1. Azure NoSQL client is delivered as a Maven POM or simple jar file

Including the library is easy for Maven devs.   Here’s the entry in the main maven repo: http://search.maven.org/#browse%7C-589510877 .  The easiest way to use and import the Java SDK for Azure is declaring a maven dependency on this POM.  If you aren’t a Maven user you can download the raw jar files from Maven or the project’s home on GitHub.

2. You can access Azure’s NoSQL storage from code running in Azure, other cloud providers, or on-prem systems.

Azure’s NoSQL product is a RESTful service delivered over HTTPS.  As a java developer the best way to use the service is using the SDK, however, the raw protocol is REST and more specifically AtomPub.  For the curious, here is the REST api documtation.  As a consequence, you may use Azure’s NoSQL api to store and retrieve data from any system with internet connectivity.  For extremely high performance, you’d want to run your app within Azure, and locate your data and app in the same datacenter (easily done in Azure with click-to-use configuration)…but if you’re able to tollerate some network hops Azure storage can be used as a data distribution hub allowing applications distributed accross the globe to share an interact with the same data store.

3. Your data stored in Azure NoSQL is available to users of other languages

Azure’s RESTful backplane is open to any computer language that can interact with sockets over HTTPS.  If you use Java to store data into Windows Azure NoSQL storage, users of other langauges can access that same data.  Today there are client SDKs for many platforms and languages including PHP, Java, Node.js, iOS, Android, and .Net.

4. Economics.  Per GB Azure NoSQL storage is an affordable place to park a large amount of semi-structured data.

Ex: 100GB and 1 million storage transactions costs $13.50 per month.  Furthermore, you don’t need to worry about the upper limit of your storage account.

5.  Windows Azure has specialized storage for files and large binary objects.

Blob storage is Azure’s solution for images, videos, virtual machine images or any other large binary file scenario.  Table storage is Azure’s solution for semi-structured data.  SQL Azure is Azure’s solution for highly structured data.  Each of these can be mixed and matched, for example, an image library having it’s meta-data (size, description, title) stored in Azure Tables and the files themselves stored in Blob storage.