Kickstart across subnets

Tue Dec 27 22:32:24 UTC 2005

NOT a dumb question at all, yes each of the dhcrelay servers can 'see'
the main dhcp server, in fact when testing this setup, I would see the
DHCPREQUEST on the dhcp server coming from the dhcrelay server - when it
worked that is. I tested this by changing one of the interfaces on a
server on the same network as the dhcrelay to bootproto=dhcp and then
ifup'd the device, and after a few moments the interface would be up
with the correct IP, aqnd logs on the dhc server indicated the
communication that had occurred (DHCPDISCOVERY - DHCPREQUEST - DHCPACK)

I need to do some tcpdumps, like you said, this might help a bit. 

Thanks for the response.

This e-mail and any attachments may contain confidential information
that is legally privileged. The information is solely for the use of the
intended recipient(s); any disclosure, copying, distribution, or other
use of this information is strictly prohibited. If you have received
this e-mail in error, please notify the sender by return e-mail and
delete this message. Thank you.

-----Original Message-----
From: kickstart-list-bounces at redhat.com
[mailto:kickstart-list-bounces at redhat.com] On Behalf Of Stuart J. Browne
Sent: Tuesday, December 27, 2005 4:26 PM
To: Discussion list about Kickstart
Subject: Re: Kickstart across subnets

Dumb questions, just for clarification.

The machines running 'dhcprelay' in each of the subnets can see the
other subnets without any issue?

I'm assuming you've got shell access to those machines, have you run any
packet sniffers across the interface when they're attemping a rebuild,
looking for dhcp requests?

----- Original Message -----
From: "MW Mike Weiner (5028)" <MWeiner at ag.com>
To: "Discussion list about Kickstart" <kickstart-list at redhat.com>
Cc: "MW Mike Weiner (5028)" <MWeiner at ag.com>
Sent: Wednesday, December 28, 2005 6:42 AM
Subject: Kickstart across subnets

> Dear list readers -
>
> I am having some major headaches trying to get this to work in my
> production environment. My "design" is kickstart off a cd, to do an
> install across NFS. The network is basically comprised of 4 vlans
> (192.168.30.0/24,192.168.31.0/24,192.168.32.0/24,192.168.34.0/24) via
> nfs - and i realize you can dhcp across subnets without some other
> 'magic' so, i have a dhcrelay running on each of the subnets. Now, for
> some reason, the 3 subnets that are not native to that of the yum
server
> arent working. The subnet the kickstart box is on works, or at least
> 'seems' to work and the install goes fairly smoothly. I am
unfortunately
> 1000miles away from these servers so its hard to tell whats actually
> going on, but i created a bootable kickstart cd for the technicians
> working in our datacenter, and whenever they use it on those 3 subnets
> that fail - i see NOTHING in any logs on any of the dhcrelay servers
nor
> dhcpd server, as if the request is not even getting relayed. And i
> really cant tell what is actually going on here when it fails.
>
> I created a ks boot cd, with all the required files (i.e. vmlinuz,
> initrd.img, isolinux.bin, etc) and the key line in my isolinux.cfg is
as
> follows:
>
> label linux
>  kernel vmlinuz
>  append ksdevice=eth1 console=tty0 load_ramdisk=1 prompt_ramdisk=0
> initrd=initrd.img network ip=dhcp
> ks=nfs:192.168.34.248:/tftpboot/ks-fc2.cfg selinux=0
>
> which seems to work, anyone see an issue with this? Once, booted, it
> SHOULD go out and try to grab in IP address from either a dhcrelay or
> the main dhcpd server (depending on the subnet). Once it has an IP
> address, it is supposed to continue on and mount the nfs share
> containing the installation tree (/repo/fedora/linux/core/2/i386/os)
and
> complete in roughly 10 minutes. Below is my ks.cfg file (also called
> ks-fc2.cfg on the tftpboot server).
>
> I am sure others have done cross subnet kickstarts, though poorly
> documented, i was wondering if someone may have sometime to help me
> debug this. I have to complete these 400+ kickstarts by the end of
> January 2006.
>
> I appreciate ANY pointers and assistance you may be able to provide.
>
> Michael Weiner
>
> # Kickstart file for Automated Fedora Core 2 installs via NFS
> install
> nfs --server=192.168.34.248 --dir=/repo/fedora/linux/core/2/i386/os
> lang en_US.UTF-8
> langsupport --default en_US.UTF-8 en_US.UTF-8
> keyboard us
> skipx
> network --device eth1 --bootproto dhcp
> rootpw --iscrypted $1$.pio2FpF$qdZBN7oU0oOG6R3oV6y5X.
> firewall --disabled
> selinux --disabled
> authconfig --enableshadow --enablemd5
> timezone America/New_York
> bootloader --location=mbr --append rhgb quiet
>
> # The following is the partition information
> clearpart --linux
> part /boot --fstype ext3 --size=101
> part /disk2 --fstype ext3 --size=12289
> part / --fstype ext3 --size=12289
> part swap --size=2048
> part /weblog --fstype ext3 --size=1 --grow
>
> %packages --resolvedeps
> @ web-server
> @ mail-server
> @ dns-server
> @ dialup
> @ network-server
> @ sql-server
> @ editors
> @ admin-tools
> @ system-tools
> @ news-server
> @ smb-server
> @ authoring-and-publishing
> @ sound-and-video
> @ server-cfg
> @ graphics
> @ ftp-server
> @ development-tools
> @ engineering-and-scientific
> @ text-internet
> kernel-smp
> grub
> e2fsprogs
> -cups
> -isdn
> -pcmcia
>
> %post
> # run rpm error report
> rpm -Va --nofiles --nomd5 >> /tmp/rpm-problems.txt
>
> # Update thyself
> yum -y update
>
> # The Last Thing happening in Kickstart:
> # get a file via TFTP which indicates we are done.
> echo "get Kickstart_end" | /usr/bin/tftp 192.168.96.11
>
>
>
> _______________________________________________
> Kickstart-list mailing list
> Kickstart-list at redhat.com
> https://www.redhat.com/mailman/listinfo/kickstart-list 

_______________________________________________
Kickstart-list mailing list
Kickstart-list at redhat.com
https://www.redhat.com/mailman/listinfo/kickstart-list