Ok, Thanks! I will give it a try!

Excellent news! I had tried various kernel and boot kernel combinations, but gave up and just used an older version of FOG. It’s nice to know we can avail ourselves of the new version and interface (and just switch boot kernels when needed). Thanks for the update!

Hi PatinMT,

FOGServer does but if you set it up as a Node it will not unless you modify the installer for node installation. Might be worth doing so if you have future setups?

Well done in solving your own issue 🙂 and posting the solution.

Are the any requirements for keeping your deployment LAN separated from your Production LAN, like multicast filtering, security, etc?

Are the machines in your production LAN already able to interact with the Fog server for client updates, snap-ins, etc?

This topic should not be in developer help, but I’ll give you some advice anyways.

You want to run your server with ProxyDHCP enabled. This detects PXE boot requests on the network, allows the computers to get an IP address from the real DHCP server but get PXE boot details from your fog server.

I have heavily modified the articles on the wiki with my experience using Ubuntu 10.04 LTS and Fog 0.32. The article title is
[SIZE=6][B][URL=‘http://www.fogproject.org/wiki/index.php?title=Using_FOG_with_an_unmodifiable_DHCP_server/_Using_FOG_with_no_DHCP_server’][SIZE=4]Using FOG with an unmodifiable DHCP server/ Using FOG with no DHCP server[/SIZE][/URL][/B][/SIZE]

there are a few notes i have taken, using the partimage program in Fog 0.32

The image must be ext2 or ext3
You can’t use extended partitions, only Logical
swap got nothing special notes

but all partitions my come after each other, and not skip blocks between them, or else the hdd fails to restore proberly.

took some tries as i used some extended partitions first, they seem to not work at all in partimage.

i heard of some people trying to replace partimage with other tools, (but never tryed it myself), for better linux support, but as normal beings, i guess we just have to wait for fog 0.33

Like kevin mentioned. If your IP phones are on a seperate VLAN they can point to different boot server than the PCs because they would be on a seperate DHCP scope…

Thank’s for the tips !!
I finally use EXT3 and everyting is working fine now.

Regards,

check the fog permissions

chown -R fog:root /tftpboot/

I fixed it by running the install script again.

OKay, let me answer my own question:

edit: /var/ipcop/dhcp/dnsmasq.local
dhcp-boot=pxelinux.0,hostname,x.x.x.x

run: restartdhcp

Hi guys,

and thank you Fernando: it’s now working!

The configuration is:

a multiboot Win7/Ubuntu 12.04 (alpha2) (anyway 11.10 is the same) deploying on a Dell Optiplex 960 Fog 0.32 (no cutomization)

The PC configuration is:
4 primary partitions sda1 and sda2 for Windows7, sda3 for swap, sda4 for linux (ext3)

The fog server configuration is 1 disk, multipartition, declaring as ostype linux
(declaring it as windows does not work)

The installations are a base Windows 7 installation, and a base Ubuntu installation
CHANGING TO GRUB-LEGACY: this was my problem: grub2 seems to be not working.

Now I’m writing some documentation, and then I’ll try to customize my Ubuntu and Windows
installation to fit my organization, but the base deploying infrastructure seems ok!

thank you again
giammy

Hi,
we deploy Centos and (K)Ubuntu sistems without problems. Are you try to upload a Ubuntu 10.xx image?
If you need deploy a ubuntu 11.xx image, you need make some changes:
*create or modify the menu.lst file
*install and configure the grub-legacy
See the next documentation (sorry only in spanish, but you can use Google traductor 🙂 )
[url]http://www.ehu.es/es/web/instalaciones/fogehu/-/wiki/main/conocimiento[/url] (see Despliegue de imágenes con Ubuntu 11.04)

It is a fairly new thing.

The problem I always had with memdisk and floppy images was that different vendors handled the pxe stacks differently. Many of my intel machines would work fine, some needed other options to work, and some wouldn’t work at all. From all the reading at the time (this has been a few years back) all the problems were because there is no good pxe standard. I haven’t used it much but gPXE (or is it iPXE now?) might be the first thing you want to load, then pass your other pxe options to it, you will probably get better results that way. From my limited knowledge of this you would probably want to chainload it (according to the ipxe.org website),

[SIZE=4][B][FONT=Lucida Grande][SIZE=18px][COLOR=#000000]Chainloading from an existing PXE ROM[/COLOR][/SIZE][/FONT][/B][/SIZE]

[FONT=Lucida Grande][COLOR=#000000][URL=‘http://ipxe.org/_detail/clipart/chain.jpeg?id=download’][COLOR=#436976][RIGHT][IMG]http://ipxe.org/_media/clipart/chain.jpeg?w=200&h=96[/IMG][/RIGHT][/COLOR][/URL]
You can chainload iPXE from an existing PXE ROM. This is useful if you have a large number of machines that you want to be able to boot using iPXE, but you do not want to reflash the network card on each individual machine.
You can build a chainloadable iPXE image using:
[SIZE=12px] make bin/undionly.kpxe[/SIZE][/COLOR][/FONT]

Good luck.