Hacker News new | ask | show | jobs
by WorldMaker 2722 days ago
CouchDB "replication" was designed to be "master-master" (or "no master" sort of) where replication is not a coordinated process but a distributed push/pull of concurrent revisions with basic conflict resolution rules.

Whereas, Mongo only built support for classic RDMS-style replication where one database serves as the master or primary, and any number of secondaries can try to keep up with it.

It's a bit like SVN versus git. Mongo when replicated still has the SVN centralized mentality, whereas CouchDB is a lot more like git.

Compare Ref: http://docs.couchdb.org/en/stable/replication/intro.html