News Press Releases

Early Customers Use BigMemory for Enterprise Ehcache to Achieve Breakthrough Memory Utilization™

Snap-in Module to Ehcache Solves Universal Garbage Collection Problem: Increases Usable Memory, Eliminates Exhaustive GC Tuning, and Enables IT to Better Meet SLAs

San Francisco—November 9, 2010—Terracotta, provider of some of the most widely used software for application scalability and performance, today announced the general availability of BigMemory™ for Enterprise Ehcache, a breakthrough solution for the universal problem of garbage collection (GC) in Java applications. BigMemory’s initial Beta program included more than 250 registered participants from 30 countries in six weeks, many of whom achieved unprecedented application performance without months of custom tuning.

Fixing Java’s Achilles’ Heel: Garbage Collection

“Garbage collection has long been Java’s Achilles’ heel,” explained Massimo Pezzini, vice-president and Gartner Fellow at Gartner, Inc. “Developers spend significant effort to work around the problem of application performance unpredictability stemming from Java garbage collection. Users will welcome structural solutions to this universal challenge.”

A Structural Solution to GC: Achieve Months of GC Tuning Results in Minutes

BigMemory is Terracotta’s structural solution to the garbage collection problem. As a pure Java add-on to Ehcache, BigMemory is an in-process, off-heap cache that is not subject to garbage collection. By efficiently storing data in-process but off-heap, where it is invisible to the Java garbage collector, BigMemory allows an application to run on a Java Virtual Machine (JVM) with less than a gigabyte of heap, while using hundreds of gigabytes of off-heap memory for fast access to data.

BigMemory’s dramatic increase in usable memory eliminates the need for costly, time-intensive workarounds. Until now, developers have struggled to optimize garbage collection by manipulating a number of configuration parameters. With much effort, developers eke out only marginally improved results which can be disrupted by changes in application code, heap size, CPU utilization or even the mix of applications running on the server.

Terracotta Customers Extol Benefits: Massive Increase in Memory Utilization without Lengthy Tuning Effort

“Running Java heaps larger than 6 GB is a challenge for any large enterprise deployment,” explained Joey Caisse, CTO of News Digital Media, a subsidiary of Rupert Murdoch’s News Corporation. “With Terracotta, we are able to reduce the heap size, and get fast, local access for large amounts of data. It’s much easier for us to manage, with less GC tuning and more predictable application performance.”

In Beta testing of BigMemory, a large e-commerce customer, who spent as long as three months tuning an application to get the maximum latency down to an acceptable level, achieved the same results in less than three hours with BigMemory.

Performance Benchmarks: In-Memory Speed at Any Size

Typically, Java applications run with 2 to 4 GB of memory. Now that 64-bit machines are the norm and servers typically ship with 8 or 16 GB of RAM, much more memory is available to applications. Using that memory to cache more data can significantly improve application performance; however, increasing the heap size of the JVM to beyond 4 to 6 GB, typically causes GC pauses long enough to seriously impact application performance.

By efficiently storing data in-process but off-heap, BigMemory allows an application to run on a JVM with less than a gigabyte of heap, yet use hundreds of gigabytes of off-heap memory for fast access to large amounts of data. In performance tests, BigMemory demonstrates steady, predictable application behavior out to hundreds of gigabytes of in-memory cache (see chart below). By contrast, the same performance tests using on-heap caches resulted in increasingly longer garbage collection pauses, which can lengthen to minutes.

Deploy with Snap-In Simplicity

BigMemory works with all mainstream JVMs and generic, low-cost hardware. Customers do not have to adopt and re-qualify a new application stack to run BigMemory because the product snaps into Enterprise Ehcache with no other changes.

“Terracotta’s mission is to provide easy-to-use, snap-in solutions to some of the most difficult challenges impacting the performance and scalability of enterprise applications, and BigMemory is the next major step,” said Ari Zilka, CTO of Terracotta. “BigMemory snaps in with no config changes and works with all major JVMs. It provides a simple way to use all the available memory on a machine in a single JVM, thus ensuring predictable latency, eliminating the application pauses that frustrate users, and enabling IT to more precisely meet established SLAs.”

Availability

For those interested in trying BigMemory for Enterprise Ehcache, please visit http://www.terracotta.org/bigmemory to download the product.