Warning message

  • The service having id "twitter" is missing, reactivate its module or save again the list of services.
  • The service having id "facebook" is missing, reactivate its module or save again the list of services.
  • The service having id "google_plus" is missing, reactivate its module or save again the list of services.
  • The service having id "linkedin" is missing, reactivate its module or save again the list of services.

Track: Unlearning Performance Myths


Day of week:

"I've been doing that for years, and it's been totally wrong all this time!!?!". Surprisingly, In our field this sort of realization actually tends to be enjoyable. I call it "unlearning", and I get a kick out of it every time it happens to me. Lately, I've been giving other people kicks too. And since many of them seem to enjoy it, we've decided to put a whole track together around the theme of unlearning myths and mis-practices around performance and it's understanding. This track will tackle and attempt to slay myths around how performance is thought about, measured, and described, as well as things that we can do to affect it. We'll cover throughout and response time behaviors, how code profiling works (and doesn't), what network protocols APIs, and associated design choices actually do well (and badly), how compilers and runtimes do or don't work, and how practical scaling interacts with performance. Each talk will include unlearning points, and some may also include helpful learning to fill in the vacuum created. Our track goal is simple: you will leave each talk knowing less than you did coming in. And that's a good thing.

Track Host:
Gil Tene
CTO and co-founder @AzulSystems
Gil Tene is CTO and co-founder of Azul Systems. He has been involved with virtual machine and runtime technologies for the past 25 years. His pet focus areas include system responsiveness and latency behavior. Gil is a frequent speaker at technology conferences worldwide, and an official JavaOne Rock Star. He pioneered the Continuously Concurrent Compacting Collector (C4) that powers Azul's continuously reactive Java platforms. In past lives, he also designed and built operating systems, network switches, firewalls, and laser based mosquito interception systems.
10:35am - 11:25am

by Gil Tene
CTO and co-founder @AzulSystems

Measuring, monitoring, and improving application responsiveness is a common need for many software professionals. Whether you develop applications or manage them, understanding application responsiveness is key to achieving successful applications and happy users. But good characterization of bad data is useless. When measurements of response time present false or misleading latency information, even the best analysis can lead to wrong operational decisions and poor application experience....

11:50am - 12:40pm

by Todd Montgomery
Ex-NASA researcher, Chief Architect, Kaazing

"The network is the bottleneck!”, “Disk is too slow!”, “We are I/O bound!" If these phrases are things you normally hear or say, then how sure are you about that? Chances are the hardware, especially I/O, you are using is loafing along and it is possible to get much more out of it. Software and hardware change rapidly. Conventional wisdom must evolve to keep up and now is a good time to revisit old (and some new) ideas with new and different perspectives.

1:40pm - 2:30pm

by Nitsan Wakart
Lead Performance Engineer for Azul Systems

In our time of need we turn to profilers to guide us in our search for bottlenecks and hotspots. We put our trust in the numbers and facts they spout. In this talk we'll discuss concrete cases in which profilers misguide, misrepresent and at times subvert the systems they aim to help us diagnose. Unlearn your trust, and relearn skepticism.

2:55pm - 3:45pm

by Cliff Click
CTO and Co-Founder of H2O

Java vs C Performance. Again. There seems to be a set of widely held wildly incorrect beliefs about Java and C performance. I say "widely held" because I hear them constantly - even after 20 years of pointing out the evidence. I say "incorrect beliefs" because they are shown wrong with even the most rudimentary scientific study, i.e. a simple benchmark. I say "wildly incorrect" because nearly always I see blatantly broken benchmarking, being used to drive home some wrong point.


4:10pm - 5:00pm

by Brendan Gregg
Senior Performance Architect @Netflix

Broken benchmarks, misleading metrics, and terrible tools. This talk will help you navigate the treacherous waters of system performance tools, touring common problems with system metrics, monitoring, statistics, visualizations, measurement overhead, and benchmarks. This will likely involve some unlearning, as you discover tools you have been using for years, are in fact, misleading, dangerous, or broken.

The speaker, Brendan Gregg, has given many popular talks on operating system...

5:25pm - 6:15pm

Open Space


Covering innovative topics

Monday Nov 16

Tuesday Nov 17

Wednesday Nov 18