Re: Diskless clients?
Dirk Eddelbuettel <[email protected]>
| Newsgroups | gmane.linux.cluster.openmosix.general |
|---|---|
| Message-ID | <[email protected]> |
On 13 September 2005 at 08:02, Terry Gliedt wrote:
| In a cluster where we use a 'gateway' (single node where tasks are
| initiated) and a number of 'client nodes' (no users defined, used only
| to run OM tasks), do the client nodes ever page?
|
| Right now we are running with extremely minimal linux installations
| (Debian sarge) on the client nodes. If these nodes will not page (all
| machines have 2-4GB), then I wonder about using diskless clients?
That works. My Quantian distro-on-a-live-cdrom (now dvd) (see the pages at
http://dirk.eddelbuettel.com/quantian for more) has this as one of its key
features:
i) Boot one (possibly "fat") central machine
ii) Initiate the openMosixTerminalserver
iii) Boot slaves over PXE -- machines can be diskless and cdrom/dvd-less
iv) Have openMosix work out of the box
v) Enjoy the 6.6 gigabytes (uncompresses) of quantitative software on
the Quantian dvd.
MPI/PVM etc also work, but may need some tweaking as the PXE replication
tends to create identical hostnames etc pp which you have to postprocess if
you want "straight" PVM/MPI without openMosix. I gave a recent conference
presentation showing different ways how to do distributed statistical
computing with R in that context (slides are on my web pages too).
As Peter said, CHAOS is an option too if you want to boot of the lite 6mb
(yes, mb) CHAOS images. ClusterKnoppix also works as Quantian uses the
clusterKnoppix kernel.
| This would allow even less effort to get a node "installed" (we'd use
| PXE boot to get the kernel up and running) and even better we can
| disable/remove the disk drives reducing power, heat and the occaissional
| disk failure (our largest single instance of failure).
|
| Will this work? Anyone have any advice on how to do this? I've not done
| it myself (yet :-)
Yes, this should work. Unfortunately, at this point, I don't have access to
real life clusters to do longer-term testing.
Hope this helps, Dirk
--
Statistics: The (futile) attempt to offer certainty about uncertainty.
-- Roger Koenker, 'Dictionary of Received Ideas of Statistics'
-------------------------------------------------------
SF.Net email is sponsored by:
Tame your development challenges with Apache's Geronimo App Server.
Download it for free - -and be entered to win a 42" plasma tv or your very
own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php