BT

Facilitating the Spread of Knowledge and Innovation in Professional Software Development

Write for InfoQ

Topics

Choose your language

InfoQ Homepage News JavaOne 2016 - Day 1 Highlights

JavaOne 2016 - Day 1 Highlights

This item in japanese

This year JavaOne live streamed sessions from four rooms for the entire duration of the five-day conference, and those presentations were also made available right after the broadcast.

Learn Java 8: Lambdas and Functional Programming by Henri Tremblay from Terracotta's EHCache team recapped the evolutoon of Java since Java 5 and generics, through Java 7 syntax simplifications, and Java 8 with Lambdas and Nashorn's JavaScript support. Keeping up with his live coding tradition, Tremblay provided many of his answers using live-coding.

Arun Gupta of Couchbase covered Docker for Java Developers. He started with the Docker mission of build, ship and run and then contrasted Docker based containers with typical virtual machines (VMs) on a Hypervisor, as shown in this image from docs.docker.com. Gupta then spoke in detail about the Docker toolbox and also provided information on swarm mode and rolling updates.

There were a several presentations on open source. One was by James Ward of Salesforce who spoke about Managing Open Source Contributions in Large Organizations. Ward talked about the why of open source and the concerns around open-sourcing, as well as strategies to mitigate those concerns, the most common being – do nothing!

There was a presentation on Automated Tuning of the JVM with Bayesian Optimization by Twitter engineers Ramki Ramakrishna, Alex Wiltschko and Jianqiao Liu. Ramakrishna first introduced the JVM tuning problem, explaining that there are some 800 tunable switches, many of which are dependent on the hardware or on each other. Out of those 800, about 250 influence performance. Ramakrishna talked about performance tuning and how it needs to be contiguous, requiring a “black-box tuning assistant” that could provide suggestions. Wiltschko then spoke about Bayesian Optimization, a machine learning approach to black-box optimization. He also provided a one dimensional tuning example shown below:

Liu then spoke on tuning the JVM performance. And finally Ramakrishna summarized the findings:

Rate this Article

Adoption
Style

BT