SCCM 2007 – OSD Configuration – Task Sequence WMI Query for Lenovo ThinkPads

During the imaging process, it becomes necessary to provide WMI queries within task sequences in order to differentiate between computer models. Typically, this is accomplished by utilizing the following query within a conditional statement inside a task sequence task item:

SELECT * FROM Win32_ComputerSystem WHERE Model = “[model name from WMI]”

Unfortunately, if you are using Levono products, this can have multiple values for the same computer familiy. (i.e.  the ThinkPad X61 has several different model “numbers,” none of which have the label “ThinkPad X61”.) In order to provide a query which will yield the commonly used model “name” for Lenovo Thinkpads, use the following query:

SELECT * FROM Win32_ComputerSystemProduct WHERE Version = “ThinkPad [model number]”

(i.e. SELECT * FROM Win32_ComputerSystemProduct WHERE Version “ThinkPad X200”)

It is also very handy to keep a spreadsheet of all the model numbers and corresponding model names for use in reporting CASE statements to make your reports easier for management to decipher.

SCCM 2007 – OSD Troubleshooting – Task Sequence has failed with the error code (0x80091007)

So everything is working fine and all of the sudden I receive the following error during the preliminary injection of a laptop model into my task sequence:

Task Sequence: TS: Windows XP w/SP3 – Model Injection has failed with the error code (0x80091007). For more information, please contact your system administrator or help desk operator.

So, I pick up the phone and dial myself.

As per just about every post I could find, I verified that all drivers necessary for pre-boot were injected into the boot image, that all necessary post-image drivers had been imported into SCCM and that they had properly replicated to all DP’s. No red flags there. I checked my task sequence (which for this particular purpose, only images and installs good and bad drivers… post image software layering is disabled… otherwise, it’s identical to my production TS.) No red flags there either.

Four days of repeated troubleshooting. I deleted the image package and re-created it – again waiting for replication to finish before changing the TS to reflect the new package name. No change.

Then, as I was explaining to someone why the OSI model is so important to troubleshooting network problems, a lightbulb appeared over my head…..

I replaced the hard drive and it works perfectly…….

The purpose of this post? Before you bang your head against a table long enough to form a permanent dent, pay your respects to the physical aspects of your work environment.

 

Nuff said….

SCCM 2007 – OSD Troubleshooting – SMS_PXE_SERVICE_POINT Message ID: 6314

Just a short note for those of you (like me) who are continually ensuring that you have the latest drivers working in your OSD environment – whether or not you adopt them. I have multiple task sequences devoted to model injection – primarily because drivers are being constantly updated as problems are resolved by vendors. Personally, I like to make sure that these drivers can be added to the system without causing problems elsewhere.

One issue I run into every so often occurs when I am repeatedly reimaging a specific model to test new drivers. I haven’t quite nailed down the cause of the issue, but I have figured out how to get past it so I can continue working.

Issue:  When imaging a specific machine multiple times and then leaving the machine information in SCCM overnight, regardless of whether you have cleared the last pXE advertisement, the machine will act as those this flag is not cleared. The machine can be deleted from SCCM, recreated, let the deletion sit overnight, etc – and the advertisement will not clear.

Reason:  It seems as though the cache on the WDS server is retaining the data from the machine in question and it is actually preventing the boot and subsequent imaging process from taking place. A message will be received in SMS_PXE_SERVICE_POINT status messages stating “The SMS PXE Service Point intructed device to boot normally since it has no PXE advertisement assigned.” (NOT my spelling… the message actually says “intructed”)

Resolution:  Restart the WDS service on the pXE server. This clears the cache and allows the box to pXE boot and image.

SCCM 2007 – Troubleshooting – WUAHandler.log – Error = 0xc8000438

After deploying software updates to a collection the following State ID appears on the report titled: “States 1 – Enforcement States for a deployment”

State ID – 3010006 – “Failed to install update(s)”

When reading the WUAHandler.log file of a machine contained within this report, the following entries are found:

Async searching of updates using WUAgent started.
Async searching completed.
OnSearchComplete – Failed to end search job. Error = 0xc8000438.
Scan failed with error = 0xc8000438.

When noting the time stamps on the first two lines from the above entries, the times are generally seconds apart.

I have found that this is usually an indication of a corrupt or outdated Windows Update Agent. Installing the new agent, followed by an update scan or application will usually solve the issue.

The latest Windows Update Agent can be found here: (scroll to the bottom of the pages for the manual downloads)

http://support.microsoft.com/kb/949104

SCCM 2007 – Status Messages – SMS_MP_CONTROL_MANAGER

Just a short note to align some field names. Within status message reports, you may occasionally receive messages which reflect an item referred to as a “GUID.”

For the purposes of reporting:

Status messages referring to client “GUID’s” are referring to the field name, “v_R_System.SMS_Unique_Identifier0.”

Personally, I think they should rename the message so that it states something other than GUID, since there are variables wiithin the SCCM database which are referred to as GUID – which can be somewhat confusing to those initially diving into the views.