RHEL6.5 provisioning without DHCP

Thu Dec 11 22:28:33 UTC 2014

Good to see the blacklist=lpfc

We always disconnected Fibre Channel after a kickstart disk on the test
system reformatted part of the SAN :(

*Peter Davey*
 <http://www.sealand-systems.co.uk/small>
psvdavey at sealand-systems.co.uk
*Tel: +44 1903 209367*

*Gsm: +44 7802 651892*www.sealand-systems.co.uk/small

On 11 December 2014 at 21:20, Fisher, Robert L (ROB) <rob.fisher at verizon.com
> wrote:

> I've been building servers using the ISO for 6 years now and we always use
> the following setup the networking in the isolinux/isolinux.cfg file:
>
>
>
> default remoteks
>
> prompt 10
>
> timeout 100
>
> display boot.msg
>
> F1 boot.msg
>
> F2 options.msg
>
> F3 general.msg
>
> F4 param.msg
>
> F5 rescue.msg
>
> F7 snake.msg
>
> label linux
>
>   kernel vmlinuz
>
>   append initrd=initrd.img ramdisk_size=124836
>
> label localks
>
>   kernel vmlinuz
>
>   append ks=cdrom initrd=initrd.img text ramdisk_size=124836
>
> label cdbuild
>
>   kernel vmlinuz
>
>   append ks=cdrom:/ks.cfg initrd=initrd.img netwait=60 lowres
>
> label remoteks
>
>   kernel vmlinuz
>
>   append ksdevice=eth0 netwait=60 blacklist=lpfc ks=
> http://100.1.1.1/ks.cfgs/ks.cfg.hostname ip=10.1.1.100
> netmask=255.255.255.0 gateway=10.1.1.254 initrd=initrd.img
>
>
>
> The entry netwait=60 is a leftover from RHEL 5 but it hasn't caused any
> problems with our RHEL 6 builds.  This entry just give the network device
> 60 seconds to enable before it tried to start networking.
>
> We use blacklist=lpfc to make sure that the OS does not try to install on
> anything other than local devices.
>
>
>
> Good luck,
>
> Rob Fisher
>
> *From:* kickstart-list-bounces at redhat.com [mailto:
> kickstart-list-bounces at redhat.com] *On Behalf Of *Larry Brigman
> *Sent:* Thursday, December 11, 2014 1:22 PM
> *To:* Discussion list about Kickstart
> *Subject:* Re: RHEL6.5 provisioning without DHCP
>
>
>
> You can set the IP address from the boot line.  You may need to know the
> mac address of the install interface if there is more than one interface on
> the server.
>
> On Dec 11, 2014 4:11 AM, "Frantisek Krecmer" <f.krecmer at gmail.com> wrote:
>
> Hello,
>
> I have following scenario:
>
> - need to install very large number of physical servers
>
> - RHEL repository is on the network filer (url install, NFS install
> available)
>
> - there is no DHCP available (and it is not possible to have one - long
> story)
> - I know serial number and IP for each server
>
> My idea was to boot from installation image
> rhel-server-6.5-x86_64-boot.iso where I modified following:
>
> - isolinux.cfg to load kickstart file
>
> label Project
>   menu label Project
>   menu default
>   kernel vmlinuz
>   append ks=cdrom:/isolinux/ks.cfg initrd=initrd.img
>
> add kickstart file containing network configuration in %pre script (get
> serial number from dmidecode, assign IP, install from network.  + configure
> hw raid etc etc)
>
> Basically, something like this:
>
>
> install
> url --url 1.2.3.4.7/rhel-x86_64-server-6.5
> ...
>
> %include /tmp/network
>
>
> %pre
> #!/bin/sh
>
> #------------------------------------------------------------------------------#
> #                     PRE-INSTALL NETWORK SETUP        #
>
> #------------------------------------------------------------------------------#
>
> echo "network --bootproto=static --vlanid=123 --ip=1.2.3.4
> --netmask=255.255.248.0 --gateway=1.2.3.5 --nameserver=1.2.3.6
> --device=eth0" > /tmp/network
>
>
>
> But. System asks for network settings when booted.
> I understand that I need to have stage2 loaded before anaconda starts %pre
> section. So, I need it to use stage2 from iso. But it looks for stage2 on
> the network when there is url as source in kickstart file. Is there any way
> how to tell installer:
>
> - start Anaconda from CD (use local install.img), run %pre, download
> packages from url ?
>
> file copied to /tmp/ks.cfg
>
> setting up kickstart
>
> kickstart forcing text mode
>
> kickstartFromUrl
>
> results of url ks, url http://1.2.3.7/rhel-x86_64-server-6.5
>
> trying to mount CD device /dev/sr0 on /mnt/stage2
>
> drive status is CDC_DISC_OK
>
> transferring /mnt/stage2/images/install.img to /tmp
>
> mounted loopback device /mnt/runtime on /dev/loop0 as /tmp/install.img
> #Perfect
>
> Looking for updates in /mnt/stage2/images/updates.img
>
> Looking for product in /mnt/stage2/images/product.img
>
> mounted loopback device /tmp/product-disk on /dev/loop7 as
> /mnt/stage2/images/product.img
>
> umounting loopback /tmp/product-disk /dev/loop7
>
> umounting loopback /mnt/runtime /dev/loop0                      #Why??
>
> no stage2= given, assuming
> http://1.2.3.7/rhel-x86_64-server-6.5/images/install.img   #you have it
> in /tmp already!
>
> setting language to en_US.UTF-8
>
> starting STEP_METHOD
>
> need to set up networking
> ...
>
> Do you have any idea if it possible to somehow change this stage2
> behavior? I found Anaconda boot options but I am unable to set stage2
> either by stage2= or inst.stage2=.
>
> (Any other idea how to set IP in ks.cfg and use it for packages
> downloading without DHCP would be also appreciated.)
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> _______________________________________________
> 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
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/kickstart-list/attachments/20141211/ca1e7112/attachment.htm>