Comments on: On the evolution of “non-relational databases”. Trends and Information on Big Data, New Data Management Technologies, Data Science and Innovation. Tue, 20 Dec 2016 11:47:00 +0000 hourly 1 By: Roland Bouman Fri, 18 Dec 2009 07:13:52 +0000 I'd like to add two things:

1) IMO, "NoSQL" is more than being merely non-relational. It is true that being non-relational is a defining characteristic of "NoSQL", but a very import additional characteristic that is present in all DBMS-es I heard of that are labelled "NoSQL", is the ability to participate in a cluster – i.e. these solutions are all inherently designed to build distributed datastores.

2) It's not just a matter of IO bottlenecks in retrieving data, or volume constraints in storing it in a single instance. The distributed architecture sort of naturally calls for parallel query execution, which is very important from a performance perspective.

kind regards,

Roland Bouman