Recently, Ryan Rawson over at Drawn to Scale wrote on their decision to base their innovative product Spire on MapR’s distribution for Hadoop. I was excited to read Ryan’s posting for a couple of reasons:
- Spire represents a significant step forward in the usability of HBase™, the database component of Hadoop.
- MapR’s distribution for Hadoop enables Hadoop to reach its true potential – to exist not only as a critical service for the enterprise, but more importantly to evolve into the next-generation platform on which highly scalable services are provisioned to support the need to process Big Data quickly and reliably.
Usage of Hadoop is exploding within the enterprise because the growth of unstructured data is overwhelming the enterprise’s capability to process it. The traditional tools (namely BI tools, relational database technology and Linux servers) and the data warehouse solutions now need to be augmented with platforms such as Hadoop. However, anyone responsible for delivering a four- or five-nines Service Level Agreement with demanding business users will tell you that reliability, availability and serviceability are core attributes for any solution architecture. It is not possible to evolve Hadoop into an effective platform unless Hadoop is capable of delivering low-latency, highly reliable service to the enterprise.
Which is where MapR comes in. As Ryan sees it, MapR has proven itself capable of delivering not only the low-latency performance required to enable Spire, it’s also delivering the traditional tenets of enterprise-grade reliability such as data protection via snapshots and mirroring, uninterrupted service in the face of pervasive hardware failure, and efficient hardware utilization to enable cost-effective scale-out as demand grows over time. MapR’s customer base has concluded that HDFS is simply incapable of meeting these goals, which is why MapR replaced HDFS with a storage services layer designed to enable Hadoop to meet these demanding requirements. Drawn to Scale’s decision to embed MapR as their platform validates this approach.
Drawn to Scale is not the only innovative company recognizing MapR’s unique capabilities. Google and Amazon.com - both of whom have become household names in large measure due to their ability to deliver highly available services to their consumers – are independently partnering with MapR to deliver Hadoop in the Cloud. The fact that both companies have chosen MapR to enable their business is a testament to MapR’s proven capability to enable a high-performance, reliable Hadoop platform.
It’s exciting to see Hadoop evolve into the next-generation platform for the enterprise. Even more so to see MapR driving that evolution. Viva la difference!