Removing Discovered Data
I recently received a request from one of my Biztalk Guys to remove two of his Biztalk Servers from the Biztalk Views, as they were used soley for QA and had constant errors. Here is how I accomplished that task:
PROBLEM
Disable BizTalk discoveries and discovered inventories
RESOLUTION
Do the following steps to disable discovery.
- Go to SCOM Console -> Authoring console -> Object Discoveries -> Scope.
- Clear all -> Type “Biztalk Server” -> Ok.
- Right click on “Biztalk Server Role Discovery” -> Properties -> Overrides -> Override -> For a specific object of class : Windows server
- Select that Biztalk server which you want to disable the discovery.
- Either wait for 43200 seconds because this rule runs in that interval or override this parameter also to smaller interval and then remove it later once we remove the discovered inventory.
- Once this discovery rule runs then open operations manager powershell command prompt - > execute “Remove-DisabledMonitoringObject”.
***Please note this process works for all Discovered Inventory, i.e. SQL, AD, Exchange.***
Best Practices for Testing Fiber – From Fluke Networks
Does your organization lack consistent techniques or adequate processes for testing fiber? To minimize costly installer/contractor callbacks, network technician troubleshooting time, and unnecessary network downtime, these three fiber-handling best practices should always be followed.
The Fiber Testing Best Practices series was designed by Fluke Networks to educate about three straightforward, yet important fiber handling best practices. This quick-read series of papers will give you the framework for implementing best practices for fiber testing within your organization:
HP Proliant Support Pack Update Script
If your an HP Shop you are most likely aware of PSP's. You are also probably not updating them as often as you should. Hopefully this write up will make your PSP updates a little less difficult as well as less time consuming. ProLiant Support Packs (PSP) represent operating system (OS) specific bundles of ProLiant optimized drivers, utilities, and management agents. These bundles of software are tested together to ensure proper installation and functionality. Many IT departments never update the PSP's in the environments, however it is recommended to update regularly to ensure optimal performance of your systems . In DEV and QA first, then after successfully tested to Production. There are a few different ways to accomplish this goal:
- Install using the Smart Start CD manually to each server- This sneaker-net option is the most time consuming, even if you have a small environment. Also, this would require user interaction on every server.
- Deploy via the HP SUM GUI, using one or two servers as the master server pushing out the source files - This option is better than the first however it still has its drawbacks. This option is limited by the performance of the master servers, also the HP SUM application is a Java app (take that how you like).
- Deploy using HP SIM. This method is the preferred method recommended by HP- Although this is the preferred method, this is almost as worse as option one in my mind. Your bottleneck is one single server; also HP SIM does a horrible job at task management. It used to take me hours to more than 12 hours to deploy a mere 100 PSP upgrades. I spent hours on the phone with HP when finally our account rep put me in contact with its Software Engineering team. They informed me that in later versions, HP SIM will use HP SUM as the upgrade engine, which is faster than what they currently have.
- Deploy using a scripted HP SUM method - My boss didn't like the answer I received from HP and he told me to come up with an option that will get all of our 4000 servers done in a window of 3 hours. I brainstormed for a few days when I came up with the below method that got it done in less than 1 hour. This option is the fastest and it provides the same functionality as if you were to deploy using SIM. Here are a few diagrams that outline this process.
Network Diagram
Process Diagram
Instructions
- Designate a server to be your Update Server, this server will be hosting the source files and scripts. Processing will not happen locally as we'll be using PSEXEC to update the servers individually.
- Download the 32bit and 64bit versions of the Support pack and extract them. Share them accordingly, for 32bit use \\Update Server\PSP , and for 64bit use \\UpdateServer\PSP64.
- Create a batch file in each of PSP folders called Upgrade.bat and add the code below to this batch file. This script will use a command line version of HP SUM to find what updates are needed and apply them accordingly. This option will not cause the servers to reboot, if you need this you can add that line of code accordingly. This batch file will also write two event log entries to indicate when the update process started and ended. **Essentially, executing this script alone will update your server. However, logging into every server, copying the PSP files, then executing, then deleting the files, then logging off would take a long time. Hence the use of PSEXEC and the rest of the process**
- Create the Share \\UpdateServer\SIMUpdateScripts\ and child shares for \32 and \64. Create the next script called Install.Bat in both 32 and 64 bit folders. This script will be used to copy the PSP files down to the D: Drive and Write Event log Entries indicating start and stop for of each local sever. It will then call the Upgrade.bat files locally and start the upgrade process. You will need to modify the script for the 64 bit version.
- Next create a Script called Delete.bat in both 32 and 64 folders. This script deletes the PSP contents from the local servers. The 64 bit version will have to be modified accordingly.
- Download PSTools and extract to c:\PStools. Then create a list of servers you wish to update in file called C:\Lists\upgrade.txt
- Finally, it's time to Execute the upgrade process. Navigate to c:\Pstools. Then run the command below. This command will launch a non-interactive command shell locally and execute the Install.Bat file on every server in your list. As I mentioned in Step 4, this script will copy the PSP contents, then call Upgrade.bat. The Upgrade Script will install all the necessary software and drivers that are outdated per that PSP.
- The Script has event logging built in so you can monitor for event 999 "Update.bat has finished". Once that is complete you can use PSEXEC to issue a reboot. After your servers have been rebooted, you can use PSEXEC to remove the PSP. For more info on usage visit the PsExec Technet site.
eventcreate /t information /id 998 /l application /d "Upgrade.bat has started" hpsum /force /allow_non_bundle_components /silent eventcreate /t information /id 999 /l application /d "Upgrade.bat has finished"
eventcreate /t information /id 998 /l application /d "Copy has started" start /wait /b xcopy \\UpdateServer\PSP\*.* D:\PSP /c /d /s /e /k /y /z /i eventcreate /t information /id 998 /l application /d "Copy has finished" CALL D:\PSP\upgrade.bat
eventcreate /t information /id 998 /l application /d "Delete has started" RD D:\PSP /S /Q eventcreate /t information /id 999 /l application /d "Delete has finished"
Server1.testdomain.com Server2.testdomain.com Server3.testdomain.com
psexec @c:\lists\upgrade.txt -u domain\username -p PASSWORD -d "\\UpdateServer\SIMUpdateScripts\32\Install.bat"
Windows Reliability and Performance Monitor
Windows Reliability and Performance Monitor is a awesome new feature built into Windows Server 2008. It's coolest features are Performance Reporting, Data Collector Sets for Performance and configuration and the Reliability Gantt chart. Here is some more info on some of the key features:
Data Collector Sets
An important new feature in Windows Reliability and Performance Monitor is the Data Collector Set, which groups data collectors into reusable elements for use with different performance monitoring scenarios. Once a group of data collectors is stored as a Data Collector Set, operations such as scheduling can be applied to the entire set through a single property change. You can schedule repeated collection of a Data Collector Set to create logs, load it in Performance Monitor to see the data in real time, and save it as a template to use on other computers.
Windows Reliability and Performance Monitor also includes default Data Collector Set templates to help you begin collecting performance data immediately.
Wizards and templates for creating logs
You can now add counters to log files and schedule their start, stop, and duration through a wizard interface. In addition, if you save this configuration as a template, you can collect the same log on subsequent computers without repeating the data collector selection and scheduling processes. Performance Logs and Alerts features have been incorporated into the Windows Reliability and Performance Monitor for use with any Data Collector Set.
Resource View
The new Resource View screen provides a real-time graphical overview of CPU, disk, network, and memory usage. By expanding each of these monitored elements, you can identify which processes are using which resources. In previous versions of Windows, this real-time, process-specific data was only available in limited form in Task Manager.
Reliability Monitor
Reliability Monitor calculates a System Stability Index that reflects whether unexpected problems reduced the reliability of the system. A graph of the Stability Index over time quickly identifies dates when problems began to occur. The accompanying System Stability Report provides details to help troubleshoot the root cause of reduced reliability. By viewing changes to the system (installation or removal of applications and updates to the operating system) side by side with failures (application, operating system, or hardware failures), you can develop a strategy for addressing the issues quickly.
Unified property configuration for all data collection, including scheduling
Whether you create a Data Collector Set for one-time use or to log activity on an ongoing basis, the interface for creation, scheduling, and modification is the same. If a Data Collector Set proves to be useful for future performance monitoring, you do not need to recreate it. You can reconfigure or copy it as a template.
User-friendly diagnosis reports
Users of Server Performance Advisor in Windows Server 2003 can now find the same kinds of diagnosis reports in Windows Reliability and Performance Monitor in Windows Vista. You can generate reports more quickly and can generate reports from data collected using any Data Collector Set. This allows you to repeat reports and assess how recommended changes have affected performance or modified the report recommendations. Windows Reliability and Performance Monitor also includes preconfigured performance and diagnosis reports for quick analysis and troubleshooting.
Windows Vista Performance and Reliability Monitoring Step-by-Step Guide [Via Microsoft Technet]
Logical/Physical Disk Foundational Performance Guidelines
As a Systems Engineer I understand the importance of Disk IOPS, whether it be Local, Directly Attached or SAN. It is obvious that if a disk is taking more than a couple of seconds to read or write then there is a problem. But what about the not so obvious metrics? How can you tell if your disk are performing out of spec, or if you need to upgrade? These are questions that I recently asked myself when my DataWareHouse started becoming sluggish. I found a Microsoft White Paper that outlined a few WMI Performance Counters that one would monitor when investigating possible disk latency.
In SCOM or Perfmon, load up the counters \%Idle and \%Avg Disk Sec Read or Write. Below are the Counter Guidelines:
\%idle
* 100% idle to 50% idle = Healthy
* 49% idle to 20% idle = Warning or Monitor
* 19% idle to 0% idle = Critical or Out of Spec
\%Avg. Disk Sec Read or Write
* .001ms to .015ms = Healthy
* .015ms to .025 = Warning or Monitor
* .026ms or greater = Critical or Out of Spec
If you notice that your disk are sustaining a Critical Spec, then you need to adjust accordingly. The answer could be looking at your RAID configuration, moving to faster disk or a faster RAID Controller, or even upgrading the number of spindles if the disk is on a SAN.
How do I know an agent has received my management pack changes?
When you import a new management pack, or make changes to an existing MP (overrides, custom rules, etc), the changes are written to the OperationsManager database. This begs the question, “How do I know when an agent has received my changes?”.
Well, there are actually an event trail and some evidence on the file system that indicate your changes were in fact received. As you might expect, the process is the same in Operations Manager and Essentials 2007.
It goes a bit like this:
* You import a new MP or modify an existing MP. The information is written to the operational database (OperationsManager).
* The RMS identifies the change and communicates this to the management servers.
* The management servers pass the new or updated management pack to the agents.
NOTE: To get a good idea of the function of the Config service on the RMS, stop the Config service, and then import an MP or make any change. You will notice the agent never receives the change. You’ll also notice state changes are no longer calculated. However, you will notice the Operations Console continues to function, MPs can be imported, etc.
* When an agent receives the MP, an event 1201 is logged to the Operations Manager log on the agent. The event includes the name and version of the MP that was received.
* When the changes have been applied, an event 1210 is logged on the agent, indicating the new configuration is now active.
You can also find evidence of the configuration change by looking on the file system of the agent. In the agent installation directory (c:\program files\system center operations manager 2007\ by default), you will find a \Health Service State\ManagementPacks\ subdirectory. In this directory, you will see copies of the management packs received. Simply by watching the “Date Modified” column (in detail view), you can see when an updated copy was last written to the file system.
Preparing Agent Trace Files
Those with large Opsmgr Enviroments sometimes come across wacky error message in the event logs derrived from the Opsmgr Agent. Often, the information in the Event Logs provide little to no help, and a call to Microsoft Premier Support is inevitable. They often ask for you to send them over the trace files for the agent. This article describes how to do so:
Scheduling the Approve of Pending Agents
I recently needed to script the approval of manually installed agents that had a certain character set in their name. Our configuration of SCOM requires that manually installed agents be approved via the console. Scripting was the best solution, as it didn’t involve human intervention.
To schedule a Powershell task, you need to create a batch file that will run your powershell script, this is the program you select in the task scheduler. Powershell.exe on its own does not have the cmdlets loaded for SCOM, therefore you have to launch Powershell and run additional commands.
In the actual script, I needed to add the OpsMgr PowerShell snap-in, create the Monitoring drive mapped to the OperationsManagerMonitoring provider, and create a connection to the RMS. I also load the OpsMgr custom Windows PowerShell script (Microsoft.EnterpriseManagement.OperationsManager.ClientShell.Startup.ps1) to load OpsMgr-specific functions.
Get Batch File and Script Here
Management Pack Objects in Operations Manager 2007
This is my breakdown on Management Packs in SCOM 2007. Below are snippets of text that I pieced together from Microsoft and other sources that best explain Management Packs. This is intended to be an overview and not an explanation of explicit MP's.
Management Packs are the building blocks which extend Operations Manager 2007 management capabilities to operating systems, applications, and other technology components. A Management Pack (MP) contains best practice knowledge to discover, monitor, troubleshoot report on, and resolve problems for a specific technology component. Management Packs include health models based on the Systems Definition Model (SDM) to analyze the performance, availability, configuration and security inputs, as well as the status of related components, to determine the overall status of components. Management Packs are available for Microsoft and third party operating systems, applications, and devices.
Use the Management Packs Objects node, in the Authoring pane of the Operations Console, to create objects that help you define how you monitor objects in Operations Manager 2007. This node also displays the monitoring settings of imported Management Packs. You can view existing attributes, monitors, object discoveries, rules, tasks, and views by clicking the appropriate leaf object under the Management Pack Objects node. You can also create new attributes, monitors, rules, and tasks from each corresponding leaf object.
Note: You cannot create object discoveries or views from the Authoring pane. Object discoveries are created using the Discovery Wizard from the Administration pane, and views are created in the Monitoring pane.
The following list describes the objects listed in the leaf nodes of the Management Pack Objects node.
Attribute
Displays a list of attributes for each object type in your Management Group. Attributes contain information that further defines an object in Operations Manager.
Monitor
Displays a list of monitors sorted by object type. Monitors continually assess the condition of specified objects. As a result of this assessment, a monitor can also generate alerts and change the health state of that object.
Object Discoveries
Displays a list of discovery objects currently in use in your Management Group. A discovery is used to dynamically find objects on your network that you want to monitor.Rule
Displays a list of rules sorted by object type. Rules collect data, such as event information, generated by managed objects.
Task
Lists the tasks that are available within your Management Group sorted by object type. Tasks are actions that run against a monitored object.
View
Displays a list of available views in the Management Group. Views display a particular aspect of monitoring settings. The displayed information in a view is the result of a query to the Operations Manager database.
Project – Custom Desk
I finally did it, I built the custom built desk I was talking about for weeks. I used the advice from the following sites and came up with my own desgin: Simplified Building Concepts , Gadget Wise-NY Times. Here is what I did in a Nutshell:
- Bought MDF from Home Depot, had them pre-cut it to my dimensions. They will make any straight cut's for you.
- I used Adjustable Legs from Ikea, "Vika Haj". They were about $15 a piece.
- I didn't have a router so I used a Wood Shaver and Sandpaper to make rounded edges.
- After assembling the legs I sanded the MDF, then Sealed with BIN Primer, sanded, applied another coat of primer, sanded, painted, sanded, painted again, then I used Polycrilic to seal.
-
- Before
-
- After
