FOG .33b PXE Boot woes

[Wolfbane8653]

Is this the example your are using?

[QUOTE]
[SIZE=5][B][FONT=sans-serif][SIZE=19px][COLOR=#000000]dnsmasq.conf real DHCP server[/COLOR][/SIZE][/FONT][/B][/SIZE]

[FONT=sans-serif][COLOR=#000000]This acts as a normal DHCP server, passing out dhcp options 209/210 to pxelinux[/COLOR][/FONT]
[FONT=sans-serif][COLOR=#000000]How it works[/COLOR][/FONT]
[LIST]
[]Client boots and net card does a DHCP Discover, it will get a DHCP offer from dnsmasq
[]Client does a DHCP Request, which dnsmasq will ACK, telling it to get undionly.kkpxe.0 (as no option 175 “Etherboot”).
[]Client then gets undionly.kkpxe.0 via TFTP.
[]ipxe then does a DHCP Discover/offer/request/ack and dnsmasq should respond this time telling it to load pxelinux.0 via TFTP (now option 175 is set, so it no longer tries to pass unidonly.kpxe.0)
[]Passes DHCP options 209/210 passed, which pxelinux later looks at for the config path and document root.
[]pxelinux.0 loads via TFP.
[*]pxelinux then takes over boot process, all data now going via HTTP.
[/LIST]
[COLOR=#000000][FONT=monospace]enable-tftp[/FONT][/COLOR]
[COLOR=#000000][FONT=monospace]dhcp-match=ipxe,175[/FONT][/COLOR]
[COLOR=#000000][FONT=monospace]dhcp-range=192.168.2.100,192.168.2.200[/FONT][/COLOR]
[COLOR=#000000][FONT=monospace]tftp-root=/tmp/memstick/host0_part1/[/FONT][/COLOR]
[COLOR=#000000][FONT=monospace]dhcp-boot=net:#ipxe,undionly.kkpxe.0[/FONT][/COLOR]
[COLOR=#000000][FONT=monospace]dhcp-boot=pxelinux.0[/FONT][/COLOR]
[COLOR=#000000][FONT=monospace]dhcp-option-force=209,pxelinux.cfg/default[/FONT][/COLOR]
[FONT=monospace][COLOR=#000000]dhcp-option-force=210,[url]http://xxx.xxx.xx.xx:81/[/url][/COLOR][/FONT][/QUOTE]

It looks like you will may need to change dhcp-option-force=209,pxelinux.cfg/default also it seems like you may need to look into a undionly.kpxe config file.
[SIZE=4][B][FONT=sans-serif][SIZE=17px][COLOR=#000000] [/COLOR][/SIZE][/FONT][/B][/SIZE]

[QUOTE]
[SIZE=4][B][FONT=sans-serif][SIZE=17px][COLOR=#000000]pxelinux[/COLOR][/SIZE][/FONT][/B][/SIZE]

[FONT=sans-serif][COLOR=#000000]DHCP options 209/210 are use to configure pxelinux. These can either come forcibly set in a ipxe script or via the DHCP server normally.[/COLOR][/FONT]
[LIST]
[]209 = pxelinux config path, “pxelinux.cfg/default” . This relative to the ‘root’
[]210 = pxelinux ‘root’, “[URL=‘http://xxx.xxx.xx.xx:81/’][COLOR=#663366]http://xxx.xxx.xx.xx:81/[/COLOR][/URL]” . This is prefixed with HTTP (in this case on port 81) , so pxelinux will use instead of slow TFTP . It can also use FTP.
[/LIST]
[/QUOTE]

[Jaymes Driver]

The suggestion to “dhcp-option-force=209,pxelinux.cfg/default” is what I already do with the default file manually, but using the pxelinux.0 file because the undionly.kpxe file can not be found to boot from.

The problem is, with these settings it still wants to look for pxelinux.0 and not undionly.kpxe.

I am attempting to utilize the new undionly.kpxe file and ipxe system, but I am being forced to load the old pxe method to load the new pxe method. I don’t mind doing this, it gets my items working, but for the best interest of the community I would like to move to the new system using undionly.kxpe that Tom has set up.

I will try the ipxe config file

[Jaymes Driver]

Alright here is a working ltsp.conf for ipxe, delete those symlinks, make sure pxelinux.0 does not exist rename it, also rename or delete your pxelinux.cfg. Make sure your undionly.kpxe is in the root of your tftpboot folder.

Edit your ltsp.conf to look like the following

[code]

Don’t function as a DNS server:

port=0

Log lots of extra information about DHCP transactions.

log-dhcp

Dnsmasq can also function as a TFTP server. You may uninstall

tftpd-hpa if you like, and uncomment the next line:

enable-tftp

Set the root directory for files available via FTP.

tftp-root=/tftpboot

The boot filename, Server name, Server Ip Address

dhcp-boot=undionly.kpxe,x.x.x.x

rootpath option, for NFS

#dhcp-option=17,/images

kill multicast

#dhcp-option=vendor:PXEClient,6,2b

Disable re-use of the DHCP servername and filename fields as extra

option space. That’s to avoid confusing some old or broken DHCP clients.

dhcp-no-override

PXE menu. The first part is the text displayed to the user. The second is the timeout, in seconds.

#pxe-prompt=“Press F8 for boot menu”, 3

The known types are x86PC, PC98, IA64_EFI, Alpha, Arc_x86,

Intel_Lean_Client, IA32_EFI, BC_EFI, Xscale_EFI and X86-64_EFI

This option is first and will be the default if there is no input from the user.

#pxe-service=X86PC, “Boot from network”, pxelinux

A boot service type of 0 is special, and will abort the

net boot procedure and continue booting from local media.

#pxe-service=X86PC, “Boot from local hard disk”, 0

If an integer boot service type, rather than a basename is given, then the

PXE client will search for a suitable boot service for that type on the

network. This search may be done by multicast or broadcast, or direct to a

server if its IP address is provided.

pxe-service=x86PC, “Install windows from RIS server”, 1

This range(s) is for the public interface, where dnsmasq functions

as a proxy DHCP server providing boot information but no IP leases.

Any ip in the subnet will do, so you may just put your server NIC ip here.

Since dnsmasq is not providing true DHCP services, you do not want it

handing out IP addresses. Just put your servers IP address for the interface

that is connected to the network on which the FOG clients exist.

If this setting is incorrect, the dnsmasq may not start, rendering

your proxyDHCP ineffective.

dhcp-range=10.0.0.10,proxy

This range(s) is for the private network on 2-NIC servers,

where dnsmasq functions as a normal DHCP server, providing IP leases.

dhcp-range=192.168.0.20,192.168.0.250,8h

For static client IPs, and only for the private subnets,

you may put entries like this:

dhcp-host=00:20:e0:3b:13:af,10.160.31.111,client111,infinite

[/code]

Save and restart dnsmasq.

[Paul Freeman]

Awesome! After making the changes the machines are now able to boot to iPXE. HOWEVER, now it is saying it cannot find /default.ipxe. FOG is set as the owner for everything in the /tftpboot folder so i cant imagine its a permissions problem. Perhaps is there another config elsewhere that needs to be changed?

[ATTACH=full]665[/ATTACH]

[url=“/_imported_xf_attachments/0/665_IMG_20140414_120339.jpg?:”]IMG_20140414_120339.jpg[/url]

[Tom Elliott]

Make sure the /tftpboot/default.ipxe file actually exists.

[Paul Freeman]

File exists and FOG has permission to use it. The contents are pointing to the IP of the Server. It was set to Net0 so i changed that to net1 as I am using eth1 for the connections but that had no effect.

This is what my default.ipxe file looks like.
[CODE]#!ipxe
chain http://192.168.1.99/fog/service/ipxe/boot.php?mac=${net1/mac}[/CODE]

[Tom Elliott]

net net0/mac is for the client’s network interface, not the FOG Server’s network interface.

Are you sure 192.168.1.99 is being hosted properly then?

Can you tftp the default.ipxe file:
[code]tftp -i 192.168.1.99 get default.ipxe[/code]

[Paul Freeman]

I was able to download default.ipxe with that command successfully. I am thinking I will start over on my FOG config using what I have learned with this thread and see where that gets me.

[Tom Elliott]

Okay, I hope all goes well.

I don’t know why it’s not pulling the default file, but before you start all over, maybe try restarting the tftpd-hpa service.

[Paul Freeman]

I had rebooted the server machine completely with the same result. Though when i try to restart the tftpd-hpa service it hangs and does not seem to want to start it again. And after a reboot when trying to restart the service the machine hangs at shutdown. Sounds like I may have other issues anyway. Would sorta make sense, i have been all over the place trying to get this thing to work.

[Jaymes Driver]

[quote=“Paul Freeman, post: 25459, member: 23545”]I had rebooted the server machine completely with the same result. Though when i try to restart the tftpd-hpa service it hangs and does not seem to want to start it again. And after a reboot when trying to restart the service the machine hangs at shutdown. Sounds like I may have other issues anyway. Would sorta make sense, i have been all over the place trying to get this thing to work.[/quote]

I would definitely start fresh. I can confirm these settings work for a linksys router. But if you are in a production environment, adjustments will need to be made. This is the most basic setting a ltsp.conf can have to boot.

I updated the wiki.

In order to use boot commands you will need to symlink undionly.kpxe to undionly.0

[code]
cd /tftpboot
sudo ln -s undionly.kpxe undionly.0
[/code]

Then edit your ltsp.conf edit these two lines
[code]
#pxe-prompt=“Press F8 for boot menu”, 3
to
pxe-prompt=“Press F8 for boot menu”, 3

You may change the number value after we get it working. 0 will instantly boot your first command.

#pxe-service=X86PC, “Boot from network”, pxelinux
to
pxe-service=X86PC, “Boot from network”, undionly.0
[/code]

[Paul Freeman]

Hello again all.

I would like to thank you all for your hard work and dedication to the FOG project and all the support you offer everyone and to myself.

After seeing that .33 was in RC status I decided to wait for the full release. Now I have a fresh install of 1.01 on linux Mint 16 running in production using Cisco router DHCP on 1 NIC.

I have followed the instructions via the WIKI for using DNSMASQ and after completion I was able to get a PC to start to load into iPXE. It had stopped at trying to load /default.ipxe and I was not sure why, so I rebooted the server and now the machines will no longer boot to iPXE. It now stops again at [I]PXE-E32: TFTP open timeout[/I]. I have verified that both TFTPD-HPA and DNSMASQ services are running.

I do not understand why it would stop working after a reboot when nothing else has changed.

[Jaymes Driver]

[quote=“Paul Freeman, post: 28676, member: 23545”]Hello again all.

I would like to thank you all for your hard work and dedication to the FOG project and all the support you offer everyone and to myself.

After seeing that .33 was in RC status I decided to wait for the full release. Now I have a fresh install of 1.01 on linux Mint 16 running in production using Cisco router DHCP on 1 NIC.

I have followed the instructions via the WIKI for using DNSMASQ and after completion I was able to get a PC to start to load into iPXE. It had stopped at trying to load /default.ipxe and I was not sure why, so I rebooted the server and now the machines will no longer boot to iPXE. It now stops again at [I]PXE-E32: TFTP open timeout[/I]. I have verified that both TFTPD-HPA and DNSMASQ services are running.

I do not understand why it would stop working after a reboot when nothing else has changed.[/quote]

What happens if you restart the services even though they are running?

[Paul Freeman]

Restarted the service and now it will boot back into iPXE. Now it stops at [I]/default.ipxe… No such file or directory ([url]http://ipxe.org/2d12603b[/url]) [/I]This reads to me like a permissions issue since the file is in the /tftpboot directory.

[Paul Freeman]

Changed ownership of /tftpboot to fog user and group and restarted the services and no change.

[Junkhacker]

can you retrieve undionly.kpxe by tftp from another computer from command line?

[Paul Freeman]

Yes, I am able to retrieve undionly.kpxe and default.ipxe via TFTP.

[Junkhacker]

can you verify that both the undionly.kpxe and default.ipxe are in the same directory and their permissions are 644

[Paul Freeman]

Both files are in the /tftpboot folder and are set to 644. Still throwing the same error. I currently have fog as the owner of the /tftpboot directory, should i change this back to root?

[Junkhacker]

so long as the file can be read anonymously, ownership doesn’t matter