Topic: DPInst with DriverPacks on network repository
Just thought I'd start a post about this since I've been working on this for a little while, with limited success. Thus far, I've tried running DPInst from C:\ and I've also tried putting DPInst on the share and running that directly. Access to the share out of the equation (read-only access is set up, and appears to be working fully as intended), I'm having a problem with the AMD CPU drivers, and (though unconfirmed) nothing else.
As it is, I get a prompt from the Hardware Installation Wizard that there was a problem installing this driver and it won't be available until after the next restart. I have tried "/c /s" as the parameters and used the DPinst.xml file included, and have ALSO tried changing to using the parameters "/c /s /sh /sw /path \\<network server>\Drivers".
I have changed the DPInst.xml file to only have the <search> and <subdirectory> parameters, using "*" as the directory as per the DPInst information on Microsoft's site. That doesn't really seem to make any difference to the CPU driver though, just things I've tried.
So far, I am of the thought that if I remove the CPU drivers from network, things should work just fine.... but I didn't have one oddity in my testing that makes me question just how well all of this is working.
I tried to leave the CPU drivers on the local computer, and run DPInst (also locally) on the local path (C:\D\CPU). It had worked before in testing using all DriverPacks together, so I thought this would be a non-issue... then I got the same error message. I actually had set up two completely separate entries in ROE for DPInst, one that used a copy of DPInst.exe and DPInst.xml on C:\ and pathing to C:\D\CPU. I hadn't even gotten to the second run of DPInst (which was completely contained on the network server) when I again got the same error. It most likely was something I did that is the problem here.
Now that I'm thinking about it, I might not have tried with the XML file supplied with DPs_Base for the CPU-only set of drivers on the local machine. This I cannot say for certain, but I do think it would be a good idea for me to try that again tomorrow when I get a chance.
I'll try to keep you all up-to-date on my status on this implementation and any idiosyncrasies I run into.