Tag Archives: MimboloveSCOM Management Pack

OpsLogix Capacity Report Management Pack Overview

Written by Tao Yang

capacity-banner-bgJust over a month ago, I have blogged and presented a webcast comparing the OpsLogix Capacity Report Management Pack and the OMS Capacity solution. Since then, an update was released on this management pack and I’d like to take a moment to provide a proper overview for this MP. For those who have not used this management pack and are looking for a solution for capacity forecasting and management, I hope you will have some ideas on the capabilities this management pack provides.

Management Pack Introduction

The OpsLogix Capacity Report MP provides OpsMgr reports that can be used to forecast trending of any existing performance data collected by OpsMgr. Same as any other OpsMgr reports, the reports provided by this MP can be accessed from the reporting pane in the OpsMgr console, under “OpsLogix IMP – Capacity trending reports” folder:

image

Installing and Configuring Management Pack

Other than the capacity report MP itself (OpsLogix.IMP.Capacity_v1.0.2.24.mpb), I was also given a zip file containing my license key. This zip file contains an unsealed MP (OpsLogix.IMP.Capacity.License.xml), which contains my license key and it is unique to my environment. I need to import the license MP into my OpsMgr management group together with the capacity report MP. Once both MPs are imported, you will able to see the reports from the folder shown in the screenshot above.

Reports

This MP offers the following reports:

  • Absolute value Report – Single instance
  • Absolute value Report – Multi instance
  • Percentage value Report – Single Instance
  • Percentage value Report – Multi instance
  • Percentage value Report – Multi instance Critical Only
  • Percentage value Report – Single instance Critical Only

I will now go through these reports.

Absolute value Report – Single instance

This report allows you to run a forecast report over any performance counters stored in the OpsMgr data warehouse DB. This report requires the following parameters:

image

  • From: The start date for the forecast analysis. The default value is Today – 30 days
  • To: the number of forecast days. default value is 30 days from today
  • Time zone: choose the time zone of your choice
  • Available Rule Languages: The default value is English, when choosing another language, the performance rules that have display strings defined in that particular language (defined in <LanguagePacks> section in management packs) will appear in the “Performance Rule” drop down list.
  • Performance Rule: this drop down list contains all the performance rules available for the language that you have chosen.
  • Counter:  this drop down list contains the counters collected by the performance rule that you have selected. As the best practice, a perf collection rule should only collect one counter, so hopefully you should only see one counter on this drop down list.
  • Object: this drop down list contains the object associated to the performance rule and counter
  • Instance: this drop down list contains a list of available instances for the performance rule and counter.
  • Managed Entity: this drop down list contains a list of managed entities associated to the performance counter instances.

The report looks something like this:

image

as shown in the screenshot above, the light blue line indicates forecasted future trending for the particular counter that you have chosen. The report also shows the forecasted change and value for the perf counter.

Absolute value Report – Multi instance

This report is very similar to the “Absolute value Report – Single instance” report. the only difference is, we can choose multiple instances in this report:

image

As shown in the screenshot above, we are able to choose multiple instances in the “Instance” section (whereas in the single instance report, we can only choose one from the drop down list). The report output displays all instances that you have selected:

image

Percentage value Report – Single Instance

When performance counters are being collected by OpsMgr, some counter values are absolute values (such as logical disk free space in MB). However, some the counter values are percentage based (i.e. % logical disk free space).the percentage value based reports are designed for the percentage based performance counters. Let’s take a look at the “Percentage value Report – Single Instance” report first. This report requires the following parameters:

image

    • From: The start date for the forecast analysis. The default value is Today – 30 days
    • To: the number of forecast days. default value is 30 days from today
    • Time zone: choose the time zone of your choice
    • Number of days for Warning level: choose the warning threshold. – if the forecasted value reaches 100% (or 0% if reverse forecast direction is set to true) within the value specified in this field, the forecasted capacity state will be warning. The default value is 60 days.
    • Number of days for Critical level: choose the critical threshold. – if the forecasted value reaches 100% (or 0% if reverse forecast direction is set to true) within the value specified in this field, the forecasted capacity state will be critical. The default value is 30 days.
    • Reverse forecast direction: by default, the forecasted capacity state is changed when the forecasted value reaches 100%. But in some cases, we are more interested when the value reaches 0% (i.e. free disk space). In these scenarios, you can specify Reverse forecast direction to “true”.
    • Available Rule Languages: The default value is English, when choosing another language, the performance rules that have display strings defined in that particular language (defined in management packs) will appear in the “Performance Rule” drop down list.
    • Performance Rule: this drop down list contains all the performance rules available for the language that you have chosen.
    • Counter:  this drop down list contains the counters collected by the performance rule that you have selected. As the best practice, a perf collection rule should only collect one counter, so hopefully you should only see one counter on this drop down list.
    • Object: this drop down list contains the object associated to the performance rule and counter
    • Instance: this drop down list contains a list of available instances for the performance rule and counter.
    • Managed Entity: this drop down list contains a list of managed entities associated to the performance counter instances.

image

As you can see from the screenshot above, I have chosen a perf collection rule that collects the % logical disk space for Windows Server 2012. Because this is the single instance report, we are only able to select one instance from the drop down list – in this case, the instance for the perf counter represents the drive letter of Windows Server 2012 logical disks. I have chosen D: drive as the instance, and selected all the D: drives on my Hyper-V hosts from the Managed Entity section.

For the third item on the report indicates according to the forecast, it will run out of space in 206.29 days. Since the warning threshold is configured as 400 days and critical is 200 days. The value 206.29 falls in between the warning and critical threshold, therefore the forecasted capacity state is warning.

For the last item (the 4th) on the report, the forecast indicates it will run out of capacity in 109.46 days, which is less than the configured critical threshold of 200 days, therefore, the forecasted capacity state is critical in this case.

Percentage value Report – Multi Instance

This report is similar to the “Percentage value Report – Single Instance”, but it allows you to select multiple instances for the perf counter you have chosen:

image

In the example above, I have chosen the “% Logical Disk Free Space Windows Server 2012” perf collection rule, which collects the % Free Space counter for Logical disks on Windows Server 2012 computers. In this case, the instance represents each logical disk’s drive letter (as highlighted). Comparing with the single instance report, we are not only able to choose a specific drive (such as C: drive), but also any other drives (as shown below).

image

Percentage value Report – Multi / Single instance Critical Only Reports

The last two reports from this MP are the “Percentage value Report – Multi instance Critical Only” and “Percentage value Report – Single instance Critical Only” reports. The only differences with these two reports comparing to the previously mentioned percentage value reports is, they filter out any items with healthy and warning forecasted capacity state, and only list the critical items:

image

So if you don’t really care about the healthy and warning items, and only want to concentrate on critical items, you may find these 2 reports handy.

Summary

The OpsLogix Capacity Report MP provides generic forecasting reports that can be used against any types of performance data collected by OpsMgr. As long as the related perf counters are being collected by OpsMgr, the reports can be used when planning future capacities. The audiences of this MP can be anyone who are using OpsMgr (i.e. server admins, network admins, cloud and fabric admins, DBAs, LOB application owners, etc).

Lastly, if you have any questions, please feel free to contact myself, or OpsLogix sales team directly (sales@opslogix.com).

OpsMgr Self Maintenance Management Pack

Written by Tao Yang

Background

I had this idea to automate some repetitive maintenance tasks that every SCOM administrator perform regularly via a management pack. So I started to write a OpsMgr self maintenance MP during my spare time around July last year, right after my baby girl Rosie was born. since then, I’ve got side-tracked many times, i.e. left for TechEd in September for a week, studying for ConfigMgr 2012 exam, studying for the 3 Windows Server 2012 MCSA exams, writing other management packs such as the Weather monitoring MP, etc. This MP has been sitting there 90% completed for months.

Few of weeks ago, I passed all 3 Windows Server 2012 exams required for MCSA, and got a phone call from my friend Dan Kregor. Dan told me he wanted to write a maintenance MP for OpsMgr, and I told him I’ve got one 90% done. So we got together, went through what needed to be included and Dan gave me some good ideas. Over last few weeks, I have spent most of my spare time working on this MP. Initially this MP was only designed for 2007 (which would also work on 2012), I then decided to build a 2012 version (based on OpsMgr 2012 MP schema) so the MP can utilise resource pools. There are also few additional rules / monitors in the 2012 version. Anyways, I’ve now finally finished both version of the MP. Below paragraphs are ripped from the MP documentation:

Introduction

OpsMgr Self Maintenance Management Pack automates some routine tasks generally performed by OpsMgr administrators on a regular basis. It also provides few rules / monitors to monitor the OpsMgr management group itself. This management pack contains 2 version.

  • The OpsMgr 2007 R2 version works on both 2007 R2 and 2012 versions of OpsMgr.
  • The OpsMgr 2012 version only works on OpsMgr 2012.

The 2012 version of this management pack is able to utilize OpsMgr 2012 resource pools and also provides additional rules and monitors than the 2007 version. For OpsMgr 2012 environments, it’s strongly recommended to use the 2012 version of this management pack.

The following workflows are included in this management pack:

  • Automatically balance OpsMgr agents among a group of management servers.
  • Automatically close aged rule-generated alerts
  • Convert all manually installed OpsMgr agents to Remote-Manageable.
  • Enable Agent-Proxy for all agents
  • Backup Unsealed (and Sealed) management packs.
  • Remove Disabled discovery objects
  • Detect staled stage change events
  • Monitoring the size of LocalizedText able from the OpsMgr operational database.
  • Detects OpsMgr management servers in maintenance mode (Only available in OpsMgr 2012 version of the MP)
  • Performance Collection rule for total number of SDK connection within the management group (among all management servers). (Only available in OpsMgr 2012 version of the MP).
  • Agent tasks for:
    • Manually backup management packs
    • Get currently connected users to the SDK service
    • Enable Agent Proxy for all agents.

All the rules and monitors from the OpsMgr Self Maintenance management packs are disabled by default. This is to ensure OpsMgr administrators only turn on the workflows that are required for the OpsMgr environments they support and configure the required parameters for workflows to suit the environment.

The agent tasks from the management packs are enabled by default.

An unsealed override management pack is provided for each version of the OpsMgr Self Maintenance MP. OpsMgr administrators can use provided unsealed override MP for customization or they can also create their own override MPs for this purpose.

Credit

I’d like to thank Dan Kregor for sharing ideas with me and taking time testing the MP.

I appreciate any feedback from the community, this is only the first release, any thoughts on additional workflows that I can build into this MP is much appreciated!

Both versions of the management packs and the documentation can be downloaded HERE.

Setting SMS_INSTALL_DIR_PATH Environment variable on SCCM site servers for SCOM SCCM management pack

Written by Tao Yang

According to the “Configuration Manager 2007 SP2 Management Pack User’s Guide for Operations Manager 2007 R2 and Operations Manager 2007 SP1” (for MP version 6.0.6000.2), An environment variable named “SMS_INSTALL_DIR_PATH” needs to be created on all SCCM site servers.

I had to do this on 80+ site servers, so I thought why not do this using a script and let SCCM to push it out to all site servers?

Therefore, I wrote a VBScript Set-EnvirVar-For-SCOM.vbs (I didn’t use PowerShell this time because it is easier to push out VBScripts via SCCM).

Source Code:

Set objWMIService = GetObject("winmgmts:" _
& "{impersonationLevel=impersonate}!\\.\root\cimv2")
Set colServices = objWMIService.ExecQuery _
("Select * from Win32_service Where Name = 'SMS_SITE_COMPONENT_MANAGER'")

For each item in colServices
ServicePath = item.PathName
Next

IF Left(ServicePath,1) = Chr(34) THEN
ServicePath = Right(ServicePath, Len(ServicePath)-1)
END IF
IF Right(ServicePath,1) = Chr(34) THEN
ServicePath = Left(ServicePath, Len(ServicePath)-1)
END IF

InstDir = Replace(ServicePath,"\bin\i386\sitecomp.exe","")

EnvVarName = "SMS_INSTALL_DIR_PATH"
IF LEN(InstDir)>0 THEN
Set objWMIService = GetObject("winmgmts:\\.\root\cimv2")
Set objVariable = objWMIService.Get("Win32_Environment").SpawnInstance_
objVariable.Name = EnvVarName
objVariable.UserName = "<System>"
objVariable.VariableValue = InstDir
objVariable.Put_
ELSE
'Wscript.Echo "This is not a SCCM Site Server"
END IF

'Wscript.Echo "Done"

This script retrieves the path of the executable (sitecomp.exe) for the SMS_SITE_COMPONENT_MANAGER service via WMI class Win32_Service, and determine the SMS Install Path based on the path to the executable. It then create a new system environment variable if the path is valid. – This means you can run this script against non-SCCM Site servers and it will not create the environment variable on these machines.

Once I’ve tested the script by manually running in on different servers, I then create a SCCM collection called “SCCM Site Servers”. This collection is a query based collection that looks for machines running “SMS_SITE_COMPONENT_MANAGER” service.

Finally, I created a package and deployed it out to this collection.