5.0

Introducing vFabric Cloud Application
Platform 5.0
11
11
• Commercial support for vFabric tc Server and Spring Framework.
• Elastic Memory for Java (EM4J). Included with tc Server Spring Edition, EM4J allows Java workloads to cooperate more fully
with ESXi advanced memory-sharing technologies. EM4J helps maintain Java application performance and reliability while
freeing memory so you can run more VMs per host, improving your consolidation ratio.
• Spring Insight Operations, a Web application that gives you real-time, detailed visibility into application behavior and
performance. Insight Operations is optimized for monitoring and tuning Spring-built applications in production.
EM4J 1.0 for Improved Memory Management of Java Workloads
EM4J revolutionizes the virtualization of Java applications.
With EM4J and tc Server, you can run more Java applications on your ESXi servers than previously possible, and Java
applications can now leverage ESXi's industry-leading memory sharing technology.
Previously if you ran Java on VMware ESXi, you had to reserve 100% of the configured memory for a VM running Java
applications or risk serious performance problems. To determine the optimum size of the Java heap, you determined the
application's peak requirement. Allocating less could mean unacceptable performance, errors, or crashes; allocating more than
was necessary wasted memory.
EM4J changes all that. It is a balloon that sits directly in the Java heap and works with a new memory-sharing interface
introduced in ESXi 5.0. EM4J allows ESXi to share available memory in the Java heap among VMs as needed, while preserving
application performance and reliability. You can size the Java heap to accommodate the peak workload, without wasting or
running out of memory. Any excess memory is ballooned away and redistributed where it is needed.
This capability lets you create more VMs and pack more tc Server instances onto a single vSphere host. It also reduces the
likelihood of out-of-memory errors that can cripple applications.
A new EM4J plug-in is available for the vSphere Web Client. The plug-in enables you to monitor EM4J configuration and
memory performance for EM4J-enabled Java workloads in the vSphere Web Client. A Workloads tab displays statistics for
Java workloads on the virtual machine selected in the vSphere Web Client inventory tree. You can quickly verify that the virtual
machine and JVM are configured correctly for EM4J and see detailed information about the JVM process and memory usage.
Alerts warn of EM4J configuration problems and suggest best practices.
Spring Insight Operations 1.5 Engineered for vSphere, Spring, and Production Environments
Spring Insight Operations gives you real-time, granular visibility into application behavior and performance in production
environments. It is a Web application that runs with tc Server Spring Edition and is a separate download.
Insight Operations graphs the health of an application over time for the entire cluster and for each server in the cluster.
Administrators see application and server problems as they occur, with detailed information about contributing events, presented
in one unified dashboard. You can quickly identify under performing servers, server errors, and trace details for specific events
and exceptions.
VMware partners with Google to integrate Spring Insight data into SpeedTracer, a Google Chrome extension that analyzes
how your application is performing inside the browser and that ties browser performance to backend performance. If your Web
application uses Ajax and other rich open Web technologies, try SpeedTracer with Spring Insight.
Main features of Spring Insight Operations 1.5:
• Automatic instrumentation of Spring applications running on tc Server expose application performance from both general and
fine-grained perspectives.
• Isolation of performance problems by application, endpoint, application server, and method.
• Tracking of performance over time, to understand whether problems are chronic or spurious.
• Response-time histograms that reveal whether performance problems affect a many users, or a small subset.