Sign in
 
HomeLibraryWikiLearnGalleryDownloadsSupportForumsBlogs
Resources for IT Professionals >
> 
> 
> 
MDT 2012 Refresh and Replace with one Deploymentshare

Unanswered MDT 2012 Refresh and Replace with one Deploymentshare

  • Friday, December 07, 2012 6:26 AM
     
     
    Sign In to Vote
    0

    We currently have a MDT 2012 environment that is used to perform Baremetals and Refresh Deployments. The USMT settings in the customsettings.ini is configured to use Hardlinking. Now we need to create a Replace scenario and since the user data that will be captured will need to be stored on a network share what is the best method to configure the current Deploymentshare and customsettings.ini to support a Replace and Refresh deployment? Currently the wizard for USMT is set to SkipUserData=YES. For refresh deployments we would like to keep it suppresed if possible. Now since a Replace deployment consist of using a Replace tasksequence and a New Computer TaskSequence is it required to not suppress the USMT wizard. Once a replace TS is executed on a client and the data is backed up to a Network share, and a new computer is put in its place can it locate the mig file on its own without manualy inputing the path into the data location to restore?

    thanks


     

All Replies

  • Friday, December 07, 2012 9:32 AM
     
     
    Sign In to Vote
    0

    Hi,

    In this cases I generaly add some "Set Task Sequence Variable" at the begining of my task sequences. Any variable defined here will override the ones defined in the custom settings.

    If I remember well for the "Replace" TS which is intended to backup you should define:

    UserDataLocation (set to NETWORK for a replace)

    UDShare (set to the share containing migration stores)

    UDDir ( The directory containing the migration stores )

    ScanStateArgs (if you want to customize USMT bahavior default is : /v:5 /o /c)

    For you second TS Define the following

    UserDataLocation (set to NETWORK for a replace)

    UDShare

    UDDir

    ScanStateArgs (if you want to customize USMT bahavior default is : /v:5 /c /lac)

    The tricky part is the UDDir, if you want MDT to be able to automatically restore data then you will have to provid ea value to UDDir which should be the same for the "capture" and for the "restore" TS.

    I generally use %OSDComputerName% for both, after the backup I manually rename the folder with the new computername. Not very convenient but with scripts and some time you could have a better automation.

     
  • Friday, December 07, 2012 3:07 PM
     
     
    Sign In to Vote
    0

    Set the scanstateargs and loadstateargs directly in the task sequence to overwrite default values found in the CS.ini

    • in your refresh task sequence set the args to use hardlinks, and userdatalocation to auto
    • in your replace task sequences set the args not to use hardlinks, set userdatalocation to network, and set UDDIR and UDSHARE as mponsot suggested

    David Sebban | Nelite North America | My Blog | Follow me on twitter : @davidsebban

     
  • Friday, December 07, 2012 6:00 PM
     
     
    Sign In to Vote
    0

    Hi,

    In this cases I generaly add some "Set Task Sequence Variable" at the begining of my task sequences. Any variable defined here will override the ones defined in the custom settings.

    If I remember well for the "Replace" TS which is intended to backup you should define:

    UserDataLocation (set to NETWORK for a replace)

    UDShare (set to the share containing migration stores)

    UDDir ( The directory containing the migration stores )

    ScanStateArgs (if you want to customize USMT bahavior default is : /v:5 /o /c)

    For you second TS Define the following

    UserDataLocation (set to NETWORK for a replace)

    UDShare

    UDDir

    ScanStateArgs (if you want to customize USMT bahavior default is : /v:5 /c /lac)

    The tricky part is the UDDir, if you want MDT to be able to automatically restore data then you will have to provid ea value to UDDir which should be the same for the "capture" and for the "restore" TS.

    I generally use %OSDComputerName% for both, after the backup I manually rename the folder with the new computername. Not very convenient but with scripts and some time you could have a better automation.

    mponsot,

    To make sure I understand you correctly

    For my "Replace" TS I should define the following Variables within the TS:

    UserDataLocation (set to NETWORK) - as shown above

    Do I also set the UDShare and UDDir values in the TS just like the UserDataLocation? Your screen shot only showed the one TS Variable so I wanted to make sure.  

    For my "Refresh" TS I should define the following Variables within the TS:

    UserDataLocation (set to Auto)

    UDShare (\\servername\share)

    UDDir (%ComputerName%)

    The ScanState and Loadstate args in the CS.ini are what are used regardless if it is a replace or refresh. But even if I set the TS var. is it ok to still leave the Userdatalocation, UDShare, and UDDir in the CS since the TS var will override them?

     
  • Friday, December 07, 2012 9:39 PM
     
     
    Sign In to Vote
    0

    Sorry for my screenshot I just made it to show you how to create a TS variable, it is not complete as you said.

    So yes you understood me correclty, set userdatalocation to network in both TS. And yes a variable defined in a TS overrides a variable defined in the custom settings.

    Regards,

     
  • Friday, December 07, 2012 10:15 PM
     
     
    Sign In to Vote
    0

    mponsot,

    For our current Refresh TS we want to contiune using hardlinking. The replace TS is the only one that will use a network location to store the data. And once the old client has had the replaced with a new client during the deployment is when we would need to provide the location to restore data.

    So if we have a TS var with userdatalocation to network how would that work? Or am I missing something? And would we need to set Skipuserdata=NO

    thanks


     
  • Saturday, December 08, 2012 5:32 PM
     
     
    Sign In to Vote
    0

    David,

    So I understand the TS Var for the Refresh and Replace, but when I image a new computer that needs the userdata from the one that I prev ran the replace TS on and I have SkipUserData=YES set how do I over come that? Will I need to have the SkipUserData set to equal NO? I would rather not have that wizard displayed but if it needs to be then it has to be.

    So the questions really comes down to how do I handle New Computer that needs userdata that was backed up

    thanks 


     
  • Sunday, December 09, 2012 9:33 PM
     
     
    Sign In to Vote
    0

    Hi,

    TO handle this I generaly use the following task sequences:

    One for the capture state (like the one in my screenshot)

    One for the second part of the replace with the location set to network, based on a standard TS Template but without the "capture state" part, this one will also be able to manage the "New Computer" scenario because if it fail to find a migration store to restore it will just raise a warning and continue

    And a final TS to manage the "refresh" scenario wich will use hardlink migration with the location set to auto.

    This way I can define all the TS variables I need and keep my wizard and my custom settings as simple as I want.

    To avoid to display the User Data screen you must define the UDShare  and UDDir to something that can be predictable, that why I generaly set it to %computername%, however for this example you must rename the folder containing your backup before you launch the second replace TS if the new computer have a different name thant the old. This part is rather tricky to automatize, even in sccm there is a manual operation to declare the computer association.

    Hope it helps

    Regards,