remote vnc-installs

[Seth Vidal]

Here's the steps I took to reinstall the serverbeach servers:

0. make sure the hosts ip is allowed to access the repos on
infrastructure. and make sure puppet allows the host to connect to its
fileserver - in the fileserver.conf file in cvs

1. login to existing system as root
2. run:
 urlgrabber \
http://infrastructure.fedoraproject.org/rhel/RHEL5-x86_64/images/pxeboot/... \
/boot/vmlinuz-install

urlgrabber \
http://infrastructure.fedoraproject.org/rhel/RHEL5-x86_64/images/pxeboot/... \
/boot/initrd-install.img


3. run this command +/- ip address changes and path changes for the
kickstart config

 grubby --add-kernel=/boot/vmlinuz-install \
--args="ks=http://skvidal.fedorapeople.org/hidden/sb1.ks lang= \
devfs=nomount ksdevice=link selinux=0 ip=64.34.163.94 \
gateway=64.34.163.65 netmask=255.255.255.192 \
dns=64.34.160.92,64.34.160.76" --title="install el5" \
--initrd=/boot/initrd-install.img 

4. run 'grub' and type:
  savedefault --default=0 --once
  quit

5. reboot the system. Watch it via ping, when it starts pinging again
after the install run:

vncview hostip:1

type in the vnc password - in this case 'vncinstall'


That's it.

-sv

[John Poelstra]

seth vidal said the following on 11/19/2007 07:18 PM Pacific Time:
>  Here's the steps I took to reinstall the serverbeach servers:
>   
Thanks for posting stuff like this!  I've been trying to do a setup like this on my
home testing network for a while, but kept running into problems.

I really appreciate how the Infrastructure team publishes things like this on a regular
basis!

John

[Seth Vidal]

On Wed, 2007-11-21 at 09:07 -0800, John Poelstra wrote:
>  seth vidal said the following on 11/19/2007 07:18 PM Pacific Time:
>  > Here's the steps I took to reinstall the serverbeach servers:
>  > 
>  
>  Thanks for posting stuff like this!  I've been trying to do a setup like this on my
> home testing network for a while, but kept running into problems.
>  
>  I really appreciate how the Infrastructure team publishes things like this on a regular
> basis!
>   
vnc kickstarts are nice b/c if you want to do any step manually it lets
you have that option and you get the extra power of the graphical
installer vs the somewhat hurky text-mode installer.

I just wish there was a way via vnc to see all of the vts during the
install.

-sv

[Jesse Keating]

On Wed, 21 Nov 2007 12:22:26 -0500
seth vidal <skvidal(a)fedoraproject.org&gt; wrote:

>  I just wish there was a way via vnc to see all of the vts during the
>  install. 
There is if you're doing a virt install and exporting the guest via
vnc, but not if you're just doing anaconda via vnc.

-- 
Jesse Keating
Fedora -- All my bits are free, are yours?

[Seth Vidal]

On Wed, 2007-11-21 at 12:34 -0500, Jesse Keating wrote:
>  On Wed, 21 Nov 2007 12:22:26 -0500
>  seth vidal <skvidal(a)fedoraproject.org&gt; wrote:
>  
>  > I just wish there was a way via vnc to see all of the vts during the
>  > install.
>  
>  There is if you're doing a virt install and exporting the guest via
>  vnc, but not if you're just doing anaconda via vnc. 
right - for bare metal, not virt.

-sv

[Chris Weyl]

On Nov 19, 2007 7:18 PM, seth vidal <skvidal(a)fedoraproject.org&gt; wrote:
>  Here's the steps I took to reinstall the serverbeach servers:
>
>  0. make sure the hosts ip is allowed to access the repos on
>  infrastructure. and make sure puppet allows the host to connect to its
>  fileserver - in the fileserver.conf file in cvs
>
>  1. login to existing system as root
>  2. run:
>   urlgrabber \
>  http://infrastructure.fedoraproject.org/rhel/RHEL5-x86_64/images/pxeboot/... \
> /boot/vmlinuz-install 
I don't follow this list closely, but are we now using RHEL for Fedora
infrastructure?  Weren't we using Fedora to run the Fedora
infrastructure (for obvious reasons)?  Why the change?

                                    -Chris
-- 
Chris Weyl
Ex astris, scientia

[Seth Vidal]

On Wed, 2007-11-21 at 12:12 -0800, Chris Weyl wrote:
>  On Nov 19, 2007 7:18 PM, seth vidal <skvidal(a)fedoraproject.org&gt;
> wrote:
>  > Here's the steps I took to reinstall the serverbeach servers:
>  >
>  > 0. make sure the hosts ip is allowed to access the repos on
>  > infrastructure. and make sure puppet allows the host to connect to its
>  > fileserver - in the fileserver.conf file in cvs
>  >
>  > 1. login to existing system as root
>  > 2. run:
>  >  urlgrabber \
>  > http://infrastructure.fedoraproject.org/rhel/RHEL5-x86_64/images/pxeboot/... \
> /boot/vmlinuz-install
>  
>  I don't follow this list closely, but are we now using RHEL for Fedora
>  infrastructure?  Weren't we using Fedora to run the Fedora
>  infrastructure (for obvious reasons)?  Why the change? 
we've been using a mix based on the service and service-life-time.

-sv

[Toshio Kuratomi]

Chris Weyl wrote:
>  I don't follow this list closely, but are we now using RHEL for
> Fedora
>  infrastructure?  Weren't we using Fedora to run the Fedora
>  infrastructure (for obvious reasons)?  Why the change?
>   
We've used a mixture of RHEL and Fedora for as long as I can remember. 
In general RHEL fits our needs better because it gives us a longer 
period before it goes EOL which means we don't have to reinstal the 
servers once a year.  When we have a need for a feature that isn't 
present in RHEL we use Fedora for the service, usually with the idea 
that we'll migrate the service to RHEL when RHEL rebases to a 
sufficiently current Fedora.

-Toshio