I evaluate some of the database capabilities for solving a large volume (many billions of records, many terabytes of data), where we will do almost primary key search queries almost exclusively.
Given its capabilities, Membase( Couchbase 1.8) looks almost like a perfect fit, and some previous tests make us believe that it is very effective for our usecase. Our main problem is that, since it Couchbase 2.0looks like a completely new direction from 1.8, product specifications may change. We like Couchbase 1.8it because it does exactly what we need very well.
We do not need representations or display / reduction capabilities. Although these are good features, we don’t need them, not what we want, if they are bad for performance at all. We ruled out CouchDBmainly because of the complexity of scaling (adding / removing nodes), which, of course, is much better in Couchbase, but also because we had some problems with this disk usage.
Does anyone know of any performance measurements made for 1.8vs 2.0without taking into account all the presentation and M / R features?
Will the plug be supported 1.8? Or 1.8dead, and we just have to move on?
source
share