[clug] Clustering Filesystem

Michael Cohen michael.cohen at netspeed.com.au
Tue Aug 23 11:45:39 GMT 2005


Hi List,
  I have recently installed a bunch of machines on my network. Each of
  these has a sizable hard disk on it. I am after a clustering network
  filesystem that has the following properties:

  1) The filesystem should has as little configuration as possible - it
  would be really nice if there was some way which nodes could discover
  each other automatically.

  2) Nodes should be able to go up and down arbitrarily without
  affecting the filesystem very much. For example I might choose to
  replicate each file on two nodes. If any one of these nodes is down,
  the file should be served off the node thats up. If both nodes are
  down, I am happy to lose access to the file temporarily. This kind of
  replication can provide quite a bit of redundancy - maybe insteam of
  raid?

  Is there anything out there that does this? Does anyone have any
  experience in implementing something similar? How stable is it?

  Michael.


More information about the linux mailing list