← Back to team overview

openstack team mailing list archive

Re: Enabling data deduplication on Swift


On Sun, Mar 11, 2012 at 3:49 PM, Caitlin Bestler
<Caitlin.Bestler@xxxxxxxxxxx> wrote:
> Restricting fingerprinting to blocks would make block level compares possible, but as I noted on an earlier reply
> it would *always* require that the blocks be transferred to perform the calculation. It is a lot harder to double
> Network bandwidth than to double storage.

hmm? why is it harder? It might be more expensive to add 10/40G ports
but with LAG's it's as possible...

 Deduplication that only saves disk space is leaving the larger problem
> of network bottlenecks unaddressed.

Doesn't that depend on the ratios of read vs write?
In a read tilted environment (e.g. CDN's, image stores etc), being
able to dedup at the block level in the relatively rare write case
seems a boon. The simplification this could allow - performing
localized dedup (i.e. each object server deduping just its local
storage) seems worth while.


Follow ups