how can I troubleshoot Windows PE booting in SCCM

[anyweb]

With all Windows PE problems and errors that you may (most likely will..) encounter during SCCM OSD Deployment, you'll want to enable the F8 feature in your Windows PE boot images, which allows you to do some basic testing by giving you access to a CMD prompt within the Windows PE session

 

So how do you enable this troubleshooting feature ?

 

locate your boot images in the Operating System Deployment section of SCCM

 

right-click on your chosen boot image (for example 32 bit) and choose Properties

 

 

Select the Windows PE tab from the properties, and place a checkmark in Enable command support then click apply

 

 

Don't forget to re-distribute your boot images to their Distribution Points after making this change, and give them some time for doing so.

 

Once distributed, when you PXE boot your clients and you want to troubleshoot any Windows PE deployment issues, press F8 once Windows PE has loadedand you'll get a command prompt.

 

In addition this will keep Windows PE operational in situations where whatever error you were getting would cause Windows PE to reboot with no warnings thereby giving you a chance to read the logs and perform troubleshooting.

 

 

SMSTS Log File location

 

 

If the task sequence completes when running in the full operating system with a Configuration Manager 2007 client installed on the computer: <CCM Install Dir>\Logs

 

If the task sequence completes when running in the full operating system with no Configuration Manager 2007 client installed on the computer: %Temp%\SMSTSLOG

 

If the task sequence completes when running in Windows PE: <largest fixed partition>\SMSTSLOG

 

Note:

 

In addition, the logs for SCCM (during the Windows PE phase) are stored in X:\Windows\Temp\SMSTSLog\ and C:\SMSTSLog and C:\_SMSTaskSequence.

 

 

 

 

More often than not, the Log file you need to be reading is the smsts.log file to see what is going wrong in your deployment...

 

After Windows PE is done and Windows installation is complete you can find the logs elsewhere such as C:\Windows\System32\CCM\Logs or C:\Windows\SysWOW64\CCM\Logs for 32bit and 64bit respectively. In those directories you'll find a log file named (for example) smsts-20081125-122856.log, the actual name of the file will change based on the date/time that you do the deployment.

 

 

I've enabled Command Prompt Support, now what ?

 

Once command prompt support is enabled you've got lots of options for troubleshooting.

 

 

network testing (ipconfig/ping...) or map a share such as

 

net use h: \\servername\share

 

If the user you are trying to logon with is a DOMAIN user then use this format

 

Enter the user name for 'servername':

DOMAIN\domainuser

 

 

disk testing (diskpart...)

read LOGS to look for errors that explain why the deployment is failing... using notepad

 

 

 

 

 

Troubleshooting Links

 

List of Custom Error Codes for Configuration Manager 2007.

 

Troubleshooting tips on Technet see here

 

Troubleshooting made easy > http://blogs.msdn.co...e-easy-way.aspx

 

Sample Troubleshooting Session

 

The Below is taken from a real troubleshooting session...

During the Windows PE phase of a Windows Server 2008 Deployment I got an error

 

Task Sequence has failed with the error code (0x00000001)

 

 

In the example above it's complaining about a Validate issue in the Task Sequence deployment of Windows Server 2008,

 

looking at the log file located at X:\Windows\Temp\SMSTSLog\ztivalidate

 

I can see that it's complaining about the speed of the processor (because this is a Virtual Machine)

 

Hint: Open the log file with SMS Trace. SMS trace is one of the tools that you can find in SCCM 2007 Toolkit. You can download it from here.

Once you open the log file it will highlight the problems in Red.

 

 

so to fix this problem required editing the Validate section of the Task Sequence and removing the check for CPU speed

 

[anyweb]

The below troubleshooting errors are for my own documentation purposes only, your results may vary:-

 

Input Error: Can not find script file "x:\sms\bin\i386\%DeployRoot%\Scripts\ZTIConfigureADDS.wsf".

 

Failed to run the action: Configure ADDS.

Incorrect function. (Error: 00000001; Source: Windows)

reason: I had configured ADDS in the wrong section of the Task Sequencer, moved it down to the correct location which is StateRestore heading after the Install Software section

 

[anyweb]

PXE Client Reboots After Booting In WinPE

 

 

 

make sure the Operating System Installation Package is distributed to the distribution point

 

 

This time I checked the SMSTS.log and found a new error

 

Content location request for WIN00018:3 failed. (Code 0x80040102)

 

followed by

 

Failed to resolve selected task sequence dependencies. Code(0x80040102)

 

 

I checked ConfigMgr and found that the package was the Windows Server 2008 Deploy Operating System Installation Package,

 

I checked that it had been distributed (it had) and updated (it had)

 

it still rebooted Windows PE, so I checked ConfigMgr's package status

 

loads of failed to access errors

 

 

 

 

workaround

 

delete the package and recreate it, then redistribute it, then verify that it's status in SCCM ConfigMGr (package status) is ok,

 

 

get it's package ID

 

 

then go back into Task Sequencer *it will alert you of the fact that it's missing anyway, as it's pointing to the old now deleted package ID*

 

 

and point it to the new Operating System Install Package

 

 

[anyweb]

got this error today just after PXE booting

 

Reboot and Select proper boot device

or Insert Boot Media in selected Boot device

 

 

The error had me confused for a bit but after rebooting a few times I eventually noticed that the PXE boot prompted me to press F12 to choose boot (like a normal lite touch or RIS menu) prior to the error,

 

 

so one time I pressed F12 and it displayed the Windows Boot Manager menu with two choices, Setup Vista or Setupy Windows Server 2008 (my WDS menu basically...)

 

 

So, why didn't it automatically PXE boot into the Server deployment, the reason was simple, even though I had imported the computers settings into the OSD section of SCCM, I had made a mistake in the MAC address, so it never would 'auto f12' the pc in question and instead, defaulted to prompting me for F12 followed quickly by the error above.

 

solution, type in the correct MAC address !

 

Once done, it will PXE boot as normal and after verifying the MAC address with the one in SCCM's database will apply pxeboot.n12

 

[anyweb]

got this error last night

 

Failed to run the last action: Install 64 bit KB951116. Execution of task sequence failed.

Unknown error (Error: 80002004; Source: Unknown)

 

while trying to install software (ignore the fact that I mention a hotfix, it actually doesnt matter), it turns out it failed because i hadn't defined my regional settings in the test task sequence, meaning that the SCCM server and Client were about 8 hours apart time wise,

 

once I set the timezone in the task sequence to the same as the server I was deploying, the error disappeared and the TS completed normally.

 

[anyweb]

another one today, and really tough to resolve....

 

Installed Windows 2008 64 bit, SQL 2005sp2 (also 64bit), sccm 2007 sp1 (32bit) configured it as outlined in my own guides and attempted to pxe boot (done in a HyperV environment) the DHCP server was running on a separate DC so we first had to manually add DHCP scope options 66 and 67 (to add these options, right click on your scope and click on set predefined options...

)

 

66= Boot Server Host Name

 

 

67= Bootfile Name

 

 

I had to set option 67 to point to REMINST/SMSBoot/x64/pxeboot.com

 

 

The client booting got a PXE error which read as follows

 

PXE-T04: Access violation.

PXE-E36: Error recieved from TFTP server

 

to cut a long story short what seemed to resolve this error was as follows

 

* Stop the WDS service

 

* delete all files within windows\temp

 

* restart WDS service

 

* redistribute the SCCM boot images and update distribution points

 

refer to the following links for more info on the above error

 

https://forums.microsoft.com/TechNet/ShowPo...17&pageid=0

 

http://forums.microsoft.com/TechNet/ShowPo...17&pageid=1

 

http://www.eggheadcafe.com/software/aspnet...-pxe-error.aspx

 

https://forums.microsoft.com/TechNet/ShowPo...4&SiteID=17

 

Update, I've seen the same error on the same server since making this post and what I have noticed is the error appears when I havn't imported a computer yet (name+mac) into osd,

 

once I import the computer using computer association, I no longer see the error above.....

[dave]

once I import the computer using computer association, I no longer see the error above.....

 

i did this twice but dosnt seem to solv te problem any sugestions ?

regards

[anyweb]

sorry, it's not clear what you did, so what did you do and what error are you getting ?

 

also please raise it as a separate post in this section thanks.

[anyweb]

new one today

 

Failed to Save environment to 80070057

 

 

the Task Sequence was forcing C: to be formatted as 9GB and D: as 9GB also (running out of space on the hyperv server)

 

Windows Server 2008 requires a partition of 10gb to install... so changed the Task Sequence from c:=9gb to c:=12gb and re ran it, this time, no error...

 

fyi.

[anyweb]

another new one today, also hard disc size related (hyperV server running out of space on C:...)

 

Failed to run the last action: Format and Partition Disk. Execution of task sequence failed.

Catastrophic failure (Error: 8000FFFF; Source: Windows)

 

 

solution, make the hard disc total size in hyperV for this virtual machine = less than the size of the space available on the hyperV host (it was greater than...)

[anyweb]

had another few today

 

1.

PXE-E55 ProxyDHCP: No reply to request on port 4011

 

very simple fix, just create the X86 Windows PE boot environment for SCCM boot image and create a new DP and update the DP,

 

in other words, make sure your server has BOTH X64 and X86 boot images even if you are only deploying X64 computers...

 

2. Contacting Server (xxx.xxx.xxx.xxx) .....

 

the dots go on for ever and your SMSPXE.log will say things like Device has been accepted but nothing else to give you a clue as to what is actually wrong,

 

the solution in my case was to reset the PASSWORD in AD users and computers for the user i specified in the MDT VBS filter (SMSinstall)

 

 

once done, error gone

 

cheers

anyweb