On May 18, 2011, at 11:55 AM, Manik Surtani wrote:
On 18 May 2011, at 09:40, Galder Zamarreño wrote:
> I'll create a wiki to compile the details/ideas of this discussion so that we can
hash this out further.
+1. I think this can be a very good optimisation, and further still, even if rehashing
needs to happen, the logic should be able to pick "preferred" data owners -
e.g., one on the same physical machine or at least rack, as opposed to, say, one across a
WAN.
I believe right now it is just the primary data owner for each entry that does the
pushing.
The primary data owner could still do the initial key+version digest push, and then the
node decides where to get the value part from which is the part that's potentially
big.
- Manik
--
Manik Surtani
manik(a)jboss.org
twitter.com/maniksurtani
Lead, Infinispan
http://www.infinispan.org
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
--
Galder Zamarreño
Sr. Software Engineer
Infinispan, JBoss Cache