Track: Evolving Java

Fresh Async With Kotlin

Asynchronous programming is on the rise. Modern software systems are connected and constantly communicating. Programming languages are adding some form of asynchronous programming like async/await. However, Kotlin had taken a fresh approach to this problem with Kotlin Coroutines.

In this talk, we’ll study various approaches to asynchronous programming, their evolution, differences and similarities. We’ll see the problem with the traditional async/await approach that is based on futures/promises and how the Kotlin’s solution that is based on concepts of coroutines and continuations is giving us safer and easier programming model.

Servlet vs Reactive: Choosing the Right Stack

When Netflix upgraded their main gateway, serving 83 million users, from Servlet based, blocking Zuul 1 to the Netty based, non-blocking Zuul 2, the results were interesting and nuanced with benefits and trade-offs. Spring Framework 5 provides a similar choice with Servlet based Spring MVC and the new reactive, web framework called Spring WebFlux.

How do you make sense of this choice? The key is to understand the trade-offs and pick the right stack for the job at hand. Of course you also need understand differences in the execution models and the impact on the programming model you use.

In this session we'll discuss these questions and also use demo code.

Lessons Debugging Serverless JVM Functions

Microservices (and now serverless) pose a lot of interesting challenges when it comes to monitoring. Decoupling of code down to the "function" level offers a *lot* of interesting opportunities in terms of efficiently architecting code, but many a time does so at a loss of context. The lose of context makes monitoring of apps and performing Root Cause analysis of issues harder than what it used to be in monolithic (and even SOA) architectures. This talk condenses some of the serverless monitoring research and practices into an experience rich talk full of actionable advice for those implementing serverless.

Continuous Optimization of Microservices Using ML

Performance tuning of microservices in the data center is hard because of the multitude of available knobs, the large number of microservices and variation in work loads, all of which combine to make the problem combinatorially intractable. Maintaining optimal performance in the face of continuous upgrades to the service, and of the platform software and hardware, makes the problem even harder. As a result, lots of performance is typically left on the table, and data center resources wasted. We share our recent experiences in applying a technique from machine learning, called Bayesian optimization, to the performance tuning problem. We describe the implementation of a service for continuously optimizing microservices in the data center using this technique.

Java 9: Tips on Migration and Upgradability

Join Bernard Traversat, Vice President of Java SE Development at Oracle to discuss migration strategies for Java 9 and the important of the new Java 9 module system to improve Java code upgradability. As the speed of deployment in the cloud is accelerating, the compatibility and upgradability of cloud services are becoming critical. The Java 9 modular runtime and module system provide a fundamental new way to assemble and protect compatibilities between Java code. This session will review the compatibility and upgradability stand taken by leading developer’s platforms (Java, JavaScript, Node.js, Python, Go), cover the new Java 9 features (module system, modular runtime, multi-jar packaging, Jdep, etc), and discuss migration strategies to migrate your existing code to Java 9. The session will provide number of useful tips to ease the migration of your code to Java 9.

Evolving Java Open Space