Ruby Performance - The Last Mile - RubyConf India 2016

Talk on how JRuby is bringing performance and concurrency to Ruby, delivered at RubyConf India 2016 in Kochi, Kerala, India.

Transcript

1. Ruby Performance The Last Mile
2. Charles Oliver Nutter @headius
5. I've spent ten years building
6. I love Ruby
7. I want Ruby to succeed
8. I believe JRuby is the way
9. Do you want performance today?
10. Do you want concurrency today?
11. Do you know JRuby?
13. What you love from Ruby • Latest Ruby language features • Mostly-same Ruby standard library • Pure-Ruby gems work great • Native gems with JRuby support • It walks like Ruby, talks like Ruby • It is Ruby!
14. With the power of the JVM • Fast JIT to native code • Fully parallel threading • Leading-edge garbage collectors • Access to Java, Scala, Clojure, ... • But it's still Ruby!
16. Do you know Jay?
17. Matz's Keynote
18. Performance
19. Concurrency
20. By 2020
21. But what about today?
22. Performance
23. What do you optimize for? • Easy to develop with: short time until first deploy • Fast startup: good response cycle at command line • Straight-line performance, many operations per second • Parallelism: utilize many cores to get more done
24. State of the Art: Production-quality Rubies
25. Production Quality? • Support for 99%+ of Ruby language features • Important parts of standard library • Runs typical Ruby applications and libraries • Healthy extension ecosystem • CRuby, JRuby are the only real options right now
26. CRuby (MRI) • Up until 1.9, AST interpreter • YARV bytecode VM introduced for 1.9.0 • GC and performance improvements through 2.x series • Ruby 2.3 is latest, released in December • Future work on JIT, GC, happening now
27. JRuby • Many redesigns since creation in 2001 • AST interpreter until 2007 • Simple AST-to-bytecode JIT until JRuby 9000 • Optimizing compiler with JIT for 9k • JRuby 9.0.5 is current, 9.1 in a couple weeks • Next-gen Truffle runtime in the works
28. Lies, damn lies, and benchmarks
29. CRuby: red/black tree 0 2 4 6 8 1.8.7 1.9.3 2.0 2.1 2.2 2.3
30. CRuby: red/black tree 0 0.625 1.25 1.875 2.5 1.8.7 1.9.3 2.0 2.1 2.2 2.3
31. JRuby: red/black tree 0 1.25 2.5 3.75 5 JRuby 9.1 CRuby 2.3
32. JRuby: red/black tree 0 1.25 2.5 3.75 5 JRuby 9.1 CRuby 2.3
33. JRuby: red/black tree 0 0.6 1.2 1.8 2.4 JRuby 9.1 CRuby 2.3 4x FASTER
34. Ruby Code
35. Ruby Code Parser JRuby AST
36. r JRuby AST Compiler JRuby IR
37. er JRuby IR JIT JVM Bytecode
38. What can we do with this?
39. Block Pass-through def foo(&b)
 bar(&b)
 end
 
 def bar
 yield
 end
40. Block Pass-through loop {
 puts Benchmark.measure {
 i = 0
 while i < 1_000_000
 i+=1
 foo { }; foo { }; foo { }; foo { }; foo { }
 end
 }
 }
41. Block Passing 0 0.55 1.1 1.65 2.2 CRuby 2.3 JRuby 1.7.24 JRuby 9.1
42. define_method define_method :add do |a, b|
 a + b
 end
43. define_method 0 0.25 0.5 0.75 1 CRuby 2.3 JRuby 1.7.24 JRuby 9.1
44. Postfix rescue foo rescue nil
45. csv.rb Converters Converters = {
 integer: lambda { |f|
 Integer(f.encode(ConverterEncoding)) rescue f
 },
 float: lambda { |f|
 Float(f.encode(ConverterEncoding)) rescue f
 },
46. Postfix rescue 0 3.5 7 10.5 14 CRuby 2.3 JRuby 9.1
47. 0 3.5 7 10.5 14 CRuby 2.3 JRuby 9.1 Postfix rescue
48. CRuby starts up the fastest
49. JRuby runs the fastest
50. And we're getting faster
51. Concurrency
52. Parallelism
53. Concurrency? Parallelism? • Parallelism happens on the harder, e.g. multi-core • Concurrency happens in the software, e.g. Thread API • You can have concurrency without parallelism • You can have both with JRuby
55. Parallelism in Ruby • On CRuby, usually process-level • Ruby threads prevented from running in parallel • Extensions, IO can opt to release lock • On JRuby, usually thread-level • Ruby thread == JVM thread == OS thread • Single-process, shared memory
56. A Mailing Queue • A simple example of concurrency • For each job, construct an email to send • Some computation added to make processing heavier • "Ruby Concurrency and Parallelism: A Practical Tutorial"
 https://www.toptal.com/ruby/ruby-concurrency-and-parallelism-a- practical-primer
57. require "./lib/mailer"
 require "benchmark"
 
 puts Benchmark.measure{
 (ARGV[0] || 10_000).times do |i|
 Mailer.deliver do
 from "eki_#{i}@eqbalq.com"
 to "jill_#{i}@example.com"
 subject "Threading and Forking (#{i})"
 body "Some content"
 end
 end
 }
58. POOL_SIZE = (ARGV[0] || 10).to_i
 
 jobs = Queue.new
 
 (ARGV[1] || 10_000).to_i.times{|i| jobs.push i}
 
 workers = (POOL_SIZE).times.map do
 Thread.new do
 begin
 while x = jobs.pop(true)
 Mailer.deliver do
 ...
 end
 end
 rescue ThreadError
 end
 end
 end
 
 workers.map(&:join)
59. CRuby: mailer * 1000 Time in Seconds 0 37.5 75 112.5 150 synchronous 4 threads 4 forks
60. JRuby: mailer * 1000 0 7 14 21 28 Synchronous 4 Threads
61. JRuby vs MRI Times Improvement 0 0.85 1.7 2.55 3.4 CRuby Forks JRuby Threads 3.37x 3.09x
62. But Threads are bad, right?
63. Most users will never Thread.new
64. You'll deploy one Rails server for your entire site
65. You'll cut your instances ten times
66. Or maybe 100 times
67. Libraries and frameworks will Thread.new for you
68. And on JRuby, you'll have more efficient apps
69. So we're done?
70. Move to JRuby
71. Now your app is fast!
72. Right?
73. It is possible to write efficient Ruby code
74. But it's very easy to write inefficient Ruby code
75. Great Features, Hidden Costs • Blocks are expensive to create, slower than method calls • case/when is an O(n) cascade of calls • Singleton classes/methods are costly and hurt method cache • Literal arrays, hashes, strings have to be constructed, GCed • Flow-control exceptions can be very expensive and hard to find
76. What happens if your code does not run fast enough?
77. You need to know your app
78. You need good tools
79. And the JVM has great tools
80. CRuby Tooling • Basic GC stats built in • Simple profilers in standard library • Some third-party tools • stackprof, ruby-prof, perftools.rb, ...
81. JVM tooling is JRuby tooling
82. JVM Tooling • Wide range of GCs: parallel, concurrent, realtime, pauseless • Built-in tools for analyzing GC, JIT, thread, IO, heap • Built-in remote monitoring via JMX • Dozens of tools out there for profiling, management, and more
83. VisualVM • Graphical console into your application • Monitor GC, threads, CPU usage • Sampled or full profiling with GUI browser • Live memory dumping, heap inspection • Ships with every OpenJDK install
86. Java Mission Control • Extremely low-overhead application recording • Analyze results offline in JMC • GC, CPU, heap events, IO operation all browsable • Commercial feature, free for development use
90. More on GC • JVM GCs are incredibly tunable with sensible defaults • Tools like http://gceasy.io and JClarity give you a deeper view • These are the best GCs and the best tools in the world
91. Finding Bottlenecks
92. Profiling
93. Profiling Tools • Command line options: --profile, --sample • JVM command-line profilers like prof • Many graphical sampling/complete profiling options • Flame graphs, stack profilers, you name it!
97. ...and this is just the beginning
98. Wrapping Up
99. Ruby is alive and well
100. CRuby continues to improve
101. Ruby 3 is very exciting
102. JRuby is performance today
103. JRuby is concurrency today
104. JRuby has tools today
105. JRuby makes you happier!
106. Thank You! Charles Oliver Nutter headius@headius.com @headius http://jruby.org https://github.com/jruby/jruby