Menu
Java garbage collection issue makes the rounds

Java garbage collection issue makes the rounds

Tackling 'universal problem,' Terracotta says BigMemory module helps improve apps performance

Tackling memory limitations imposed on Java applications, Terracotta on Tuesday will offer what it describes as a breakthrough in solving the "universal problem" of Java garbage collection.

The company will release a beta version of BigMemory for Enterprise Ehcache, an add-in module to the open source Ehcache Java caching library now owned by Terracotta. Compatible with popular Java Virtual Machines, BigMemory provides an off-heap cache to free Java applications from memory and performance constraints of garbage collection, thus providing a large memory footprint for Java applications, Terracotta said. Garbage collection is bypassed via Terracotta's technology.

[ The annual JavaOne technical conference is being held later this month, sans Google. | Keep up with the latest Java developments with the JavaWorld Enterprise Java newsletter. ]

Garbage collection provides Java memory management, reclaiming objects no longer used by a program. But Terracotta describes garbage collection as the "Achilles' heel" of Java applications, impacting performance of applications and limiting access to data.

"Java is kind of stuck in the small memory world because of Java garbage collection. People typically run with small heaps [of 2GB to 4GB]," said Mike Allen, senior director of product management at Terracotta.

Benefits of BigMemory include:

  • Increasing usable memory, enabling applications to easily cache 64GB or more of data in an off-heap store not subject to garbage collection. Application speed is increased by enabling applications to run on fewer JVMs with bigger caches.
  • Provision of performance and predictability for service-level agreements, via improved application throughput.
  • Elimination of tuning and inefficient workarounds that try to optimize around garbage collection.

An industry analyst, though, was not sold on Terracotta's solution, suggesting Java shops already have developed work-arounds to Java memory problems.

Java memory limitations and garbage collection problems "are long-term issues and Java shops have learned to work around them using manual testing and tuning techniques," said Matthew Aslett, senior analyst for enterprise software at the 451 Group. "While there will be some that will jump at the chance of an alternative, others are bound to be skeptical and may prefer to stick with their own work-arounds, at least until Terracotta can deliver real-world performance figures that demonstrate its claims to deliver predictable latency times." 

BigMemory for Enterprise Ehcache will be generally released in October.

Another vendor, Azul Systems, also attempts to solve the issue with Java performance via the company's Managed Runtime Initiative, which is intended to improve execution of managed runtimes via enhanced interfaces and functionality across components of a systems stack, such as the managed runtime, OS kernel, and hypervisor. Azul's Zing software, billed as an elastic runtime platform for Java applications, includes open source components of the initiative.

But Azul, Allen said, offers an "intrusive solution."

"They're trying to solve Java garbage collection using one approach. We're looking at it in a very different way," he said.

Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.

Join the newsletter!

Or

Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.

Tags open sourceGooglejavasoftwareapplication developmentDeveloper World

More about GoogleOracleProvision

Show Comments
[]