would like to use something like a distributed hash table, where i
define in which specific nodes specific data will be put. The aim is to
see if with this mapping of data the query response time gets improved.
I would like to use memcached for a
prototype. I think the servers operate independently and the clients
determine how to distribute the keys.
What client do you suggest and how much memory should i use in the server to not evict my data?
thanks a lot!