groups, helping you to optimize the management infrastructure.

Infrastructure Planning and Design streamlines the planning process by:

  • Defining the technical decision flow through the planning process.
  • Listing the decisions to be made and the commonly available options and considerations.
  • Relating the decisions and options to the business in terms of cost, complexity, and other characteristics.
  • Framing decisions in terms of additional questions to the business to ensure a comprehensive alignment with the appropriate business landscape.

System Center 2012 Visio Management Pack Designer for System Center Operations Manager

Want to design your own simple management packs for SCOM (OpsMgr) from scratch but, like me, found the authoring kit to be like a mythical Greek maze filled with monsters?  Well I have great news … the Visio Management Pack Designer (VMPD) is finally here!!!!

I blogged about this tool at MMS earlier this year.  You drag and drop what you want done, and it’ll do all the hard work for you.  It’ll be a great addition to any OpsMgr admin/consultant toolkit.

The System Center 2012 Visio MP Designer—VMPD—is an add-in for Visio 2010 Premium that allows you to visually design a System Center Operations Manager Core Monitoring Management Pack. VMPD generates a Management Pack that is compliant to the MP Authoring Best Practices by simply dragging and dropping Visio shapes to describe your application architecture and configuring them through the Shape Properties.

– Visually describe your application architecture which generates classes and discoveries.

– Visually add monitoring to create monitors and rules.

– Start quickly from our pre-canned single server and multi-server patterns.

What Impresses Me Most About the Veeam nworks Management Pack for System Center …

… is the sheer amount of information that it provides.  I previously talked about the monitoring.  That’s great for the reactive side of things.  When I managed infrastructures, I like to take some some to see who things were trending so I could plan.  That’s where reports come in handy, and there’s no shortage of those in this management pack:

image

On my client’s site, we had an alert about latency on a HBA in one of the hosts.  I wanted to give the client some useful information to plan VM placement using affinity rules to avoid this from happening again.  One of the cool reports allows you to create a top-bottom chart of VMs based on a specific performance metric.  The below report was created with with the VMGUEST IOPS metric and shows the top 25 disk activity VMs.

image

As usual with OpsMgr, the report could be scheduled for a time period, and/or saved as a web archive, PDF, word file, etc.  I like this management pack.  Sure, it is pricey (I was told over EUR400/host socket being monitored), but it’s good.  BTW, Veeam did release a 10 socket (enough for 5 hosts with 2 CPUs each) management pack for free, which is available to you under two conditions:

  1. Be a new customer to Veeam AND
    2. Be a SCOM 2012 customer (not SCOM 2007)

Yesterday’s Fun In OpsMgr: Failed to store data in the Data Warehouse

Actually, the full error in the alert in this System Center Operations Manager 2007 R2 install was:

Failed to store data in the Data Warehouse.Failed to store data in the Data Warehouse. Cannot resolve the collation conflict between "SQL_Latin1_General_CP1_CI_AS" and "Latin1_General_CI_AS" in the equal to operation.

A bit of quick checking and I found that the SQL server instance had the default and incorrect collation of Latin1_General_CI_AS while the OpsMgr databases had the correct collation of SQL_Latin1_General_CP1_CI_AS (check the properties of SQL Server and the databases in SQL Management Studio to verify).

And this pretty much explained why reports from new management packs weren’t appearing in OpsMgr.  The odd thing is that this problem went unnoticed for over 6 months and many management packs functioned perfectly well.

I knew what was ahead of me: a SQL rebuild.  So that’s what I did, with some guidance from a blog post by Marnix Wolf, MVP.  I veered a little from the guidance he gave.  I opted to start with a new SQL Reporting DB because it was easier to do this and I had no customisations to rescue.  So I didn’t restore it, I didn’t run ResetSRS, and I just needed to reinstall OpsMgr Reporting and supply the details.

Interestingly, the OpsMgr Reporting installed froze about half way through.  There were no visible issues, no performance bottlenecks, no clues, nothing to explain the setup hang … except for the Application Log in Event Viewer.  There McAfee reported that it was preventing lots of .Net stuff.  Uh oh!  I temporarily disabled the McAfee protection and the installer wrapped up almost immediately.

Once everything was back I verified that monitoring worked, that the datawarehouse was still OK, and that reports were repopulating and working.  But then a flood of alerts came in:

Microsoft.EnterpriseManagement.Common.UnknownServiceException: The service threw an unknown exception. See inner exception for details. —> System.ServiceModel.FaultException`1[System.ServiceModel.ExceptionDetail]: Execution of user code in the .NET Framework is disabled. Enable "clr enabled" configuration option. (Fault Detail is equal to An ExceptionDetail, likely created by IncludeExceptionDetailInFaults=true, whose value is: …

That looked nasty but the fix was easy enough.  As Alexy Zhuravlev said, run this on the SQL server against the OperationsManager database:

sp_configure @configname=clr_enabled, @configvalue=1
GO
RECONFIGURE
GO

After that, everything was okey dokely and the SQL 2008 R2 DB was updated to get it OpsMgr 2012 ready Smile

Deployed Veeam nworks Management Pack For vSphere

I deployed the Veeam management pack for System Center Operations Manager with a client on their site site yesterday to monitor VMware vSphere.  It was my first production deployment of the solution.  It was pretty simple:

  • Deploy collectors
  • Discover vCenter servers/hosts
  • Monitor
  • Run reports

Oh and the reports!  There’s so many of them with lots of information.  It’s a very nice management pack.  And it accomplishes what the client wanted: they have visibility into VMware from System Center.

Does it work?  Yeap; it detected read latency on a HBA, an oversubscribed VMFS volume (based on potential growth of thin VMDKs), and a full VMFS.

Citrix XenApp Monitoring Management Pack For System Center Operations Manager – Import Fails

I’ve been doing some on-site work this week with System Center Operations Manager.  The customer has some XenApp servers they wanted to monitor them.  Due to lockdowns, we had to do manual agent installations (with approval done in the OpsMgr console).  The management pack is a free download from Citrix via the MyCitrix site.  We also found the management pack files in a subfolder on the root of ISO.  The import failed with this error:

Citrix Presentation Server Management Pack could not be imported.

If any management packs in the import list are dependent on this management pack, the installation of the dependant management packs will fail.

The requested management pack was invalid. See inner exception for details. Parameter name: managementPack.

The solution is simple enough.  Increase the size of the OperationsManager database transaction log file from 100MB to 1,000MB, and then reattempt the import.  It worked for me.

Technorati Tags: ,

System Center Operations Manager Saves The Day … Even In A vSphere Site

Way-back-when, I deployed Microsoft Operations Manager 2005 just after it had RTMd.  My boss, the IT infrastructure manager, decided that we should do our initial agent deployment in our DR site.  We had Windows and HP ProLiant management packs imported.  The DR site in question was rarely visited: pretty much whenever we needed to install something new or when we had a test invocation scheduled.

Within minutes, the agents started reporting degraded hardware: memory DIMMs, RAM, and PSUs.  That won my boss over.  Once the network team were happy, we started deploying agents 17 worldwide sites.

This week I’ve been involved with a proof-of-concept deployment of OpsMgr 2007 R2 CU5 in a VMware environment.  The customer wanted to see how it would handle monitoring of a critical service that had received significant investment and attention from the business.  The management group was built, and agents were deployed to the application servers (any consultant who deploys agents to hundreds machines at once is being negligent because the customer will reject the un-tuned full of noisy alerts monitoring solution).  A handful of management packs were imported, including Windows and SQL Server.  And within minutes we had detected an issue.  The SQL log file for the application was not able to expand and the critical LOB app was about to fail.  Nice timing Smile (I swear I didn’t cause it!).  The customer’s IT staff were on it and the problem was avoided.  Then a day later, once the data warehouse was populated with some info, I ran some performance reports and identified a vCPU bottleneck in the SQL server VM and a recommendation was made there.

To quote Charlie Sheen: WINNING!

It would be easy to think that SysCtr is irrelevant to VMware.  Sure, in my opinion System Center + Hyper-V exceeds the alternative.  But, elements of SysCtr + vSphere easily exceeds vSphere by itself or with some overpriced point solution with a “v” badge stuck on it.

We know the business values applications (or services).  They couldn’t care less about vSphere VS Hyper-VS any other virtualisation.  Now this customer has SLA monitoring/reporting on this particular LOB application and thanks to the early warning from OpsMgr, it’s sitting nicely at 100%, and the IT department’s customer is a happy camper.

Monitoring the Hybrid Microsoft Cloud

The Microsoft Hybrid cloud, as it stands currently, is a mixture of a Hyper-V private cloud with an Azure public cloud, managed by System Center App Controller (formerly Concero).  One of the key pieces of the Microsoft solution is monitoring the health of the application (that the business really cares about) using System Center Operations Manager (OpsMgr).

Management packs make monitoring of Hyper-V, Windows, SQL, Exchange, CRM, hardware, storage, etc, easy.  You can put together end user perspective monitoring from the basic ping test to the advanced synthetic transaction, build service-centric distributed application models, and provide SLA monitoring of the LOB applications.  That’s got the private cloud covered.

There is also a management pack for Azure.  This allows you to monitor the availability, health, and performance of your public cloud services.  Let’s face it – even if Microsoft does/did provide a monitoring solution within Azure – can you really use a monitoring solution that is a part of the thing you are monitoring, i.e. the Microsoft public cloud?  I say no – and that’s the first reason why you should use OpsMgr and this management pack.  The second reason is that it allows you to integrate your monitoring of public and private clouds, giving you that mythical single pane of glass for monitoring.

  • The features of this management pack are:
  • Discovers Windows Azure applications.
  • Provides status of each role instance.
  • Collects and monitors performance information.
  • Collects and monitors Windows events.
  • Collects and monitors the .NET Framework trace messages from each role instance.
  • Grooms performance, event, and the .NET Framework trace data from Windows Azure storage account.
  • Changes the number of role instances via a task.

The prerequisites of it are:

  • The management group must be running Operations Manager 2007 R2 Cumulative Update 3.
  • The Windows Azure role must be published with full trust level. For more information about Windows Azure trust levels, see Windows Azure Partial Trust Policy Reference.
  • Windows Azure Diagnostics must be enabled. For more information about Windows Azure Diagnostics, see Implementing Windows Azure Diagnostics.
  • Windows Azure Diagnostics must be configured to forward diagnostic data to a Windows Azure storage account. For more information about configuring Windows Azure Diagnostics, see Transferring Diagnostic Data to Windows Azure Storage.
  • Microsoft .NET Framework version 2.0 or newer must be installed on the computer that you designate as the proxy agent when you configure the Monitoring Pack for Windows Azure Applications.

Operations Manager 2007 R2 Downloads – RHEL 6 Support

Cumulative Update 5 for System Center Operations Manager (SCOM/OpsMgr) 2007 R2 was released last night.

And now RedHat Enterprise Linux 6 is supported (catching up with Hyper-V)!

There are more details on the CU5 support page.

A new Cross-Platforms management pack was also released.  Oddly, the download page mentions RHEL4 and 5 as supported but not the newly CU5 supported RHEL 6.  Don’t worry, the MP’s word document does mention that RHEL6 is actually supported by the new MP version.  It also mentions a few other fixes and new features of the MP.

System Center Operations Manager 2012 Beta

OpsMgr/SCOM 2012 beta has been launched.  It will …

“… help you manage your data centre and cloud environments by:

  • Delivering flexible and cost effective enterprise-class monitoring and diagnostics while reducing the total cost of ownership by leveraging commodity hardware, with standard configurations to monitor heterogeneous environments.
  • Helping to ensure the availability of business-critical applications and services through market-leading .NET application performance monitoring and diagnostics plus JEE application health monitoring.
  • Providing a comprehensive view of data centres, and private and public clouds.

Feature Summary

  • Predictable performance and availability of critical applications
    • End-to-end views of application health and topology
    • Establishment of application service-level delivery (SLAs)
    • Precise identification of application errors

    Flexible and cost-effective infrastructure monitoring

    • In-depth monitoring, diagnostics, and reporting for heterogeneous environments
    • Integrated network device monitoring and alerts
    • Simplified management infrastructure
  • Comprehensive monitoring for your data centre and cloud—on your terms
    • Integrated physical, virtual, and cloud management
    • Common console across data centre and clouds
    • Rich reporting”

The things that look most interesting to me from the TechNet videos that I’ve watched are the more fault tolerant/simplified management server groups, and built-in network monitoring.  The second product I ever worked with in my career was a networking product that my then employer started to sell (but stopped soon after we started training sales people).  OpsMgr appears to work similarly: feed in a bunch of network device credentials, point it at one “seed” device, and let OpsMgr discover the rest of the network from there based on what it finds.  And after that you have port and protocol level fault and performance monitoring.  I guess we’ll find out more once we get the beta installed.