On 10/16/2012 01:39 PM, Peter Robinson wrote:
one may say "disk storage is nothing these days"
iw ould say: mulitply it with 20, 50, 100 virtual machines
on really expensive SAN-storage where "disk space is cheap"
is not true
And I would say : get an entreprisey deduping san
for production under load not really a good decision
even if, save ressources is always a good idea over
the long
That depends. The process of actually de-duping is some what intensive
depending on the process but if you clone off a template it's not and
you get other advantages like improved use of cache and less I/O on
the disk array so it's very dependent. I generally see more
improvement than loss.
Moreover, I was under the impression that BTRFS is going to do de-duping
internally and transparently, because it keeps block checksums anyway so
detecting duplicate data comes cheap.
--
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxxx
https://admin.fedoraproject.org/mailman/listinfo/devel
