using SCCM 2012 RC in a LAB - Part 8. Deploying Windows 7 X64

[Oilers93]

Hello,

 

When creating the Deployment there was a part about setting the clients to "Access content Directly from a DP" instead of "Download Locally when needed..". My question is, does setting the option to "Access content directly from the DP" work for multicasting or just when unicasting? I thought I had read somewhere that when Multicasting you can only use the "Download locally when needed.." deployment setting.

 

The reason I ask is that earlier in July we had been using multicast with the "download locally when needed" setting and when we would start imaging, the clients would bomb out and overall maybe 10% would successfully re-image (all that failed would have some hash check errors). Since then we've switched to unicasting and using the "access content directly" deployment setting which has been incredibly reliable (95%+ success rate) but the network takes quite a hit when imaging a larger sized lab. This isn't a huge deal right now as our students and staff are away on summer holidays but once school is back in session we can't afford to be flooding the network should we have to re-image during the day.

 

Is there something that i can do to keep the reliability from unicasting but decrease the stress on our network by using multicasting?

 

Thanks in advance for any help with this.

[Samuel Romariz]

Hi,


I need help. When I deploy SO Windows 7, using hardlink, failed.

How Can I attached file smsts.log, in this blog. Can you help me please?

Regards,

[anyweb]

if you are having hardlink issues then please raise a new topic outlining your problems, to include a log file simply click on the 'more reply options' in the bottom left corner of the post.

[keeop]

Hi All,

 

Having some OS deployment issues. Everything was working fine until I was messing about with the MBAM stuff. Bottom line, where I was just using standard port 80/443 for the SCCM/IIS web site, I am now running this web site on alternative ports. I have set these ports in the the 'Site Properties' section, along with ticking the 'Custom Website' checkbox. All standard SCCM software deployments, updates, policy updates etc. work fine. No problems there. However, when now trying to do a Windows build - either using the TS I created following this post or the HTA one, no joy. Going through the logs it seems that I'm getting a 403 error when trying to access the package on the web server, but for some reason it's still trying to use port 80 during this process. So, I'm getting something like:

 

'Error downloading file from http://myserver:80/PACKAGE/FileNameto C:\_SMSTaskSequence\FileName etc...'

 

As you can see, it is trying to use port 80 even though an alternate port has been specified in SCCM. What am I missing? Is there anywhere else I need to change this port value for this to work? My only guess is that there's something that needs to be changed in the PXE Boot Image, as at this stage, this is the only thing being 'accessed' - we're pre-SCCM client connectivity so I assume it doesn't know about the alternative port at this point. If I create a new boot image, will that use the new alternative port information? Is that what I maybe need to do?

 

Cheers.

Keeop

[keeop]

OK, I've managed to find a workaround for this......

 

Does seem to be a bug with SCCM. I haven't installed any update roll-ups yet so maybe that will fix it? Anyway, I have set the boot image to be available from the PKG$ share as well as from the standard distribution point access. This now works. The boot image is found and downloaded and after that, SCCM uses the alternative port as it should. Not too much hassle.

 

Cheers.

Keeop

[snghoshi]

Hi,

 

im successfully deploying my image to target computers.

 

i now would like to know if there's a way i can add network drivers to this image so that i can be able to deploy the same image to different computers. at the moment its only applying the image to the same computer (from which i captured the image). now i want to deploy this image to a different model.

[anyweb]

you shouldn't be capturing the image from a real computer but build (and then capture) it from a virtual computer instead, that way you keep it driver agnostic and 'clean'

 

you can then deploy that clean image to any hardware you want using customized Apply Driver package steps which are set to install drivers for that model only, for example as I do here for the Surface Pro 3.

[philljp]

Hi anyweb,

 

I don't understand why you explicitly state a username for the SMSTSUdaUsers variable on the collection? Surely that would mean every computer built from that collection would have that user linked to it?

 

Thanks

[Peter van der Woude]

That's correct. I think it's just an example to show the concept. You can even leave that variable empty to make it show a dialog box for entering a user name. For some more ideas see: http://www.petervanderwoude.nl/post/different-methods-to-set-the-user-device-affinity-for-usage-during-and-after-a-task-sequence-in-configmgr-2012/

[anyweb]

Hi anyweb,

 

I don't understand why you explicitly state a username for the SMSTSUdaUsers variable on the collection? Surely that would mean every computer built from that collection would have that user linked to it?

 

Thanks

 

yup but it's just to show you one way that you can set variables in your lab, you can set it on

• the collection
• the computer
• the task sequence itself

and instead of hard coding it as I've done in that example you can try making it more dynamic for example:

%somedomain%\%someuser%

as I do in the CM12 UEFI BitLocker FrontEnd HTA