Hello:
Could sometbody tell me the differences between LTSP and Stateless Linux project? How they compares? In wich situations is more recomendable one type of the architecture or other?
Best regrards, -Adriano
Adriano Galano wrote:
Hello:
buenas,
Could sometbody tell me the differences between LTSP and Stateless Linux project? How they compares? In wich situations is more recomendable one type of the architecture or other?
you can get detailed information from : http://people.redhat.com/~hp/stateless/StatelessLinux.pdf http://fedora.redhat.com/projects/stateless/
Xose Vazquez Perez wrote:
Adriano Galano wrote:
Hello:
buenas,
Could sometbody tell me the differences between LTSP and Stateless Linux project? How they compares? In wich situations is more recomendable one type of the architecture or other?
you can get detailed information from : http://people.redhat.com/~hp/stateless/StatelessLinux.pdf http://fedora.redhat.com/projects/stateless/
What version of FC3 will go with current Stateless cvs? Should current FC3 release candidate work with current stateless cvs? (Instructions say to use FC3T1).
On Wed, 27 Oct 2004 18:29:23 +0200, Xose Vazquez Perez xose@wanadoo.es wrote:
Adriano Galano wrote:
Hello:
buenas,
Hello Xose (Buenas ;-)):
Could sometbody tell me the differences between LTSP and Stateless Linux project? How they compares? In wich situations is more recomendable one type of the architecture or other?
you can get detailed information from : http://people.redhat.com/~hp/stateless/StatelessLinux.pdf http://fedora.redhat.com/projects/stateless/
Yes I read it, thanks. But I have doubs about when apply LTSP or Stateless. What are the advantages of Stateless over LTSP deployment?
best regards, -Adriano
-- Hello, this is Darl McBride, and I pronounce Linux as UNIX.
On Wed, 2004-10-27 at 21:30 +0200, Adriano Galano wrote:
On Wed, 27 Oct 2004 18:29:23 +0200, Xose Vazquez Perez xose@wanadoo.es wrote:
Adriano Galano wrote:
Hello:
buenas,
Hello Xose (Buenas ;-)):
Could sometbody tell me the differences between LTSP and Stateless Linux project? How they compares? In wich situations is more recomendable one type of the architecture or other?
you can get detailed information from : http://people.redhat.com/~hp/stateless/StatelessLinux.pdf http://fedora.redhat.com/projects/stateless/
Yes I read it, thanks. But I have doubs about when apply LTSP or Stateless. What are the advantages of Stateless over LTSP deployment?
A simple distinction is that with LTSP the computer is just a dumb terminal displaying programs being run on a more powerful server. Stateless installs the OS image on the client where the programs are run. This allows a person to detach the computer from the network and still have it be usable. When he/she plugs back in the user's changes are synced back to the server. It is Stateless because root setting are never changed. A user can go to another computer, log in and come up with the same desktop they had. You can do cool things like upgrade to the latest hardware simply by plugging it into the network and booting up with a stateless Linux image. It basically gives the benefits of central management of OS images that LTSP gives while still retaining the benefits of programs running directly on users hardware. So it is the best of both thin and fat client technologies. Of course there are drawbacks such as need for powerful hardware at the client and bigger disks but with hardware prices dropping so much that is not as much of a concern.
Also Stateless Linux is much more than a project. It is a whole mentality on how Fedora is now developed. For instance a lot of the configurations tools for things like printers and network are being designed so they do not require root state.
On Wed, 2004-10-27 at 16:16 -0400, John (J5) Palmieri wrote:
A simple distinction is that with LTSP the computer is just a dumb terminal displaying programs being run on a more powerful server. Stateless installs the OS image on the client where the programs are run. This allows a person to detach the computer from the network and still have it be usable.
Don't confuse the "cached client" mode with stateless linux in general. The idea is that we treat an NFS root filesystem with only an X server installed (similar to LTSP) in the same framework as an NFS root filesystem with a full set of apps installed, or the cached client mode, or a live CD mode. The definition I would give of stateless linux in general is "sharing the same OS instance between multiple machines" (which implies the OS instance is read-only, and contains no per-machine state - those are the things that require OS changes)
Havoc
On Wednesday 27 October 2004 15:27, Havoc Pennington wrote:
On Wed, 2004-10-27 at 16:16 -0400, John (J5) Palmieri wrote:
A simple distinction is that with LTSP the computer is just a dumb terminal displaying programs being run on a more powerful server. Stateless installs the OS image on the client where the programs are run. This allows a person to detach the computer from the network and still have it be usable.
Don't confuse the "cached client" mode with stateless linux in general. The idea is that we treat an NFS root filesystem with only an X server installed (similar to LTSP) in the same framework as an NFS root filesystem with a full set of apps installed, or the cached client mode, or a live CD mode. The definition I would give of stateless linux in general is "sharing the same OS instance between multiple machines" (which implies the OS instance is read-only, and contains no per-machine state - those are the things that require OS changes)
Havoc
Sounds like the way I had Redhat8 and RedHat 9 at my library. one large read-only nfs share for /, and a 128MB ramdisk for /tmp so users could save files, that kind of thing.
devel@lists.stg.fedoraproject.org