Subject: Re: [vserver] ramFS Host
From: "Michael S. Zick" <mszick@morethan.org>
Date: Thu, 3 Jan 2008 09:48:47 -0600

On Thu January 3 2008 09:25, Michael S. Zick wrote:
> On Thu January 3 2008 09:06, Raoul Bhatia [IPAX] wrote:
> > as it is possible to boot a whole linux system from network with no local
> > harddisk (pxe should be the keyword), i see no issue with that besides
> > being a little more complicated for updates :) (as you need to patch the
> > file on the server and re-pack it - afairc)
> > 

Sorry. I missed your point in my first reply - -
You meant the VServer set-up and configuration files in the host context - -

Good point, thanks!

The factory install reserves the last cylinder (as /dev/sda4) for a staging
area during BIOS updates.

Perhaps something similar for additions to the "base set-up and config" files;
an additional "reserved I.E: unseen by other than host context" few cylinders;
I do have auFS patched into 2.6.24-rc6 (http://aufs.sourceforge.net) for added
flexibility.

Your point that it would be a lot of trouble to rebuild the init ramFS archive
each time you wanted to tweak a firewall rule or configuration setting was 
a good one.  Thanks.

Mike
> 
> Actually, I want to make updates easier than the factory setup - -
> The machine ships with unionFS overlaying / within the init script - -
> Which protects their software install from inadvertent change - -
> It also prevents the user from making intentional changes - -
> 
> I want to convert mine from a "tech toy" into a for-real, personal computer.
> 
> > please tell us about your progress!
> > 
> 
> Progress can be followed here: http://eee-developers.com
> 
> Mike
> > cheers,
> > raoul
> > 
> > On Thu, 3 Jan 2008 08:30:42 -0600, "Michael S. Zick" <mszick@morethan.org>
> > wrote:
> > > Group,
> > > 
> > > A somewhat abstract question today ...
> > > 
> > > The usual protocol for a kernel booting with an initial ramFS (early
> > > userspace);
> > > is for the setup and whatever initialization be done in the ramFS image;
> > > then switch_root (which removes the initial ramFS contents) to the "run
> > > time"
> > > storage media (yea, everyone here knows that) ...
> > > 
> > > But the size of a truly minimal VServer host image (logging, sshd, ntpd,
> > > network tables)
> > > is not very big (target system is a mini-laptop (EeePC) with at least
> > > 0.5Gb ram) - -
> > > Why "switch_root" for the host?
> > > Why not just bring up the first (most likely only) user context as a
> > > vserver on the
> > > run-time media - let the host context continue to run in ramFS?
> > > 
> > > Technically, it looks like it should work - -
> > > Anybody tried this sort of setup?  Any practical problems encountered?
> > > Of course, this setup would hold a few Mbytes of disk buffers out of the
> > > pool,
> > > but the machine has a lot of ram, I don't think they would be missed.
> > > 
> > > Mike
> 
>