VM Client can't PXE boot
Let me just preface this by saying that I had it working before the switch was replaced. We were using a simple set up of a Linksys WRT54G with DD-WRT and a basic 10 port NetGear 10/100 switch. After switching to this Catalyst 2960, I haven’t been able to get my VM to PXE boot.
I can ping the fog server from the physical machine.
I can ping it from the VM when it comes up in Windows
I can access the fog interface from within the VM so I know it sees it.
When the PXE boot bombs out, it comes up with the PXE-E51 and M0F and once I saw a PXE-E53.
I’m not really sure what the issue could be. I’m following a guide that was hosted here some time back regarding imaging Windows 7 machines. I even tried recreating my VM to see if it was a problem with that but no dice. I’ve also created the vmware_user local account on the physical machine.
I’m at a loss here. It’s using bridged networking (suggested by the guide) so I’m not really sure what else it could be. Maybe I’m just missing something here big time.
I also saw a suggestion to turn on portfast on each interface of the switch. Unfortunately, that’s already done.
Any clues or tips?
Ben Warfield last edited by
Did you ever get this fixed? If not… I think I know what’s going on.
The new switch has port security enabled. By default, port security on Cisco switches only allows one mac address per port. The port is allowing the windows 7 machine to connect initially, but blocks your VM when it attempts to connect because the VM has a different mac address.
Try this. On your switch, enter these commands.
en (enter enable mode)
conf t (configure from the terminal)
interface f0/2 (tell it which interface you want to work with)
switchport port-security maximum 2
This tells the switch that you will have 2 mac addresses using this port. You’ll come across this command at some point in your CCNA studies :)
Let us know if this works!
It’s the licensing thing – [url]http://fogproject.org/forum/threads/how-do-you-deal-with-licensing.5560/[/url]
I’ll do what I can. Link me to the other thread.
I do have that specified and actually, I just figured this thing out. I was going through the steps of enabling SPAN on the port and as I was setting the destination port, the switch didn’t like the port security being on if I was going to monitor it. I disabled switchport security on both ports and bam, the VM boots into the PXE menu.
I don’t know that I really need this feature on anyhow but then again I’m only studying for the CCENT, i’m no network engineer :\
Thanks for the follow up and support chad. Now if you can spare some time to help me with my other issue ;)
OK. Does the WRT54G allow you to specify the Option 66/67 or next-server-name, filename that is needed to pxe boot or do you have FOG running ProxyDHCP to handle that?
Thanks, I’ll give this a look. I also found a guide on enabling SPAN/RSPAN to monitor the port. I’ll try this and post what I find.
Also, I don’t believe the FOG server is set up as a DHCP server (I believe that was an option during initial setup?) so the WRT is actually providing DHCP if I remember my setup correctly.
You might try adjusting the portfast or security settings then. The Cisco 2960 may be complaining that it has what looks to be an edge port (single device, no chance of looping) with multiple MAC addresses trying to use that port. This is just a hunch.
Yep, that’s correct.
That WRT54G is a wireless router/ap, and shouldn’t have anything at all to do with FOG. Is your Win7 machine wired to the cisco 2960 and is the FOG server wired to the 2960 also?
There’s also the Linksys WRT54G router attached to this Cisco switch. This thing is completely off of the production network so nothing else attached besides that and other workstations getting imaged.
I guess if I could draw it out it would look like this:
Linksys Router -> Cisco 2960 24 port PoE switch -> Workstations
Tell me if I’m understanding this correctly.
You have a VM inside a Windows 7 workstation, connected directly to a Cisco 2960 with no other switching/networking gear between the Workstation and the Switch (other than patch panel/patch cables)?
Here’s the error message I see when the VM is attempting to PXE boot.
%PORT_SECURITY-2-PSECURE_VIOLATION: Security Violation occurred, caused by MAC address 000c.2994.8afd on port FastEthernet0/2
It says this repeatedly.
Here’s the configuration on the port, just like every other port:
switchport mode access
switchport port-security aging time 2
switchport port-security violation restrict
switchport port-security aging type inactivity
macro description cisco-desktop
spanning-tree bpduguard enable
Are you familiar with what I should be seeing? I do notice that when the VM is attempting to get an IP, there are repeated errors regarding port security. I’m no longer in the office but I can paste the error message here when I’m back in the office. When I’m attempting to boot the VM, I see the same error message over and over.
I’m logged onto the switch on the same machine so I can see it happening.
I’d ask my networking guy but to be completely honest, I’m actually studying for CCENT/CCNA and I’d like to see if I can’t figure this out. Additionally, when I saw him configuring this switch he was using Cisco SDM (or whatever GUI you can use to configure these switches).
Thanks for your help chad.
This may have to do with portfast or spanning tree protocol. Can you ask you network guys to monitor the port while you try to pxe boot and analyze the traffic?
It is configured for a bridged network. I had also tried NAT just to see if that would make a difference.
Is your VM configured for a bridged network adapter or NAT’ed?
From everything I can tell, yes.
I have a FOG server plugged into the Cisco switch and a separate machine running Windows 7 plugged into the same switch. That W7 machine is running VMware Workstation, receiving its IP via DHCP. Everything looks good networking wise. I can also PXE boot with that very same W7 machine.
Are the FOG server and the VM on the same subnet?
We did in fact go to a managed switch. Most of the ports were configured by someone using what I believe was Cisco’s SDM.
I’m using the WRT-54G to pass DHCP to all the hosts connected on the switch if that makes any sense. It probably doesn’t make much sense from a networking standpoint but it works for every host we’ve connected to it so far aside from this VM. We’ve imaged upwards of 200+ machines from desktops to laptops to thin clients and only this VM is having issues connecting.