I have noticed a timing issue. once you click start now, it would sometimes take my server around 45 seconds before it turned pending. Are you starting the pxe boot that close after you start the task?
Is this the same machine that will work sometimes and other times not work?
These are all different machines, some machines just seem to catch while others fail and requires a reboot. I've even removed the Bare Metal server and added the device back in hoping to get it resolved. Eventually it would catch.
During my testing though, on the same machine it is not catching 100%.
I've not notice the timing of the process turning to Pending. Normally I would start the task and then start my PXE boot. I will try your wait suggestions. Will post back.
Try to take a look to c:\program files\landesk\pxe\system\pxeconfig.xml on your PXE Representative.
This should (or better, MUST) specify MAC addresses that will be automatically recognized ("LANDesk Managed Provisioning" instead of "press F8..")
Sorry to but in, I have roughly the same issue. If I open the pxeconfig.xml file it has an old MAC address and IP from an earlier task. THe core does not update this file to the MAC for the current machine. Should this clear out automatically? Also can it contain multiple addresses?
How many PXE Reps do you have deployed? If you have multiple PXE Reps, they all must be on when the job is scheduled or all Provisioning tasks will fail.
These were some of the difficulty I encountered earlier, and through the forum rectified the problem. PXE Rep must be on and working, for example: I have a PXE REP deployed to a server but that server has since been formatted or off, PXE and Provisioning will not work properly.
This problem seems to be a timing issue, and I have not had the time to test this theory out. Since it does eventually catch, it does indicate a timing between starting the task and PXEing the bare metal server.
I'm brand new to provisioning and sloggin' through all the docs trying to get it to work...
I'm having your problem 100% of the time on my first test machine - I did find on teh core server configure menu -> services -> Os Deployment
this screen has were you can pick what machines are Holding Queue... I've not see that in any of the docs and I'm wondering if that has any effect on what you area dealing with. I've just moved my PXE over and re-deployed - I'll see what happens.
I had to scale back all of my PXE Reps to just a few. When one of them dies I deleted the machine from the LANDesk database and Provisioning worked flawlessly. I now only have 3 reps running and it works great.
A couple of things I would look at are:
- Only 1 PXE rep per subnet. (Although this would not cause the issue you are having, it would cause plenty of other issues.)
- Make sure your PXE reps are on machines that are ALWAYS on.
- A powered off or unreachable rep will cause some flaky provisioning issues.
- Do not choose PXE as the default boot choice, make it one time deal. Set the local drive as the boot choice by default.
If these things do not help, export your template (Sanitize it so that no confidential info is put on the web) and upload it here so we can see if it is a template issue. We are using LANDesk provisioning very successfully and I am more than happy to help you get successful with it as well.
So if i have 30 pxe reps all on different subnets on a Core Server and some of them are turned off Could that cause grief with Provisioning.