The relational database is dead. Okay, that may be a stretch, but thanks in large part to the demands of massively-popular consumer internet services and sophisticated enterprise applications alike, more and more developers are finding that, for certain applications, moving away from relational databases is not only desirable, but practically necessary.

One of the biggest benefits of 'NoSQL' solutions is that many do away with schema. In other words, developers don't need to define a rigid structure for data, as they do with relational databases. When dealing with certain kinds of data, this is ideal. Additionally, for some applications, relational databases have less-than-stellar performance and can be very difficult and expensive to scale.

The good news is that developers and companies evaluating whether it makes sense to adopt a NoSQL approach have a growing number of options, many of them free and open-source solutions. Here are five of the most prominent worth considering.


Another Apache Software Foundation project, Cassandra is a distributed database that allows for decentralized data storage that is fault tolerant and has no single point of failure. In other words, "Cassandra is suitable for applications that can't afford to lose data."

Who's using it: Facebook, Twitter, Reddit, Digg


Modeled after Google's BigTable database system, Hypertable's creators aim for it to be the "open source standard for highly available, petabyte scale, database systems." In other words, Hypertable is designed for storing massive amounts of data reliably across many cheap servers.

Who's using it: Zvents, Baidu,


MongoDB is a document-oriented database that uses a JSON-style data format. It's ideal for website data storage, content management and caching applications, and can be configured for replication and high availability.

Who's using it: Foursquare,, Etsy, The New York Times


A product of the Apache Software Foundation, CouchDB is another document-oriented database that stores data in JSON format. It's ACID compliant, and like MongoDB, can be used to store data and content for websites, and to provide caching.

Who's using it: BBC, Skechers, Meebo


Memcached is an in-memory caching system ideal for storing relatively small amounts of data that applications would otherwise need to retrieve from a database. Originally created for LiveJournal, Memcached is now one of the most popular caching tools for large consumer websites.

Who's using it: Wikipedia,, Flickr, Craigslist

Patricio Robles

Published 13 April, 2011 by Patricio Robles

Patricio Robles is a tech reporter at Econsultancy. Follow him on Twitter.

2641 more posts from this author

You might be interested in

Comments (5)



Bummed that Riak was not mentioned here - solves a larger problem set than any one of these alone.

over 7 years ago


will wharfe

Tools like memcached take the strain off the relational database by writing data to memory at the same time that they are written to the dbase - so that next time someone comes along, you don't have to query the database. They don't replace the dbase.

over 7 years ago


Carlos J. Ochoa Fernandez

I am very sorry, but I am totally disagreeing with your article. The use of documental, distributed or other kind of data archives were used in the early 70.s and did not resolve the problems of modelling data and performance that happen today in the IT world. The issue is what is the expertise of web developers in the Object Oriented Data Base world, and the competences needed. If you have good people, good skills and know how in ORDB, and know how to balance those skills in the web world, you will be a winner. This for sure. In conclusion the ORDB is not dead, you have to know how to lead with.

Carlos J. Ochoa

over 7 years ago

Patricio Robles

Patricio Robles, Tech Reporter at Econsultancy


I accurately described Memcached as a caching technology. What you are caching often comes from a database, but that database doesn't necessarily have to be a relational database. Additionally, it's often used to cache data from other sources: responses from API calls, rendered pages, etc.


Is the relational database dead? Of course not. But if you think that there's no other game in town, you're not looking at what some of the companies listed in this post are doing.

I also think it's a huge mistake to assume that you can compare the technology of the 1970s to the technology of today. Not only is the technology different, the applications are. Data-rich consumer internet applications used concurrently by millions of people, for instance, did not exist in the 1970s.

over 7 years ago



I think you're pretty right in the sense that thi is the end of relational model as the only way to go. Graph databases, Key/Value store are proving very interesting answers to use cases where integrity is not problem #1. But the NewSQL approach ( VoltDB, MySQL Cluster, etc...) is providing interesting answers to the same use case ( huge scalability, transparent sharding, etc... )

over 5 years ago

Save or Cancel

Enjoying this article?

Get more just like this, delivered to your inbox.

Keep up to date with the latest analysis, inspiration and learning from the Econsultancy blog with our free Digital Pulse newsletter. You will receive a hand-picked digest of the latest and greatest articles, as well as snippets of new market data, best practice guides and trends research.