[infinispan-dev] HotRod, ClientIntelligence and client-side key location

[Manik Surtani]

On 23 Mar 2010, at 11:04, Mircea Markus wrote:

>>> 
>>> * For addresses, use ${HASH_FUNCTION} on address.hashcode()
> not sure how CH will look in near future (are we adding virtual nodes? if so hash code on address might be useless). 
> It might be a good idea dropping this from first HR version and approach it once we are clear on how improved consistent hashing will look?


I'm hoping we won't have to go into virtual nodes.  But yeah, this part of HotRod will be in flux anyway until we have settled on a hash function.  The rest of the protocol should be stable enough to warrant work on server and client impls.

Cheers
Manik
--
Manik Surtani
manik at jboss.org
Lead, Infinispan
Lead, JBoss Cache
http://www.infinispan.org
http://www.jbosscache.org