Restarting the WDS service

[Vancouverite]

I'm finding that I'm having to restart the WDS service on my SCCM servers far, far too often for my liking.

 

Initially, in my test/lab environment, I encountered that problem that I know many other people have seen where, after reimaging the same machine multiple times, eventually it fails to pxe boot (PXE Boot Aborted) and clearing the last PXE advertisement doesn't resolve this. In that scenario stopping and restarting WDS would work and after that imaging would continue as normal.

 

I put this behaviour down to the fact that the same small number of machines were being imaged and reimaged multiple times over a short period of time.

 

But now that SCCM is being used in production I'm finding exactly the same behaviour. Attempts to reimage a newly-imported computer for the first time, in a collection with all previous advertisments cleared, are still aborting until the service is stopped and restarted. This is happening with both the primary and second sites.

 

Does anyone have any thoughts as to what might be happening? I really don't want to have to have Helpdesk restart services on a server everytime that they need to image something.

[Peter van der Woude]

That is probably caused by the WDS/ PXE Cache. By default the cache is being held for 60 minutes. Take a look here for more information: http://support.microsoft.com/kb/2019640

[Vancouverite]

That is probably caused by the WDS/ PXE Cache. By default the cache is being held for 60 minutes. Take a look here for more information: http://support.microsoft.com/kb/2019640

 

The thing is, now that I'm using SCCM in production I'm seeing this problem happen on machines that haven't been reimaged for months and that have never been imaged using SCCM/WDS (they would have been imaged using ZENworks from a completely different server). I did consider changing the PXE Cache setting but had decided against it because I really couldn't see how that could be the cause. Or am I missing something?

[Vancouverite]

Clients at that site are now failing to pick up a boot image with the error:

 

PXE-T04: Access Violation

PXE-E36: Error Received from TFTP Server

 

 

DHCP is located on a different server to SCCM. Option 60 isn't configured and Option 67 is set to \smsboot\x86\wdsnbp.com

 

Does anyone have any thoughts as to what might be causing this?

[lord_hydrax]

I believe Peter is on the right track for your initial issue, check how it resolved it for me:

 

http://www.windows-noob.com/forums/index.php?/topic/3181-wds-and-sccm-pxe-booting-issue/

 

The way I see it WDS has its cache of info about the collections that it updates hourly and by making that reg change it resets every 2 minutes instead which is practically instant in most cases.

 

I was on the same boat as you for months having to always restart WDS and while it didn't bother me hugely, what DID frustrate me was that my support guys had to constantly log on to my server and restart the service.

 

As for the other issue, my first guess was DCHP was on the server but obviously that's not the case. I have definitely also had that issue before, I think it was the permissions to the RemoteInstall folder where your boot images are stored. Within this folder there is a folder called SMSIMAGES and I have added Users with read access to this. Probably situational a bit with the way I have my user accounts for SCCM setup and Authenticated Users might be more practical. Anyways maybe give that a try?