MapR Technologies, Inc., provider of the Converged Data Platform that fully integrates analytics with operational processes in real time, announced that Enterprise Strategy Group (ESG) published an ESG Technical Review that evaluated the performance and scalability of running MapR-DB in the cloud, and compared results with two open-source NoSQL database offerings.
ESG audited and analyzed tests to validate the performance capabilities of MapR-DB, Apache Cassandra and Apache HBase in the same testing environment. The goal of testing was to first create a common hardware configuration, and then configure each of the databases with identical consistency levels to ensure a fair comparison. Results from five core workload tests run on three similarly configured deployments include:
- MapR-DB outperformed Cassandra and HBase, with over 10x performance in some cases
- MapR-DB yielded an average performance improvement of 2.5x more operations/sec than Cassandra and 5.5x more than HBase
- When looking at latency for one of those workloads, MapR-DB proved to be the fastest, and the most predictable
- Cassandra yielded higher latency and HBase proved unreliable in delivering fast or consistent latency
When leveraging NoSQL databases to meet the performance requirements of demanding operational analytic workloads, organizations want more than just speed—they also want consistency. This enables a predictable end-user experience that meets the needs of a dynamic business that continues to crave a real-time response,” said Mike Leone, senior validation analyst, ESG. “ESG validated MapR-DB’s ability to deliver faster, more consistent performance in an AWS cloud environment compared with open source offerings HBase and Cassandra.”
Testing by ESG was completed using the Yahoo Cloud Serving Benchmark (YCSB) version 0.12. YCSB is an open source, industry-standard framework for evaluating the performance of key-value and cloud-based serving datastores. Tests were performed in Amazon EC2 using the recently released Enhanced Networking Adapter (ENA), highlighting that MapR-DB can leverage additional cloud resources to deliver higher performance for applications.
As part of the MapR Converged Data Platform, MapR-DB is an evolved, enterprise-grade NoSQL database delivering high performance, reliability, and scalability for big data analytic workloads. It is capable of extreme scalability to provide real-time data analytics without the extra complexity and cost of a traditional database architecture. With MapR-DB, organizations gain a responsive big data analytics platform with multi-modal support for JSON and open-API functionality.
We built MapR-DB from the ground up for data-intensive applications. MapR Engineering team built data structures that ensure consistent low latency and are capable of extreme scalability to provide real-time data analytics without the extra complexity and cost of a traditional database architecture,” said Anoop Dawar, vice president, product marketing and product management, MapR Technologies. “The outstanding results found in ESG’s testing prove MapR is consistently addressing many of the most significant challenges organizations face while building modern data intensive applications.”
Sign up for the free insideAI News newsletter.
