It appears in my experience that, in the finish during the day, most NoSQL databases are in their core key/value stores, meaning you ought to have the ability to develop a layer that could be NoSQL database agnostic.

That layer would just use CRUD procedures (put, set, remove), but would expose more complex features, and you'd have the ability to switch with minimal effort the actual DB be it Mongo, Redis, Cassandra, etc.

Would building something similar to this have value to a lot of people, and will it already exist?


There's kundera and DataNucleus

It's difficult to construct such factor.

Like a thought experiment, I would recommend that you simply take, for instance, Redis, MongoDB and Cassandra, and design an API of these layer.

These NoSQL solutions have drastically different qualities plus they serve different reasons. Attempting to develop a common API on their behalf is much like creating a common API for SQL database, spreadsheet document, plain text file and gmail.

When you can easily develop something, it'll completely pointless.

Different needs demand different tools.

UnQL means Unstructured Query Language. This is an open query language for JSON, semi-structured and document databases.

NuoDB is definitely an elastically scalable SQL/Acidity database that utilizes a vitalOrWorth model for storage. It runs on the top of Amazon . com S3 today (in addition to standard file systems) and may support any KV store in principle. For now it's access technique is SQL, however the system could readily support other data access languages and techniques if that's a typical requirement.

Craig Morris, NuoDB Corporation.