After handing over all the lists on the application and current client side infrastructure, the first question that came is: “ok I see a lot of applications listed that are only installed on one machine, can you give us the machinenames?”.

ehm well, it should be possible…

in the ACT GUI you can find the computername if you click on the application, but I had to find a way to automate this

So I installed the SQL management studio express and dived into the ACT database.

Here is the SQL query to give you all applications installed on only 1 machine:

SELECT DISTINCT
Application_Computer_Count.computerCount,
Application_Computers_Query.machineName,
Application_Report_Complete_vw.AppName,
Application_Report_Complete_vw.AppID

FROM         Application_Computer_Count INNER JOIN
Application_Report_Complete_vw ON Application_Computer_Count.appID = Application_Report_Complete_vw.AppID INNER JOIN
Application_Computers_Query ON Application_Report_Complete_vw.AppID =Application_Computers_Query.appID

WHERE     (Application_Computer_Count.computerCount= 1)

ORDER BY
Application_Report_Complete_vw.AppName

What is more interesting is the following query, which will give you all applications installed on 1 pc marked with 1 or more issues:

SELECT DISTINCT

App_Issue_Counts.appID,
App_Issue_Counts.activeIssuesCount, Application_Computer_Count.computerCount,
Application_Computers_Query.machineName,Application_Report_Complete_vw.AppName

FROM         App_Issue_Counts INNER JOIN

Application_Computer_Count ON App_Issue_Counts.appID =Application_Computer_Count.appID INNER JOIN

Application_Computers_Query ON App_Issue_Counts.appID =Application_Computers_Query.appID INNER JOIN

Application_Report_Complete_vw ON App_Issue_Counts.appID =Application_Report_Complete_vw.AppID

WHERE     (App_Issue_Counts.activeIssuesCount >=’1′) AND (Application_Computer_Count.computerCount = 1)

ORDER BY
Application_Report_Complete_vw.AppName

another thing with ACT 5.6, was that it did not seem clear to me why none of the 16 bit applications were listed in the collected data

Luckily there is a way to find them with another SQL query, checkout (and credits to): http://www.theexperienceblog.com/2011/04/14/how-to-find-16-bit-applications-in-your-act-inventory

The last thing for now on ACT is that the collector is not capable yet to work with windows 7 sp1.

you need to import a sql script to make this possible

Here is the content of this script (the import can be done via osql.exe -E -i<thenameofthesqlscript> -S”<sqlserver\instance>” -d<dbname> :

INSERT INTO.[dbo].[OS]
([osID]
,[osName]
,[majorVersion]
,[minorVersion]
,[buildNumber]
,[servicePackName]
,[servicePackMajor]
,[servicePackMinor]
,[csdVersion]
,[productType]
,[suite]
,[publishedDate])
VALUES
(N’6.1.1′
,N’Windows 7 SP1′
,6
,1
,0
,N’Service Pack 1′
,1
,0
,N’Service Pack 1′
,0
,0
,’2011-02-22′)
GO

Also remember, the infliction point to test applications during a migration is:

COST of failure X PROBABILITY of failure > COST of testing

Of course I updated my Deployment Share after the upgrade. Updating a second time, and completely regenerate the boot images got my attention that I didn’t reimported the boot images into WDS.

After reimporting the new generated boot images with MDT 2012 my Task Sequences showed up and I was able to start deploying again!

So don’t forget to reimport your WinPE boot images after upgrading to MDT 2012!

Windows could not parse or process unattend answer file [C:\MININT\unattend.xml] for pass [windowsPE]. A component or setting specified in the answer file does not exist.

According to the MS forums this should be solved in MDT 2012 beta 1. So we upgraded to MDT 2012 beta 2, but unfortunately the error remains.

Deleting the <UpgradeData> node from the unattend.xml, like Michael mentioned, fixed the issue. Windows ThinPC deployed successful when this node was removed.

This can be solved by disabling the time synchronization between the host, set the date to one in the past, and restart de SQL services afterwards.

net stop vmictimesync
date 24/2/2011
net start mssql$pocserver
net start mssql$microsoftscm
net start mssql$maps
net start vmictimesync

source: http://social.technet.microsoft.com/Forums/en-US/w7itpropoc/thread/6ccb11d4-e2cf-4af3-8df1-d2fdfb294d33

Edit: Proof of Concept Jumpstart Kit v3.0 was released, start your download managers!

But we’ve got an issue with the Reader again.

Launching a PDF file from within the bubble gave an Access Denied error (using a physical Adobe Reader 10.1):

There was an error opening the document. Access Denied

This issue exits only when the PDF file is stored on a “personal folder”, like the Desktop-folder for example. (local User NTFS permissions?).

This isn’t an App-V issue but a Reader X feature called Protected Mode that gave this error.

It can be solved with some basic OSD scripting, but the first step is here to stay: “Know your application”. As Reader X gave this error, we can disable this feature with a registry key. Most of the work is to find out what registry key to set. To keep this short, here is my OSD script that disables Adobe Reader Protected Mode when opening Reader from within a App-V application:

<REGISTRY>
<REGKEY HIVE=”HKCU” KEY=”SOFTWARE\Adobe\Acrobat Reader\10.0\Privileged” NOREDIR=”FALSE”>
<REGVALUE REGTYPE=”REG_DWORD” NAME=”bProtectedMode”>00000000</REGVALUE>
</REGKEY>
</REGISTRY>

This goes into the VIRTUALENV-tag.

Happy sequencing!

However I still had some weird scenarios, even with server OS, eg: After successfully patching a server a few days earlier, the compliance status would still report those patches as missing (although they had been installed successfully…)

This is where I made the link with the so-called WMI stability hotfixes (credits to Kim Oppalfens!  and also http://ccmexec.com/2011/08/suggested-hotfixes-for-wmi-related-issue-on-windows-platforms ).

To try and avoid these kind of problems I have a two-phase approach:

- in the Task Sequence I will install the WMI stability hotfix directly

- I created a separate Task Sequence with all stability hotfixes and have spread that on the rest of our clients.

Note : this is the list of available WMI fixes per OS

2003+XPx64: kb933061
2003 32b: kb933061
XP 32b (SP2): kb933062
Vista/server 2008: which one is it?
Server 2008 R2/W7: 981314 + 977357 + 2465990 (SP1 only)

some more info about this WMI stability hotfix TS:

* first I created a package with 5 different command lines (and 5 programs) pointing to 5 different subfolders

* then I created a TS with 5 separate run command line entries (linked to the 5 programs in the package)

Every Step then has a condition linked to the appropriate OS.

as we have no direct need to integrate MDT, I just needed the functionality of the MDT scripts in SCCM.

Start by copying the MDT scripts and put them into a package.  you do not have to create a program for this package. put it on at least one DP.

Edit the TS:

- create a ‘run command line’ step: cscript.exe ZTIgather.wsf /localonly. browse and select your MDTscripts package.

(the localonly parameter is needed or this step will fail with return code 8000)

-in another existing step I have used the variable ‘IsDesktop’ to determine when to apply the step:

See, you do not have to reinvent the wheel again !

The eventlog shows two records about this issue:

The winlogon notification subscriber <GPClient> is taking long time to handle the notification event (CreateSession).

And:

The winlogon notification subscriber <GPClient> took xxx seconds to handle the notification event (CreateSession).

I stumbled upon some possible solutions. The first was updating the NIC driver. As I’m using the wireless network card, I’m only updating this one.

Updating NIC driver

Old driver:

New driver:

Drivers are updated but the logon time wasn’t reduced by a single second.

Disable Computer GPO’s

Second try was moving the computer object to a unmanaged OU (without policies applied), run gpupdate /force and rebooted some times .. but the problem still exists.

GpNetworkStartTimeoutPolicyValue

Third try: adding GpNetworkStartTimeoutPolicyValue with value 1 to the registry:

Maybe it’s a service or application on the computer itself .. ? I disabled some 3rd party (non Microsoft) services, but the problem remains.

Disable IPv6?

–> no luck

Disable CardReader?

–> no effect

Disable Autotuning?

netsh int tcp set global autotuninglevel=disabled

Didn’t solved it either, so I enabled Autotuning again:

netsh int tcp set global autotuninglevel=normal

Patches/Hotfixes?

KB2617858: Unexpectedly slow startup or logon process in Windows Server 2008 R2 or in Windows 7

KB2525332: You encounter a long logon time after you enable the "Do not automatically make redirected folders available offline" Group Policy setting in Windows 7 or in Windows Server 2008 R2

–> This hotfix fixes the slow logon issue for me! (I’m using redirected folders!)

If 1Day = 30 hours then

WScript.Echo “Stay tuned at The D-Spot for some posts about the new beta product!”

End If

http://www.microsoft.com/download/en/details.aspx?displaylang=en&id=27841