Fedora Project Considering "Stateless Linux"

Havoc Pennington writes "Red Hat developers have been working on a generic framework covering all cases of sharing a single operating system install between multiple physical or virtual computers. This covers mounting the root filesystem diskless, keeping a read-only copy of it cached on a local disk, or storing it on a live CD, among other cases. Because OS configuration state is shared rather than local, the project is called 'stateless Linux.' The post to fedora-devel-list is here, and a PDF overview is here."

Comments

Re:How is this different from NFSRoot ?

[babbage]

To clarify, for those who haven't used it or read the link above, NFSRoot is not the same as a thin client, and it isn't a matter of just running X sessions remotely.

At boot time, the system uses PXE to pull down a read-only root filesystem, from which the kernel is launched, and sets up a ramdisk for read-write filesystems (/var).

All applications come down over the network from the NFS server and run locally; processes that emit log data end up writing back to a central server.

If client has catastrophic hardware failure, no big deal -- reconfigure DHCP to say that foo.domain.org is dead and now a machine with a different MAC address is going to be the new foo.domain.org, and the next time that user boots with the new computer, everything (minus anything they put in their /usr/local that wasn't backed up somehow) is back without missing a beat.

If the NFS server goes down, the client machines may freeze, but in theory they should resume cleanly when the server comes back up. (In practice this does't always work, but it works a lot better than I would have expected it to -- a lot of the times, things really are just Fine.)

Stateless Linux seems to offer most of the same things that most of the Linux users at my company are already doing today with NFSRoot. The main blind spot we have with NFSRoot is a good way to handle laptop users, but at my company there aren't many of those so we can deal with that. Aside from that case, the way we use NFSroot seems to be basically identical to what RedHat is proposing with this Stateless Linux proposal.

So -- can anyone that has read up on both explain the difference? It isn't obvious to me...