DELIVER SLAs WITHOUT TRADEOFFS
Does your application have some tight SLAs to meet, but you're always facing tough tradeoffs to meet them? Terracotta gives you a way to meet tight SLAs, without having to sacrifice availability, throughput, or scalability. Terracotta is an application memory store that's as safe as transactions, but without all the development complexity or performance degradations that tend to come with transactions, and it lets you scale your architecture without a major redesign.
A SIMPLER AND MORE EFFICIENT APPROACH TO SCALABILITY
Terracotta's unique approach to making applications simultaneously highly-available and scalable is at the core of its ability to deliver tight SLAs without tradeoffs. Terracotta does this by hooking directly into the Java Virtual Machines (JVMs) that are running your application, and, based on its configuration, mapping pre-determined parts of memory from one JVM to another in an extremely efficient, just-in-time manner. Only what needs to be shared is shared, and only where and when it is needed.
NO NEED TO CHANGE APPLICATION CODE
This shared memory pool between JVMs results in an application that is resilient to server failure and other faults. Since the shared data is exchanged in such an efficient manner, Terracotta can achieve high availability at high scale, without sacrificing application performance. Also, since Terracotta hooks into the JVM to do this, there is no API to add to your code to use Terracotta, and many applications currently using Terracotta have required no changes to use it.
