[Linux-cluster] Re: 2 Nodes, IDE, no SAN Solution?

[Benjamin Marzinski]

On Wed, Oct 06, 2004 at 05:56:27PM +0200, Adrian Immler wrote:
> > Do you mean 2 PC's with each a separate IDE drive?
> 
> i am interested in a system with more than 2 seperate PC's each with its
> own raid system ...
> 
> some software that ties all those raids together to one big. coda? afs?
> lustre? intermezzo? pvfs? i do not need a system that serves "one share
> at the end" but a system where X PC can take part (config dependent) and
> where each server gives some defined HDD space to "the big share" and
> with some striping and mirroring. and with a general journal like
> gfs.... just a dream ? its not intended for clients, just a big "drive"
> between some servers where they share their drives to one big ...

If I'm understanding you correctly, there is no good way that I know of to
do this now. But we are working on changing that.

right now you can use gnbd or iscsi to export block devices to other nodes.
(I don't know if there are any software iscsi targets that allow you to also
use the exported block device locally. gnbd does. The unh-iscsi target might)
Every node could export a spare partition to all the other nodes. If you had
N nodes, you would have N block devices, all the same size (1 local, and N-1
remote).  If you had cluster aware mirroring, you could set up a mirror on
top of all those devices, and then you would have a consistent image on each
node.  However, to make this reliable, you can't have "the big share" as you
called it. If there is information that only exists on one device, It's likely
that the entire cluster would crash in the event of one machine going down.

As soon as cluster aware mirroring is available (it's being worked on)
This should be easily doable.

-Ben
> 
> sorry for the bad english ;]
> 
> --
> Linux-cluster mailing list
> Linux-cluster at redhat.com
> http://www.redhat.com/mailman/listinfo/linux-cluster