Cassandra NoSQL Database Lands on HP Moonshot Servers

Technology disruptions usually are a good thing for the channel. But when it comes simultaneously packaged in the form of both hardware and software, the opportunity created for solution providers is usually much more than the sum of its parts.

This week DataStax, which provides a commercially supported implementation of the open source Cassandra database, announced the NoSQL database is now available on Moonshot servers, a new architecture for servers developed by Hewlett-Packard.

Matt Pfeil, chief customer officer for DataStax, noted the combination of an open source NoSQL database designed to replace traditional relational database using a more efficient class of servers from HP—in this case the Moonshot servers that incorporate low-energy versions of Xeon processors from Intel packaged in a system-on-a-chip (SOC) design—creates a compelling value proposition that is going to be difficult to ignore.

DataStax, which this week is also celebrating the fifth anniversary of its founding, already has 500 customers. Cassandra is an open source project with well more than 500 users. DataStax Enterprise is an implementation of Cassandra that comes with management tools and commercial support, and has been gaining momentum as a high-performance alternative to relational databases that can support big data applications.

At its core, Cassandra is a distributed database that combines a key/value store with a columnar database architecture. It doesn’t support joins and queries the way a traditional relational database does, but it does enable masterless replication between nodes to enable applications to scale in ways that don’t require complex sharding of a relational database. The end result is a database platform capable of supporting high-performance applications without having to pay for commercial relational database licenses from, for example, Oracle or IBM.

One of the things that many organizations have yet to appreciate about Cassandra, which was developed at Facebook, is how extensible it is. In the months ahead Pfeil said, solution providers should expect to see time-series and graph database functionality layered on top of Cassandra.

Working with HP is critical for DataStax, Pfeil noted, not only because HP tends to dominate the high end of the server market, it has a massive presence in the channel that DataStax hopes to leverage.

Of course, when it comes to databases, IT organizations have no shortage of options, including more than few that HP has either developed or resells. The good news is that most organizations today are employing a broad mix of database technologies and platforms. That not only makes it easier to sell a NoSQL database such as Cassandra, it also provides solution providers with all kinds of data management services opportunities that should make managing all those diverse database platforms a lot easier for all concerned.