On Jul 7, 2010, at 6:48 AM, Michael Segel wrote: > > Non DFS used tends to be logging or some other information on the disk.
It can also be the result of hadoop not properly parsing df. [ZFS pools, for example.]
On Jul 7, 2010, at 6:48 AM, Michael Segel wrote: > > Non DFS used tends to be logging or some other information on the disk.
It can also be the result of hadoop not properly parsing df. [ZFS pools, for example.]