Steven W. Carter wrote:
Now that Cobbler is getting to a more finalized and stable
configuration, we should resume development and discussion on the
Stateless Linux project.
Is there anything in particular that needs to be addressed that is
blocking further development on this? I'm extremely interested in this
project and I believe that it has a lot of value in almost all fields of
distributed computing.
https://fedorahosted.org/k12linux/
LTSP has been using small parts of stateless in narrowly defined ways
for the past year now. I'm currently using read-only root, NFS root
boot with DHCP supplied root paths.
As for the single greatest problem preventing stateless from being
successful?
Our read-only root is very problematic. We would be in significantly
better shape for all sorts of stateless efforts if we had some kind of
union mounts. We are the only major distribution without unionfs or
aufs. I am aware that neither will ever hit the upstream kernel though
so this is a challenge for us.
Other things possibly related to stateless that I'm working on:
https://bugzilla.redhat.com/show_bug.cgi?id=424591
Add NBD root dhcp mount to mkinitrd. NBD (network block device) has
been used in production in other distros for a few years now and has
been shown to increase performance by like ~40% over NFS root.
We are *very* close to "hybrid" thin clients with select local
applications working out of the box.
Warren Togami
[EMAIL PROTECTED]
_______________________________________________
Stateless-list mailing list
[email protected]
http://www.redhat.com/mailman/listinfo/stateless-list
