Forums

SCOM 2007 R2 and Monitoring Individual Process Performance

Joseph H. Schwendt II — Tue, 27 Jul 2010 22:48:37 GMT

Hi,

I'm have a hard time trying to figure out how to monitor the performance of an individual process across my environment. I'm looking to track at least CPU and Memory utilization, but also tracking process id would be helpful to tying performance back to log files for post analysis.

In my case, I'm looking to monitor lots of instances of w3wp.exe (50+/server) spread among lots of servers (50+). To attempt to do this with Windows PerfMon is an excercise in total frustration. I'm easily able to monitor these counters in aggragate using the built in management packs, I just can't figure out how to break out individual processes.

I'm certainly aware that this sort of thing incurs significant overhead, so I'm looking to keep the polling interval on the larger size (say every 30 seconds or so). Note that I'm not looking to alert on any of this, nor am I concerned with number of instances of a specific process running.

Any help is greatly apprecaited!

Thanks,

Joe

Re: SCOM 2007 R2 and Monitoring Individual Process Performance

curtmcgirt — Fri, 25 May 2012 00:26:55 GMT

i was looking for this too. that MP was written for IIS 2003, but for anybody else who looking for an update for IIS 7, i just

1. opened this xml in the authoring console

2. added the IIS 7 MP as a reference

3. changed the target of the Rule and Monitor to "IIS 7 Application Pool" instead of iis 2003 application pool.

4. pushed the MP to my management group from the authoring console

5. adjusted the monitoring views to use IIS 7 application pool instead of IIS 2003 application pool.

everything seems to be working.

Exchange 2010 SP2 and the Exchange MP v14.2.71.0

LarryAlthouse — Thu, 24 May 2012 21:42:59 GMT

Is anyone monitoring Exchange Server 2010 SP2 with the last release of the Exchange 2010 MP (v14.2.71.0). One of our teams is getting ready to push Exchange SP2 and we're being asked if there will be any problems on the OpsMgr side, but I cannot find any information that addresses this directly. Please let me know if you are running this configuration and if there were any problems in OpsMgr going from Exchange 2010 SP1 to SP2.

Thanks!
Larry

Re: Exchange 2010 SP2 and the Exchange MP v14.2.71.0

Shaun Collins — Thu, 24 May 2012 23:11:34 GMT

Larry, there has been no update to the Exchange 2010 MP for SP2 to my knowledge, but I would think this should not be a big deal since a service pack does not usually bring new features, but patches. I would suggest simply running in test and moving forward based on results.

Re: Core MP version for CU5

curtmcgirt — Thu, 24 May 2012 21:42:27 GMT

thanks Pete. i should read that before tomorrow, and maybe i'll ask him directly; he and Cameron are scheduled to be at the inaugural DFW User Group meeting tomorrow. but i didn't see your name on the list, sir.

Core MP version for CU5

curtmcgirt — Wed, 07 Mar 2012 01:22:10 GMT

i remember installing cu5. pretty vividly. the Agents By Version view tells me i did so also. but my System Center Core Library MP is still 6.1.7221.61. shouldn't it be .81?

according to this it should be http://support.microsoft.com/kb/2495674

"When a new management pack is created after CU4 or CU5 is installed and then an attempt is made to edit the management pack in the Authoring console, the Authoring console cannot edit the management pack because it cannot find the latest version of the Microsoft.SystemCenter.Library management pack (build .61 for CU4 and build .81 for CU5)."

my ws-management library, web application monitoring library, and microsoft generic report library are all .81, but those are the ones that get manually imported during the upgrade. was my Microsoft.SystemCenter.Library supposed to get upgraded too? how did it not?

Re: Core MP version for CU5

Pete Zerger — Thu, 24 May 2012 21:21:53 GMT

Best resource is likely Kevin Holman's recently posted "OpsMgr 2007 CU6 - My Experience' article at http://blogs.technet.com/b/kevinholman/archive/2012/05/18/opsmgr-2007-r2-cu6-rollup-hotfix-ships-and-my-experience-installing-it.aspx

Re: Backup metadata enumeration failed - sharepoint error

Tommy Gunn — Thu, 24 May 2012 21:03:35 GMT

Sameer, kind of an obscure issue I think. I've seen this error in my environment and the iwalkthru.com was helpful. Also see this one - http://www.lennaerts.com/?p=41

Backup metadata enumeration failed - sharepoint error

Sameer — Tue, 22 May 2012 17:17:14 GMT

We are using DPM2010 and Sharepoint 2010. Recently since last couple of days, we started seeing this error - "Backup metadata enumeration failed" for our Sharepoint servers. All the other backups are working fine, except the sharepoint.
We tried rebooting the server, restarting the services, trying this - http://www.itwalkthru.com/2010/10/dpm-2010-and-sharepoint-2010-metadata.html, but nothing has helped so far.

Anyone else experienced this and resolved it?
Thanks for your help in Advance
-Sameer

Re: Backup metadata enumeration failed - sharepoint error

Sameer — Thu, 24 May 2012 20:44:18 GMT

Not much DPM help on SCC? :)

Re: Core MP version for CU5

curtmcgirt — Thu, 24 May 2012 20:12:28 GMT

assuming i missed something in installing cu5, which step of installing cu6 should i pay close attention to to check the version of my system center library to make sure it was applied correctly?

Re: performance alerts for missing performance data

curtmcgirt — Thu, 24 May 2012 20:05:34 GMT

ok i finally got the overridden 2008 monitor to work on my bad agent by downloading a cpu stresser utility and making the machine run at 5% cpu for a few minutes. (the agent is grossly oversized with 16 processors that haven't been above 1% utilization for the past two weeks, save for this morning's spike. ah, "standards.")

now the question remains, why has this monitor repeatedly failed to detect 100% cpu utilization on this agent and how can i troubleshoot it any further than i have already?

performance alerts for missing performance data

curtmcgirt — Wed, 18 Apr 2012 05:17:23 GMT

this morning serverX spiked to 100% cpu usage. the scom performance collection rule only happens every 5 minutes, but the monitor checks every 2 minutes. the performance graph shows it going from 28% to 100% in one interval. but at that point (according to the performance graph) the agent somehow kept heartbeating but stopped reporting performance data. the cpu usage monitor was not triggered, and the health service unavailable monitor was not triggered, and no alerts were generated at all. the server was unusable and had to be rebooted, but zero alerts were generated in scom, and the "list top cpu consuming processes" recovery task was never run, so we have no idea what caused the issue. all these are not good things. what's a good answer to this kind of situation? i know the performance monitor and the performance collection rule are two separate entities. is there any way to see what, if anything, the monitor detected while the rule was collecting nothing?

Re: VMM2012 - Server 8 Hyperv cluster can be added but Server 8 Hyperv Host servers error

Bryan Heath — Thu, 24 May 2012 17:16:50 GMT

Well I managed to get one of my Server 8 servers to take a SMS agent after tweaking some cecurity settings. Oddly it was a clustered VM host. I may have to allow these to bake heh.

VMM2012 - Server 8 Hyperv cluster can be added but Server 8 Hyperv Host servers error

Bryan Heath — Wed, 16 May 2012 22:46:54 GMT

Have any pointers or an MP for server 8?

Error (12704)
Unable to connect to the Hyper-V service on the server VMHOST01.lab.local.

Recommended Action
Ensure that the Hyper-V services (vmms, nvspwmi, or vhdsv) and the Virtual Machine Manager Agent are installed and running on the server VMHOST01.lab.local.

Re: VMM2012 - Server 8 Hyperv cluster can be added but Server 8 Hyperv Host servers error

Bryan Heath — Thu, 24 May 2012 16:54:31 GMT

I have the OM MP working now Thanks Stan! . Has anyone had luck installing the SCCM agent on server 8 servers? I have the CTP for Server 8 and the SCCM 2012 that MSFT gave me at MMS.

Re: performance alerts for missing performance data

curtmcgirt — Thu, 24 May 2012 15:49:29 GMT

this happened again this morning on the same server. this time, however, the performance collection continued, and i can see the cpu at 100% for over an hour, and the Processor Queue length over 100 for over an hour. the thresholds for the cpu monitor are 95% and 15, respectively, over 2 or 5 samples. but the monitor did not trigger.

i did some further digging, and found that i have not received a high cpu usage alert from *any* windows 2008 agents since early november of last year. this is a couple of weeks after the latest OS MP came out, which would coincide with about the time i installed it. i am still getting cpu alerts from 2003 agents. i found a 2008 server that's averaging about 40% cpu usage, and i overrode this monitor so that the cpu threshold is 1% and the processor queue length is 1, and to trigger after one sample. the monitor has not triggered yet. i'm watching perfmon on the agent and i can see the cpu hovering around 35%.

i did the same overrides on the 2003 monitor for a 2003 server, and the monitor triggered a couple of minutes later.

so now i'm wondering if the CPU monitor is broken for 2008? but surely somebody else would have noticed this before now.

MSSQL SPN Loop

MarcoL — Thu, 24 May 2012 14:13:55 GMT

Dear All,

We have just upgraded our hosted SCOM environment to 2012 and in the process moved all databases to a new SQL 2008 R2 server. SCOM is now reporting however, that i am missing an SPN on the SQL server. No problem, i'll just add it, one would think. But after adding it, SCOM complains that i have an misplaced SPN.

hmmm, so the initial warning is:

Date and Time:	5/18/2012 10:55:46 AM
Property Name	Property Value
HasIssue	true
MissingSpnList	MSSQLSvc/DTB01.domain.local:49343
MisplacedSpnList
DuplicateSpnList

So I add the spn with "setspn -A MSSQLSvc/DTB01.domain.local:49343 dtb01"

Next thing I get:

Date and Time:	5/24/2012 3:00:45 PM
Property Name	Property Value
HasIssue	true
MissingSpnList
MisplacedSpnList	MSSQLSvc/DTB01.domain.local:49343 - DTB01$
DuplicateSpnList

What am I doing wrong here?

All input is welcome, as long as its constructive :)

Kind regards,
Marco

A lots of: "Connection with the Operations Manager Root Management Server failed."

Jelena L — Thu, 24 May 2012 11:18:47 GMT

Hi all,

Do you have any idea what could cause the following everyday error:

Connection with the Operations Manager Root Management Server failed.

Error: A Management Pack in the Management Group has been added, upgraded, overridden, or deleted. A reconnection will be needed to detect the changes.

Number of occurrence: 1

Log Name: Application
Source: MSExchangeMonitoringCorrelation
Event ID: 717
Computer: SCOM

Thanks!

Receiving alerts after putting the server under maintenance mode.

Krish — Wed, 23 May 2012 20:44:58 GMT

Hi,

I am using SCOM 2007 in my enviroment. I am receiving alerts after putting the server under maintenance mode. Please help me to fix this.

Thanks,
Krish

Re: Receiving alerts after putting the server under maintenance mode.

lolek — Thu, 24 May 2012 07:19:32 GMT

which alerts?

maybe this article will help you:
http://support.microsoft.com/kb/942866