With help From Geroge1421 we discovered that there was a rouge DHCP server connected to the network. It was running from a VM on a regular PC. Network admins were able to locate the port relates to the MAC address of the rouge server.

Now PXE is booting to FOG smoothly

I tired options ‘Windows 10’ as well as ‘other’. None of them worked. I even tired DD method for cloning it showed the same errors.

Hi,

I have the following setup. Thought it would be easy for everyone to understand with the below diagram than explaining in long sentences. FOG manages only OS1 (which is on the SSD) and OS2 is isolated in a separate network and managed with different tools (which cannot provide any multi disk imaging facilities.)

The SSD mounts as dev/sdc and the raid0 device /dev/md126 is created using /dev/sda and /dev/sdb (pulled these information from debug mode).

I want to capture both of these OSs and image workstations with exact same configuration. I’ve tired the following and none of them are able to capture both OSs.

1. primary disk set to /dev/md126
   a. ‘Multiple Partition Image - All disks’ - > fail with the error message | could not find any disks (/bin/fog.upload)
   b. ‘Multiple Partition Image - Single disk’ -> Capture successful. Tested deploying directly on to /dev/md126 on another workstation successfully.
2. Primary disk set to /dev/sdc
   a. ‘Multiple Partition Image - All disks’ fail with the error message | could not find any disks (/bin/fog.upload)
   b. ‘Multiple Partition Image - Single disk’ - > did not try this option
3. Primary disk left blank
   a. ‘Multiple Partition Image - All disks’ capture only /dev/sdc
   b. ‘Multiple Partition Image - Single disk’ - > did not try this option

I can capture from raid0 without any problem if I do a ‘Multi Partition - Single disk’ capture. Fog doesn’t like when we set a primary disk in ‘Multiple Partition Image - All disks’ - >capture fails with the error, could not find any disks (/bin/fog.upload). I’ve captured (and deployed) an all disk capture with a VM (with multiple OSs in them) with no raid configuration successfully.

As I understand all disk capture mode doesn’t see or ignores the md126 device. However, I can see the mdadm container loading and starting messages ( assembled with 2 drives) disks before the actual capture starts.

Can you please shed some light on this matter? How can I resolve this issue without capturing and deploying the two OSs separately?

FOG is up to date running version : 1.5.4

Thank you in advance

With help From Geroge1421 we discovered that there was a rouge DHCP server connected to the network. It was running from a VM on a regular PC. Network admins were able to locate the port relates to the MAC address of the rouge server.

Now PXE is booting to FOG smoothly

@george1421 Thanks. Starting to do that now : )

I checked PXE on multiple PCs they all get a unique IP assigned in 192.168.30.xxx range. My DHCP settings are in 172.16.1.xxx range

Server

• FOG Version: 1.4.4
• OS: Ubuntu server LTS 16.04

Client

• Service Version: hyperV/HP 800 G1
  (Same error on VM and workstation)
• OS: Windows 7

Description

DHCP is served from a separate windows server. Option 066 (ip for fog server) and 067 are (undionly.kpxe) is correctly configured. Initially IP address and option 066 are correctly passed onto the PXE boot process.

Then when ipxe configures net0 with completely different information with IPs not even in the network scope. (see the below screenshot)

I manually configured ipxe ‘config’ command and it worked fine.

DHCP is configured with a second failover server (windows). I’ve double checked there are no other proxyDHCP or a rouge DHCP server connected to the network. Could someone shed some light into this issue. Is there any thing should I manually configure in FOG server?

Thanks In advance.
K