JProfiler Get rid of your performance problems and memory leaks! Modern Garbage Collection Tuning Java Performance Training Courses COURSES AVAILABLE NOW. We can provide training courses to handle all your Java performance needs Java Performance Tuning, 2nd ed The classic and most comprehensive book on tuning Java Java Performance Tuning Newsletter Your source of Java performance news. Subscribe now! Enter email: Training online Threading Essentials course JProfiler Get rid of your performance problems and memory leaks!

Back to newsletter 214 contents

Performance is one of those quality attributes that used to be very hard to cost. But it's gradually become easier, not least because for many cases now you can directly measure the cost benefit of a performance change. For example, previously you would infer that getting the page display time down from say 1.4 seconds to 1.2 seconds would produce a revenue gain, and maybe even guess at a number for that. Then cross your fingers, and hope and try and see (and fail) amongst all the noise whether there was a revenue gain attributable, likely from data several days or weeks old.

But now, you deploy a canary version and direct a portion of traffic to that version and directly compare the result of the change on real- time transactions. You also have direct cloud running costs and you make a change that you can see immediately costs less than previously. Last year Segment gave details of the performance engineering exercise they did that that saved them a million Dollars annually - directly measurable cost benefits from performance engineering. Amazon and Google have tested small performance improvements (eg 100ms faster page display time) and determined the exact percentage improvement in revenue that improvement generates. Priorities can be defined very clearly in many cases based on how much money a performance improvement will save or generate. Welcome to the world where the performance engineer is now a cost-saving profit centre.

Now on to our usual Java performance news section, tools, articles, talks. And of course the tips from this month's articles and talks, as ever are extracted into this month's tips page.

A note from this newsletter's sponsor

News

Java performance tuning related news.

• Evolution of Strings in Java to Compact Strings and Indify String Concatenation - details that might interest you if you're into String internals and compiler and JIT optimizations of strings
• Project Loom, adding lightweight threads (fibres) to the JVM
• Startup Challenges by Claes Redestad - sum of improvements so far from JDK 8 through JDK 11, and a look at what new features like condy might mean for the near future
• An introduction to distributed tracing and Zipkin

Tools

Java performance tuning related tools.

• B3 propagation, standard for distributed tracing
• Java distributed tracing implementation compatible with Zipkin backend services
• OpenTracing (distributed tracing) API for Java
• OpenCensus vendor-agnostic libraries to provide observability for your services

Articles

• Hot code is faster code (Mark Price/Aitu Software)
• Back to the CompletableFuture: Concurrency in Action (Dmitry Vinnik/Devoxx)
• CompletableFuture Timeouts in Java (Grzegorz Piwowarek/4Comprehension)
• Top Performance Challenges in Distributed Architectures (Andreas Grabner/Devoxx)

Jack Shirazi

Back to newsletter 214 contents