Java Application Scalability and Performance
What Limits Java Scalability?
Innovation in Java Virtual Machines (JVMs) hasn’t kept up with the scalability needs of Java applications. Online products and services such as portals, Ecommerce and trading systems have heavier transaction loads and data-intensive features that were not foreseen even ten years ago. At the same time, IT organizations are being asked to lower costs and increase flexibility through virtualization, Cloud and x86 replatforming initiatives.
Conventional JVM limitations:
- Inefficiency: can't scale beyond 2-3 GB of heap memory per instance without long garbage collection (GC) pauses and poor application performance.
- Rigidity: fixed memory allocation size (i.e. -Xmx) for each JVM instance.
- Instability: inconsistent behavior outside of a small operating range.
- Complexity: poor visibility and management within and across a 'proliferation' of JVM instances.
- Topology: ill-suited for performance-centric virtualized and Cloud deployments.
In contrast, the Zing™ JVM allows each instance to scale to hundreds of GBs of heap memory and tens of cores seamlessly and elastically.
Getting Started With Zing
Azul Case Studies
 |  |
 |  |
 |  |
