thanks for your answer. i have already read this document but, for me, is not clear the concept. What is the PXE rapresentative? i have thinked that the core server managed the pxe process. Our default agent ( already deployed to all client) not have the pxe service enabled.
What is the way for start a new client ( without the agent installed ) via PXE and allow it to contact the core server ( for example for capture the image ) ? It is enough that under Self-electing subnet service --> PXE i enable the subnet and allow the mac address of the client and then create a new task ( afater added mac address also to bare metal devices ) ?
sorry for all my questions and many thanks
I go over some of the comms-flow of the PXE process as part of this video here -- [Tech Brief On-Demand Webinar 2016] Provisioning with LANDESK Management Suite .
The PXE rep is a subnet-specific device that picks up the PXE-request in that subnet & essentially asks the Core "Hey, this guy is booting up. Does he have any Provisioning job to do? If so, I'll tell him to boot into WinPE".
There's also a few other bits to help with understanding the provisioning process in the video & the presentation PDF that should hopefully help you a little bit with making sense of things.
thank you very much for your answer. Unfortunatelly i still don't understand how i should configure the core server for the pxe. I tell you my scenario:
What i miss?
Many many thanks
Ah OK - so you don't "configure the Core" as such for PXE - it's important that it's a client that does it (among other things, needs to be a device local to that subnet).
So basics to cover then:
Does that make more sense?
A lot of the provisioning stuff can break in a single step, and the whole thing doesn't appear to work. For instance, if your PXE rep can't resolve your Core because your DNS is wonky, "none of it appears to work" ... the net-cause tends to be very binary ("it doesn't appear to work") but there's a lot of little steps that can work, and a single individual one be at fault to block you.
Hope that helps a bit (and don't worry - takes a little while to get your head around everything. One thing you're not going to be short on with EPM is 'new technologies/things to learn') .
ok, make more sense but, how elect a PXE rep?which is the procedure?
I just want to add a note here:
I recently had trouble myself and found that the local firewall on the PXE REP was blocking UDP Port 67 and there for causing a timeout. Once I opened this up, everything worked like a charm.
Ah OK - so THAT is described here -- About whats new in OS Provisioning in LANDESK 2016.3 .
In short, the process goes like this:
In terms of "election" - that's an algorithm that gets used here (again - all automatic) -- I doubt that'll be overly interesting to share, though there *IS* an override value option available should it be needed (I'd argue it's easier to just have a limited set of "reliable devices" elected to have CSEP enabled, and not have CSEP on "known unreliable" devices in the first instance).
Does that help / make sense?
Yes, make sense and now it works. thnak you. Now i have another issue: int hte PXE Rep the service Landesk PXE Service start and stop after few seconds. I have reinstalled also the machine, alwasy same issue. Any ideas?
Have a look at the logs & co mentioned here -- How to Troubleshoot Self-Electing PXE Services .
If need be, get in touch with support (since other than mentioning you're on EPM 2017, I don't know if you're on 2017.1 or 2017.3, nor what SU you're on). They may be able to get you a service update that may help you along, among other things .
Ok Thanks, now both services are stable. Another question: i have allowed only 2 mac address to start via pxe in the subnet network but if i start up another client not allowed via pxe it can start: there is a way to allowed the PXE only for some client?
I'm not entirely sure what you're asking here.
That should cover everything.