Here is a problem I am suddenly running up against. RDP 3.00 BL20pG3 servers
I have been working with both Blade servers and RDP since the very beginning. I deliver training for both products for HP for the last 4 years, so I well versed in both products but something is driving me crazy the last week.
When trying to deploy any kind of job (script or image) to BL20pG3, I get some kind of authentication error while loading winpe during the pxe boot. I can get an IP no problem and find my pxe server at the F12 prompt early in the pxe boot stages. I can get the WINPE boot iso with no problem. It loads and loads and loads for what seems like forever (nothing wrong just regular winpe load time). Right at the end of hte winpe load when the aclient appears on screen it hangs on the authenticating connection. The aclient screen is displayed with an error stating that it could not connect as a device has stopped responding. I can open a small dos box and notice that the machine is hung trying to map an X:\ drive. I have been able to duplicate the problem on 4 different deployent servers that were otherwise working perfectly fine with diff models of servers (DL380G4 and BL30p).
Anyone seen anything like this before?? Is there an issue with the nics in the blade server?? They are the nc7781 broadcomm nics. Same ones that are in the Bl30p's that are working perfectly fine.
I know it is normally trying to map F: drive. That is why i mentioned that it was trying to go to X. I found that odd. Also, the first thing I tried to do was map F manually ---- doesnt work.
Can you map to say H: (or any other letter) with the WinPE account? If not then you may need to focus on things preventing the login. If you can map to another drive letter then something is grabbing drive F:. I think you can run explorer from the C: drive (C:\windows\explorer.exe).
I just ran into a case where a account's KVM was taking the F: for virtual media connections.
I had not tried any other drive letter. Good to know what the X drive is though. Thanks for that info. What I find strange though is that I have used several different server models and this behaviour only occurs with the BL20p G3's that I work with on a regular basis.
Thanks. I had already been given this fix from Jason McGee out of Houston, but it did not really work well at all. Now instead of all of the 20ps failing only most of them do now.
What are the other 20p's doing? Do they fail to map a drive still? Do they have an IP (or is it the standard MS IP - which means you don't have an IP)?
I have this same problem as well in RDP 3.00. The problem seems to stem from WinPE not getting a valid IP address. About 50% of the time WinPE will come up with the 169.254.xx.xx address that will not work with any valid IP addresses on the network. It is something in the Altiris WinPE that is not configured properly.
When the server PXE boots it is able to get a valid IP address from DHCP, but once WinPE loads it all changes and may not have a valid IP address.
We have our own version of WinPE and adding the Aclient to our WinPE does not produce this problem is goes away and I can boot a hundred times and never encounter this failure.
I have also looked in Altiris forums as well and there are others reporting this problem. I personally think it might be happening more on Blades than on regular servers.
This really is not good and hope HP or Altiris can get to the bottom of this issue for resolution.
I applied the hotfix at the bottom of this thread (backed up files first) and now I get more information but still no luck in obtaining an IP.
The error is...
=================================== System Error 1060 has occured.
The specified service does not exist as an installed service.
The service name is invalid. ===================================
It appears that its trying to start the "DHCP Client" service under WinPE...can anyone confirm/deny this and advise of any resolution or way of troubleshooting this further?
Check your network driver is included in your WINPE image. If not, which I believe is the case, you wouold need to add it be editing your WINPE image file.
I've just been rebuilding our new lab environment and have found myself in the same situation for the past week.
I've never ran into this problem in the past couple of years of building the server and running image jobs until this time.
I've gone through all the suggestions around this site to no avail thinking that it had to be some sort of environment problem.
In the PXE manager I edited the WinPE boot and I wound up putting the FQDN at the end of the username (username@domain.com) and it is now up and working fine.
