Cant deploy to Lenovo T470 on v1.2.0

[Jarvis2018]

Im looking to update fog so newer hardware can boot to it.

The current system is Version: 1.2.0 (havent updated the kernel) and was installed in 2016.

We are building images fine on older hardware with Windows 10 (hp desktops and laptops) on Legacy Bios  but we now have new Lenovos to build.

Lenovo Desktop PCs boot fine and we can install base image no issue but laptops on Legacy Bios get stuck on “attempting to send inventory”.

We know virtually nothing about the linux install or the Fog install other than the passwords to connect to it.

Im wondering whats the best way to update FOG to get newer hardware to boot (like the lenovo laptops).

Im not too bothered on switching to UEFI unless I have to and if I do have to will all devices currently set to Legacy have to be changed ?

Appreciate any help…

[george1421]

I’m going to suspect that the older kernels you are running with 1.2.0 don’t support the hardware of the T470, probably the disk and network interfaces. If your T470 disk is nvme based, then 1.2.0 will flat not work for you. You will need to build up a new FOG server running 1.4.0 or later.

[Jarvis2018]

@george1421 Thanks for the reply.

The problem with building a new Fog Server is no one here has a clue about how to move the images and devices and config from the old server to the new one or even if that is possible.

Is that possible ?

[george1421]

@Jarvis2018 There is a wiki page that describes how to migrate your FOG server, make sure you read it completely before migrating: https://wiki.fogproject.org/wiki/index.php?title=Migrate_FOG

Its ideal to spin up a new server so you will have your original one still intact in case something goes wrong with the migration.

[Jarvis2018]

Will give a look.

So the only thing on the new server then is to change the IP on DHCP or is it best to change the new server to the old IP address.

[george1421]

@Jarvis2018 There is a risk when you change the FOG Host system IP address after FOG is installed. You can fix the hard coded bits in fog, but its really best of FOG IP addresses are set before fog is installed. It can either be assigned static IP address manually or through a reservation in dhcp. That part doesn’t matter.

If you want you can move the 1.2.0 fog server’s IP address (understand that will disable imaging), and then spin up a new FOG server at the old IP address of the 1.2.0 fog server. This way all of the clients will connect to the new fog server right away. With FOG clients post 1.2.0, the clients will tattoo themselves to the 1.3.0 or later FOG server. So you don’t want to reinstall 1.5.4 multiple times because the certificates will be different. 1.2.0 clinets didn’t support certificate tattooing.

[jay172]

Hi,

We had an issue with pxe booting this model, We don’t use an uefi pxe boot file. The workaround that worked for us was to turn off secure boot and turn on legacy mode, create or deploy an image. Then turn the settings back to normal.

[Sebastian Roth]

@jay172 Which FOG version do you use?

@Jarvis2018 See there is another hint for you!

[jay172]

@Sebastian-Roth

FOG Configuration
FOG Version Information
Running Version: 8179

Kernel Versions
DefaultMember
bzImage Version: 4.6.2
bzImage32 Version: 4.6.2