Published by Gbaf News
Posted on August 4, 2015
New Java performance services tailored to improve Cassandra response time, reliability, up-time, and throughput
News highlights:
Azul Systems (Azul), the award-winning leader in Java runtime solutions announced the availability of Cassandra HealthCheck, a new business consulting service designed to help enterprises eliminate Java-related barriers and improve the runtime consistency, reliability, and up-time of their Cassandra deployments. As an example of the unique capabilities of this new service, Azul recently achieved over 56x improvement in worst-case latency, 25% improvement in throughput, and dramatic reduction in client disconnects due to node timeouts for a large-scale Cassandra deployment with a customer in the financial services industry.
Cassandra HealthCheck (Cassandra HC) consists of a combination of onsite and remotely-delivered runtime analytic services. Leveraging both open source and proprietary tools, Azul consultants have deep Java experience helping IT organizations and DevOps teams optimize Java runtimes and system and application configurations to significantly improve Cassandra deployments, including components leveraging Cassandra data stores such as Apache Lucene, Solr and Spark.
Apache Cassandra is a highly performant, scalable open source distributed database management system written in Java. However, many aspects of Cassandra performance, consistency, and availability depend on the capabilities of the underlying Java runtime platform. Legacy Java Virtual Machines (JVMs) may suffice for some use cases, but enterprises that require highly consistent and reliable low-latency Cassandra deployments are often limited by the JVM itself, specifically the detrimental impact of garbage collection. The Java garbage collection process freezes the application while memory is defragmented and compacted, resulting in random Cassandra pauses that lead to response time inconsistency, increased time to data consistency, and can even trigger serious conditions like cascading node failures, where multiple stalled Cassandra nodes trigger cluster-level outages and crashes.
“We are seeing strong interest in the Apache Cassandra NoSQL database, particularly for applications that require a combination of low-latency performance and a highly available, distributed architecture,” commented Matt Aslett, research director, data platforms and analytics, 451 Research. “Garbage collection issues can hit any Java deployment, but are particularly significant where low-latency performance is a requirement. With Azul’s new Cassandra HealthCheck offering, Cassandra users can leverage the company’s JVM and performance tuning expertise, allowing for more consistent and reliable deployments while focusing their attention on the application itself rather than spending time trying to tune away JVM issues.”
“Many users do not understand the negative impact a Java runtime can have on Cassandra response time, reliability, and throughput,” said Scott Sellers, CEO and President of Azul Systems. “With Azul’s Cassandra HealthCheck service offering, our specialists can quickly show enterprises how to maximize the value of their Cassandra investment and eliminate many of the performance and reliability challenges facing Cassandra deployments today. Better Cassandra deployments ultimately means better business performance.”
Explore more articles in the Top Stories category
