Joyent and Hadoop: Making Big Data Better

January 23, 2013 - by rachelbalik

Surely you’ve heard the famous saying, “Ask not what your cloud provider can do for you, ask what your cloud provider can do for Hadoop.”

What? You haven’t? Well perhaps that’s because it’s not quite famous yet, but once word gets out about the new Apache Hadoop-based solution we launched today, we expect this phrase to be top-of-mind for companies who are looking for better performance from Hadoop and have ambitious goals when it comes to managing big data.

Our new Hadoop-based solution empowers companies to be more effective with Hadoop by reducing costs and dramatically decreasing query response times. In early measurements by Altoros Systems, a software product development services company, Hadoop clusters on Joyent Cloud produced a nearly 3X faster disk I/O response time versus identically-sized traditional virtual machines.

Alternatively, companies looking to reduce their infrastructure footprint can cut costs by two-thirds without detracting from performance metrics.

How it Works

Running Apache Hadoop on Joyent accelerates response times of distributed and parallel processing and improves the scaling of Hadoop clusters executing intensive, data analytics applications. Through the use of its operating system virtualization and CPU bursting technology, Joyent is able to extract better response times and deliver results to data scientists and analysts faster.

Why It Matters

Early users of the new solution are already reaping the advantages for massive, parallel and distributed workloads. By helping to address new interactive workloads not currently addressed in Hadoop, Joyent is dramatically reducing the time required for businesses to gain insight from their most vital data. Joyent’s Hadoop solution will also offer a more consumer-friendly experience that can be added to existing Joyent Cloud implementations and easily configured to meet customer storage needs.

What’s Next?

This is our first foray into the big data business - and you can be sure it’s not our last. In the words of Jason Hoffman, our CTO, “We intend to continue bringing our technical expertise to the market and reverse the typical understanding of big data implementations - that they’re expensive and hard to use. We’re committed to meeting the insatiable demand for faster analytics and data retrieval, changing how computing functions for the enterprise.”

The Apache Hadoop solution is available for Joyent customers today. For more information, visit: Joyent.com/hadoop.