Azul Systems Solves Java’s ‘Warm-Up’ Problem

Azul Systems has announced ReadyNow!, new technology that, the company says, solves the “warm-up” problem prevalent in many Java applications. ReadyNow! is included in the latest version of Azul’s Zing runtime for Java, version 5.9, and is intended to ensure peak application performance is available without long “warm-up” periods and during other critical times.  According to the company, ReadyNow! is well-suited for use in the financial markets that require peak system performance at the critical moments during market open as well as during unpredictable surges in volume or trading behavior.

According to Azul, the Java “warm-up” problem refers to an aspect of the behavior of a Java Virtual Machine (JVM), the software responsible for running Java applications. As a Java application is launched, the JVM is responsible for compiling the Java application into a form that can be run on a computer or server, and as the application continues to run the JVM will recompile important parts of the application code to improve performance.  However, the company says, a problem exists as often a Java application has not yet been compiled optimally when peak performance is actually required.  In practice, the Java application needs to “warm-up” before it is ready to reliability perform at the level needed.

In financial markets, this “warm-up” period can result in lost trading opportunities, lost liquidity, lost revenue, lost market and risk data, and ultimately loss of customers.  These losses can accumulate and have a significant impact on a firm’s market presence and standing in this highly competitive sector.

To compensate for these problems, firms utilize various techniques including executing “warm-up” test data that attempts to mimic real market conditions, issuing and canceling small live trades, and various other programming practices. These workarounds for the “warm-up” problem carry real operational risks, however, sometimes executing accidental trades that can cost millions or in extreme cases cause market “flash crashes.”  And yet, even after assuming these risks, the systems may still not realize peak performance at market open due to the unpredictable and dynamic characteristics of a real-world trading environment.

The goal with ReadyNow! Is to provide traders, market makers, and risk analysts with the confidence that their Java-based systems, and their software, will be  “warmed up” when they are needed most, without unexpected stalling and without assuming operational risk.