Ihave monitored both my posts and a half dozen others trying to resolve their PXE boot issues. One thing in common is that they tend to be piecemeal. So, I thought I would start over with one post that clearly identifies what I have done. I did the following: 1.Remove PXE role from only site server in the lab. 2.Uninstall WDS and reboot the server. 3.Delete the residual D:\RemoteInstall folder from previous PXE installation. 4.Install WDS on the site server and reboot without configuring WDS in any way. 5.Delete all PXE related logs in SMS folders (e.g. pxecontrol.log, pxesetup.log, etc.). 6.Clear the PXE related status messages in SCCM 2007 to get a clean start with logs. 7.Assign PXE role to primary server with 192.168.1.0 boundaries. 8.Check PXESetup.log for installation successful indication and no errors. 9.Check PXEMSI.log for any errors. 10.Deploy the x86 and x64 boot images to the SMSPXEIMAGES$ distribution point. 11.Verify the boot images are on both the normal and SMSPXEIMAGES$ server shares. 12.Verify the status message confirm successful boot image distribution. 13.Verify the reference image is on the normal server share. 14.Verify the physical files are deployed to D:\RemoteInstall\SMSBoot\x86 and x64. 15.Create a Task Sequence called Reference.Image.Install using the x86 boot image. 16.Advertise the Reference.Image.Install sequence to a collection called Lifecycle. 17.Add a PXE client MAC to a collection called Lifecycle using the Import Computer wizard. 18.Verify Lifecycle collection includes the new PXE client record. 19.Boot the PXE client.At this point, the client displays a series of things shown in the image at http://www.myitforum.com/forums/upfiles/12811/9E0C885127A7422595EAB26F8531269A.jpg. While monitoring the trail of periods trek across the screen, the SMSPXE.log shows "Device has been accepted" messages. Everything appears good (i.e. IP addresses, etc.) and it is clearly interacting with the server, as illustrated by the log entries. It simply does not get "approval" to proceed. Ideas?

I've seen a similar behavior, but I can't remember how I solved it I had to configure something in WDS directly (although the SCCM docs say that you do not have to touch WDS itself). Just open the 'windows deployment services' MMC and see what happens (IIRC there was a popup that pointed me in the right direction).

I have done a few interations of this. Including a couple of rebuilds from scratch. The WDS console does have an important feature that got me "further" into the boot/reimage sequence. When I set the WDS to automatically "approve" known machines and "prompt" for pending machines, things got better. When the machine was in AD, using its MAC or GUID in the "Managed Computer" tab, the WDS considered this as a "known" computer and approved. It then proceeded with the WinPE boot. The boot triggered, but it never launched the imaging process. When the machine was not in AD, it appeared in the "pending" queue in the WDS console. If I "approved" it, the client would proceed to the boot process. Again, it never launched the reimage. So, using WDS only, I have at least gotten to the WinPE boot phase. With SCCM\WDS\PXE, it gets to the same point of "Contacting Server...." It has a message and numeric identifier that looks EXACTLY like the approval message using the WDS only scenario. The only problem is, there is no "approval" mechanism that involves manual intervention. I have added the new client to the Lifecycle collection (its approval) and the logs indicate that it found the correct advertisement, but it never proceeds.

Had the same problem as the progress of PXE clients did not get past "Contacting server...". Turned out the x64 boot image package were not distributed to the DP correctly (install pending..). Removed the package from the actual DP, re-distributed it, verified status and then PXE booted my clients without problems. Did not touch the WDS console.My blog: vegardhw.net | Twitter: @vegardhw | LinkedIn: /vegardhw