On Apr 11, 2008, at 7:43 AM, slitz wrote:
I've read in the archive that it should be possible to use any
distributed
filesystem since the data is available to all nodes, so it should be
possible to use NFS, right?
I've also read somewere in the archive that this shoud be possible...
It is possible. The performance will be much lower on large clusters,
but it will work. Just use file:///path/to/my/data/input as the input
path. It also works for output paths. Note that this assumes that the
nfs file system has consistent names across the cluster.
-- Owen