1

tall order this request but I am hopeful...

I have a bunch of servers (exactly 20) that each have about 2-7TB of free space, and that space is not in any danger of being used up soon. I was wondering if there is some application that I can allocate some percentage of that spare disk to and it would would manage replicating the files to be redundantly stored in various places.

I would imagine the distributed network RAID type file-system would present itself much like nfs, and require some lock manager, or centralised management server to facilitate all this or alternatively could use some multicast protocol.

I looked at a few distributed systems like GFS, but they seem to require reformatting which is not an option.

1
  • Do note that you don't need to reformat your drives to play around with GFS...it does need block devices, but you can provide this using files and the loopback driver (losetup and friends). On the other hand, there is a fair amount of configuration work necessary to get GFS going, and if the other solutions are simpler, go for it. Also, GFS isn't designed to be "resilient" (it assumes stable storage), so possibly not a reasonable solution in any case.
    – larsks
    Dec 10, 2010 at 17:23

1 Answer 1

1

It doesn't look like I can delete this question, which I would like as I have found something very similar which answers most of my question. All be it from a roundabout way as the original question is asking about windows desktops; distributed resilient storage system using existing linux spare disk space

It looks like Tahoe; http://tahoe-lafs.org/trac/tahoe-lafs

and GlusterFS; http://www.gluster.org/

are going to be my main research points to resolve this question for me.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .