Ok cool, I look forward to reading it :) One thing, when you say route requests to specific instances .. does that imply that node relationships can't span instances ?
e.g. say if node 1 and was connected (at depth 1 or deeper) to a node that was stored in another instance would the neo4j path/traversal api work across the split ? On Sun, Feb 20, 2011 at 11:48 PM, Jim Webber <[email protected]> wrote: > Hi JT, > > The default way of scaling Neo4j is to use the an HA cluster and route > requests (via some domain-specific strategy) to specific instances to > ensure that caches stay warm (that is "cache sharding"). > > So scaling for Neo4j is a function of having enough RAM to comfortably hold > the working data set, and if necessary creating rules to logically partition > interactions with that data set so that warm caches are normal. > > I've written this up, and as soon as I can add diagrams I'll post it to the > Neo4j blog later this week. > > Jim > _______________________________________________ > Neo4j mailing list > [email protected] > https://lists.neo4j.org/mailman/listinfo/user > _______________________________________________ Neo4j mailing list [email protected] https://lists.neo4j.org/mailman/listinfo/user
