News today from the database vendor Couchbasethat it is giving its customers much more flexibility when it comes to finely tuning database scalability. Before looking at the news, it’s worth taking a step back and looking at the broader database market.

In the past few years, we have seen a huge rise in the number of so-called NoSQL databases. These databases, as the name implies, model data in ways other than the traditional tabular approaches. Readers will likely know that traditional databases follow a tabular approach (rows and columns) and, while this makes them great for structure data (name, rank, serial number), it makes them less well suited for unstructured data. As the world increasingly moves to looking at data in a far broader context than before, NoSQL databases are increasingly a great tool to use.

And the number of individual NoSQL databases is huge – the Wikipedia entry shows many dozens, and this number is rising every day. And while some of these products are differentiated through the way they work (object store versus document store, for example), others would see to be a “me too” approach.

Couchbase aims to be a general-purpose NoSQL database. It comes with a lot of different things of value to different groups – developers will like Couchbase’s mobile offering, while administrators will like the built-in caching layer and auto-sharding. Couchbase has seen commercial success from brands like Disney, Neiman Marcus, and Verizon.

So being a general purpose product means Couchbase has very broad appeal, but it also means it needs to offer a high degree of flexibility. That required flexibility is part of the reason for Couchbase’s announcement today that it is adding multi-dimensional scaling to the Couchbase product. By offering multi-dimensional scaling, Couchbase is allowing customers to isolate their index, query and data services.

What this means is that customers can precisely tune the hardware for each of those three distinct operations – index, query, and data. Traditionally the same compute resource has hosted all three operations, the result of which is that they are essentially competing with each other. The impacts of that being that relational databases perform well with high query complexity and low scalability while traditional NoSQL offerings do the opposite – performing best with low query complexity and higher data scalability.

With this change Couchbase is hoping to give customers the very best of both worlds – they can provision specific query, index and data machines and hence optimize the hardware those operations are running on. For Couchbase, it ideally means that more customers will fulfill their requirements with one product (Couchbase) given its increasing flexibility to cover their different use-cases.