centos 5 PXE install server w/kickstart issues

swanyjim · 04-23-2009, 03:28 PM

hi all,

I am trying to install a PXE install server on centos 5.2 that will also run kickstart. I have the PXE server up and running well that install manually just fine. when I try to run the kickstart it hangs while getting an IP. I have serchaed the net with no luck.

My

dhcpd.conf is:

Code:

ddns-update-style interim;
option domain-name-servers (comcast DNS servers) ;

default-lease-time 86400;
max-lease-time 604800;


authoritative;

subnet 10.1.10.0 netmask 255.255.255.0 {
        range 10.1.10.210 10.1.10.211 ;
	# enable for PXE boot
        filename "pxelinux.0";
        option subnet-mask 255.255.255.0;
        option broadcast-address 10.1.10.255;
        option routers 10.1.10.201;
        allow unknown-clients ;
        allow booting ;
        allow bootp ;


}

host computer {
hardware ethernet (client mac);
fixed-address 10.1.10.210;
}

ks.cfg is:

Code:

#platform=x86, AMD64, or Intel EM64T
# System authorization information
auth  --useshadow  --enablemd5 
# System bootloader configuration
bootloader --location=none
# Partition clearing information
clearpart --none  
# Use graphical install
#graphical
# Firewall configuration
firewall --disabled
# Run the Setup Agent on first boot
firstboot --disable
# System keyboard
keyboard us
# System language
lang en_US
# Installation logging level
logging --level=info
# Use network installation
url --url=ftp://10.1.10.201/install/centos/i386/5.2
# Network information
network --bootproto=dhcp --device=eth0 --onboot=on
# Reboot after installation
reboot
#Root password
rootpw --iscrypted $1$jIU5BNX.$yg3ZCh8oi0vRHknpzWo3b/

# SELinux configuration
selinux --disabled
# System timezone
timezone  America/New_York
# Install OS instead of upgrade
install
# X Window System configuration information
xconfig  --defaultdesktop=GNOME --depth=8 --resolution=640x480

%packages
@base

/etc/xinetd.d/tftp :

Code:

service tftp
{
	socket_type		= dgram
	protocol		= udp
	wait			= no
	user			= root,rhinstall 
	server			= /usr/sbin/in.tftpd
	server_args		= -s /tftpboot
	disable			= no 
	per_source		= 11
	cps			= 100 2
	flags			= IPv4

The output of /var/log/messages is :

Code:

Apr 21 21:38:55 server xinetd[7627]: START: tftp pid=7629 from=10.1.10.210
Apr 21 21:38:55 server xinetd[7627]: file descriptor of service tftp has been closed
Apr 21 21:38:55 server in.tftpd[7630]: tftp: client does not accept options

I am not sure where else to look however ANY help would be more than appreciated!

Thanks

kentyler · 04-23-2009, 03:40 PM

Your definition of the client is a fixed-address client and outside of the defined subnet that has the options which allow bootp and boot.

Try to remove your fixed-address or add the options you wish the fixed address to have access to within the hosts definition.

host computer {
hardware ethernet (client mac);
fixed-address 10.1.10.210;
filename "pxelinux.0";
allow booting ;
allow bootp ;
}

swanyjim · 04-23-2009, 03:54 PM

awesome thanks for the reply.

I have added it to the the config in that section. However the machine does boot over the network and begins the install(look for the init image and such) then when it moves on durring the text portion of the install is when it hangs. Would it still matter at that point?

kentyler · 04-24-2009, 07:36 AM

Check the permissions of the files and /tftpboot directory I'm almost sure they have to have full allowed permissions.

boethius · 04-26-2009, 09:26 PM

Quote:

Originally Posted by swanyjim

awesome thanks for the reply.

I have added it to the the config in that section. However the machine does boot over the network and begins the install(look for the init image and such) then when it moves on durring the text portion of the install is when it hangs. Would it still matter at that point?

Personally I don't see the point of using fixed-address statements unless of course you're using DHCP to assign permanent IPs to the device, which I assume is why you're using it.

Anyway, if it gets beyond the PXE boot screen and goes into Anaconda to try and begin the kickstart process but doesn't proceed beyond network autoconfiguration the most likely problems are:

1. There are two or more ethernet devices in the server and anaconda is detecting a different ethernet device as eth0 versus whatever the PXE boot process detected first. If you're in an environment where you're constrained to a single ethernet drop per server, then try specifying the kickstart ethernet interface; e.g.,

append ksdevice=eth0
... or ksdevice=eth1
... or ksdevice=eth2

then change your Kickstart file:

network --device eth1 ...

etc.

I'd suggest having multiple menu option lines for each ethX interface and kickstart file that specifies individual different ethernet ports.

2. If you happen to be using Cisco switches enable portfast on all non-trunking ports. Portfast will speed up the DHCP process greatly on Cisco switches.

swanyjim · 04-27-2009, 05:32 PM

hey thanks to all who replied so diligently!!
it ended up being the permissions of the directories where my installation medium was kept.

if anyone would liek a further explanation let me know.

Thanks again guys!!

stevetokyo · 07-15-2009, 04:42 PM

@swanyjim,.. I would like further explanations,.. I have give the files and folder permission to 777 but I still get the same problem.

deadeyes · 07-15-2009, 05:02 PM

Quote:

Originally Posted by stevetokyo

@swanyjim,.. I would like further explanations,.. I have give the files and folder permission to 777 but I still get the same problem.

Please give more details on what you get as error/in log files/...

What did you configured and how.

Also check that your firewall is not blocking it (I had this problem)

stevetokyo · 07-16-2009, 12:37 AM

Trying to configure PXE Setup for Client on Centos 5.3

I made lot's of mistakes,.. But now it is working. I'm still yet to configure the kickstart part,. working on it now.
These were my mistakes.
1. I put the "pxelinux.cfg/default" file in the wrong location. Because I was using two different tutorials.
2. Wrong permissions on the folders.
3. Wrong syntax on the "pxelinux.cfg/default"

I was using this tutorials http://wiki.centos.org/HowTos/PXE/PXE_Setup and from Red Hat Enterprise Linux Administration UNLEASHED page 51 by (Tammy Fox).

I went back through everything and it is working now. Thank God it was on my Lab!!!!

Thanks for your help and concerns.

deadeyes · 07-16-2009, 01:00 PM

I see that you do everything manually.
In redhat (centos) there are some packages that make you live easier.
If you install system-config-netboot this will supply you with pxeos and pxeboot and system-config-netboot
These tools you can use from the shell and system-config-netboot in graphical mode.

This url shows you everything you need to do for a PXE install over http/nfs/ftp:
http://www.redhat.com/docs/manuals/e...de/ch-pxe.html

If you want to easily create a ks.cfg file, then just install centos (RHEL) and copy the anaconda-ks.cfg as your ks.cfg. (be sure to edit if you need automatic partitioning working)

What do you mean it was on your lab?

stevetokyo · 09-01-2009, 01:57 AM

@deadeyes thanks,..the link you gave was very helpful. About my lab,.. I mean not in a live environment(my network at home). Once again Thanks. And sorry for replying this late.