Monday, January 27, 2014

SCOM event 10801 – Discovery data couldn’t be inserted to the database

 

This came up at a customer recently, generally you will get an event that contains some GUIDs that represent the HealthServiceId and DiscoveryId.

SCOMDiscoveryError

It can be troublesome to track down the exact workflow that is responsible for this error.  However, there is some PowerShell that can help:

First, take the GUID of the DiscoveryId, then, at a OpsMgr PowerShell prompt type:

[guid]$g = "a8cc450e-fdfa-9ff0-e46f-0b8a3f58a57b";
Get-SCOMDiscovery | ? {$_.Id -eq $g};
Obviously substitute your DiscoveryId in there.  The results should look a little like this:
image

Check the DisplayName property.  You can now look at the workflow in the console, or VSAE to see what it is doing.

Friday, June 21, 2013

RD Connection Broker Redirection Timeout (ConnectionRedirectTimeout) and Sysinternals

 

I was working on a VDI project for a customer this week.  The solution was for a public library, the software used by the library added a custom GINA to Windows, that is to say it replaced the default logon screen, like so:

image

A VDI solution was proposed using the pooled desktop scenario, whereby a connection broker was used to distribute the connections amongst several RD virtualisation hosts, all using a template ‘gold’ image.

Problem…

Now, because of the custom GINA, the RD connection broker would correctly redirect the connection to the Windows 7 VMs, but… would disconnect the session after around 6 minutes as there had been no login activity.  This was a problem in this scenario as the library terminals spent most of their time at the login screen, wanting for a member of the public to come up and use them.  The problem source was identified as the RD connection broker, not the VM itself by two methods:  First, RD connections directly to the Win 7 VM stayed active for ever, secondly, the following log entries were spotted on the RD connection broker:

image

What could be the solution to this problem?  I like to think I know MS products pretty well and thought that there must be a parameter for this.

Saturday, April 6, 2013

Note to self: ConfigMgr 2012 SP1 Management Point is ASP.NET v4

 

I spent far too long trying to work out how to solve this problem and felt somewhat stupid afterwards!!

I am putting this up here as a reminder that, when deploying the ConfigMgr 2012 SP1 management point role on Windows 2008 R2 – to remember to register ASP.NET v4 with IIS,  If you don’t do this, you will see something like this in your mpcontrol.log:

Call to HttpSendRequestSync failed for port 80/443 with status code 500, text: Internal Server Error

If you get this, and your site system is running Windows 2008 R2, then you might have forgotten to register ASP.NET version 4 with IIS.  Prior to SP1 this was not an issue, as the management point web application ran using .Net version 3.5.  This is actually a descendent of .Net version 2, which was the default ASP.NET option in this OS.  When .Net v4 gets installed by the site server as part of the site system’s deployment, the ASP.NET version is not changed.  With Server 2012, the default is .Net v4 so we don’t have to do this.

In order to fix this, we need to run the following command from an elevated command prompt (run as administrator):

cd /d %windir%\Microsoft.Net\Framework64\v4*

aspnet_regiis –i –enable

Hey presto

If you still have issues then try removing and re-adding the MP role..

P.S.  My excuse for getting caught out by this is that I have been working with Server 2012 for so long that it slipped my mind

Winking smile

Friday, February 22, 2013

Hyper-V: Managing Failover IP Configuration with PowerShell

So it seems that managing failover TCP/IP configuration in Hyper-V replica is a bit of a pain.  There is no VMM integration (why?) and to manage anything but a small lab style deployment you are going to need some PowerShell.

Here is a basic script that will set the failover TCP/IP configuration on a set of VMs on a cluster, then talk to another cluster and set the same configuration on the replica machine.

Tuesday, February 12, 2013

ConfigMgr 2012 SP1 Computer Associations with PowerShell

Blimey!  It’s been a while, but here is a quick PowerShell script I made to create computer associations using the new PoSH module included in ConfigMgr 2012 SP1…

This might be useful when you don’t want to import computer information into ConfigMgr, using the SDK you can create the association between computers that are already known to SCCM.

It takes input from a CSV file that must contain at least two headings:  SourceComp and DestinationComp.  The CSV file must be called machines.csv and be located in the current working directory.

Wednesday, June 27, 2012

MCSE: Private Cloud

 

Like a lot of other people, I decided to take the two new private cloud exams (71-246/71-247) whilst at MMS this year.  The exams were in beta at the time so nobody really had any idea what was in them, now they are out of beta status the exam numbers are: 70-246 and 70-247.

What a pleasant surprise it was to see the emails arrive in my inbox a few days ago saying that I had passed!  There had been a bit if a buzz going round when Kevin Holman said on his blog that his results are in and can be checked on the Prometric web site, I had done so and read that I had passed but you always think that there could be a mistake.  So when I received the emails and checked my transcript on the Microsoft site then it was a welcome relief.

Because I had already qualified as a MCITP: Server Administrator on 2008 Server then I already had met the pre-requisites for the new MCSE qualification.

 MCSE(rgb)_1460

Smile

For information on the exams, go to the following site:  http://www.microsoft.com/learning/en/us/certification/cert-private-cloud.aspx

Suffice to say they were tough, sometimes you have an idea on whether or not you have passed – well, this time I had none.  The syllabus casts a wide net and you will have to know the entire System Center 2012 product set to some degree.