Faster than C?

Presentation given at nodedublin.

Transcript

1. Faster than C?
   Parsing binary data in JavaScript
   Felix Geisendörfer 18. Oct, 2012 - NodeDublin
   

   View Slide

2. @felixge
   transloadit.com nodecopter.com
   

   View Slide

3. My Battle
   

   View Slide

4. JavaScript vs C
   

   View Slide

5. Good vs Evil
   

   View Slide

6. Good Parts vs Evil
   

   View Slide

7. Bad Parts vs Evil
   

   View Slide

8. Faster than C?
   

   View Slide

9. Sorry for the
   “title bait”
   

   View Slide

10. JavaScript vs C-Addons
    (for parsing binary protocols)
    

    View Slide

11. The Story
    

    View Slide

12. early 2010
    

    View Slide

13. No MySQL module
    for node.js
    early 2010
    

    View Slide

14. All we had was
    NoSQL Libraries
    early 2010
    

    View Slide

15. View Slide

16. Pure JS / No C/C++
    

    View Slide

17. Before Buffers
    became usable
    

    View Slide

18. The Parser was using
    JavaScript Strings
    

    View Slide

19. Node.js Trivia
    

    View Slide

20. “ Buffers” used to be
    called “Blobs”
    

    View Slide

21. For 3 min and 15 sec
    

    View Slide

22. RIP Blobs
    ✞
    

    View Slide

23. Sun Dec 13 08:39:20 2009
    -
    Sun Dec 13 08:42:45 2009
    

    View Slide

24. Anyway
    

    View Slide

25. mysql can be done
    without libmysql
    

    View Slide

26. View Slide

27. No good deed goes
    unpunished
    

    View Slide

28. Sir Isaac Newton
    

    View Slide

29. Third Law of Motion
    

    View Slide

30. “When a first body exerts a force F1
    on a second body, the second body
    simultaneously exerts a force F2 = −F1
    on the first body. This means that F1
    and F2 are equal in magnitude and
    opposite in direction.”
    

    View Slide

31. Third Law of Github
    

    View Slide

32. “When a first person pushes a
    library L1 into a remote repository,
    a second person simultaneously
    starts working on a second library
    L2 which will be equally awesome,
    but in a different way.”
    

    View Slide

33. <3 Github!
    

    View Slide

34. View Slide

35. Benchmark
    • Parse ~180 MB / 100.000 rows of MySQL result data
    • 5 Columns: id, title, text, created, updated
    • -> create 100k objects with 500k keys + 500k values
    

    View Slide

36. 0
    500
    1,000
    1,500
    mysql−0.9.6 mysql−libmysqlclient−1.5.1
    benchmark
    mbit
    benchmark
    mysql−0.9.6
    mysql−libmysqlclient−1.5.1
    

    View Slide

37. Of course.
    

    View Slide

38. libmysql = C
    

    View Slide

39. my library = JavaScript
    

    View Slide

40. C > JS, right?
    

    View Slide

41. But V8!
    

    View Slide

42. And Crankshaft!!
    

    View Slide

43. Node.js !!!1!
    

    View Slide

44. Was I
    living a lie?
    

    View Slide

45. Kind of
    

    View Slide

46. V8 / Node
    = Tools
    

    View Slide

47. Performance
    is not a tool
    

    View Slide

48. Performance is
    hard work &
    data analysis
    

    View Slide

49. 0
    500
    1,000
    1,500
    mysql−0.9.6 mysql−libmysqlclient−1.5.1mysql−2.0.0−alpha3
    benchmark
    mbit
    benchmark
    mysql−0.9.6
    mysql−libmysqlclient−1.5.1
    mysql−2.0.0−alpha3
    

    View Slide

50. Third Law of Github
    

    View Slide

51. View Slide

52. 0
    1,000
    2,000
    3,000
    4,000
    mysql−0.9.6
    mysql−libmysqlclient−1.5.1
    mysql−2.0.0−alpha3 mariadb−0.1.7
    benchmark
    mbit
    benchmark
    mysql−0.9.6
    mysql−libmysqlclient−1.5.1
    mysql−2.0.0−alpha3
    mariadb−0.1.7
    

    View Slide

53. Time to give up?
    

    View Slide

54. NEVER!
    

    View Slide

55. New Parser
    

    View Slide

56. 0
    2,000
    4,000
    6,000
    mysql2 new−parser
    benchmark
    mbit
    benchmark
    mysql2
    new−parser
    

    View Slide

57. Third law of Github?
    

    View Slide

58. Endgame
    

    View Slide

59. Last bottleneck:
    Creating JS Objects
    

    View Slide

60. Also: MySQL Server saturated
    

    View Slide

61. Anyway
    

    View Slide

62. How to write fast JS
    

    View Slide

63. Does not work
    

    View Slide

64. Profiling
    • Good for spotting small functions with stupid
    algorithms performing many iterations
    • Bad for complex functions with many
    primitive operations
    

    View Slide

65. Taking performance advice
    from strangers
    • Good for ideas & inspiration
    • But useless when applied cargo-cult style
    

    View Slide

66. Does Work
    

    View Slide

67. Benchmark Driven
    Development
    

    View Slide

68. Benchmark Driven Development
    • Similar to test driven development
    • Use it when performance is an explicit design goal
    • Benchmark first > benchmark after !
    

    View Slide

69. 1 function benchmark() {
    2 // intentionally empty
    3 }
    

    View Slide

70. 1 while (true) {
    2 var start = Date.now();
    3 benchmark();
    4 var duration = Date.now() - start;
    5 console.log(duration);
    6 }
    

    View Slide

71. Benchmark Driven Development
    • Next step: Implement a tiny part of your function
    • Example: Parse headers of MySQL packets
    • Look at impact, tweak code, repeat
    

    View Slide

72. Example Results
    • try...catch is ok
    • big switch statement = bad
    • function calls = very cheap
    • buffering is ok
    

    View Slide

73. Favorite
    

    View Slide

74. 1 function parseRow(columns, parser) {
    2 var row = {};
    3 for (var i = 0; i < columns.length; i++) {
    4 row[columns[i].name] = parser.readColumnValue();
    5 }
    6 return row;
    7 }
    

    View Slide

75. Make it faster!
    

    View Slide

76. 1 var code = 'return {\n';
    2
    3 columns.forEach(function(column) {
    4 code += '"' + column.name + '":' + 'parser.readColumnValue(),\n';
    5 });
    6
    7 code += '};\n';
    8
    9 var parseRow = new Function('columns', 'parser', code);
    

    View Slide

77. ->
    

    View Slide

78. 1 function parseRow(columns, parser) {
    2 return {
    3 id : parser.readColumnValue(),
    4 title : parser.readColumnValue(),
    5 body : parser.readColumnValue(),
    6 created : parser.readColumnValue(),
    7 updated : parser.readColumnValue(),
    8 };
    9 }
    

    View Slide

79. eval = awesome
    

    View Slide

80. eval = awesome
    

    View Slide

81. new Function = awesome
    

    View Slide

82. Data Analysis
    

    View Slide

83. Data analysis
    • Produce data points as tab separated values
    • Add as many VM/OS metrics as you can get to every line
    • Do not mix data and analysis !!
    

    View Slide

84. Recommended Tools
    • node benchmark.js | tee results.tsv
    • R Programming language (with ggplot2) !
    • Makefiles, Image Magick, Skitch
    

    View Slide

85. Why?
    

    View Slide

86. 0
    2,000
    4,000
    6,000
    mysql2 new−parser
    benchmark
    mbit
    benchmark
    mysql2
    new−parser
    

    View Slide

87. ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ● ●
    ●
    ● ● ●
    ● ● ●
    ● ●
    ● ● ●
    ●
    ●
    ● ●
    ● ● ●●
    ● ● ●
    ● ● ● ●
    ● ●
    ● ●
    ● ●
    ● ●
    ● ●
    ● ●
    ● ●
    ●
    ●
    ● ● ●
    ● ●
    ● ● ● ●
    ●
    ● ●
    ● ● ●
    ● ● ●
    ● ●
    ●
    ● ●
    ● ● ●
    ● ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ● ●
    ● ●
    ●
    ● ● ●
    ● ●
    ●
    ●
    ●
    ●
    ●
    ● ●
    ● ●
    ● ●
    ● ● ●
    ●
    ● ● ●
    ●
    ● ●
    ●
    ● ● ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ● ●
    ●
    ● ●
    ● ● ●
    ●
    ●
    ● ●
    ● ●
    ● ●
    ● ● ●●
    ● ●
    ●
    ● ●
    ● ●
    ● ● ●
    ●● ● ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●● ● ●
    ●
    ● ● ●
    ● ●
    ●
    ●
    ●
    ●
    ● ● ● ●
    ●
    ●
    ●
    ●
    ●
    ●
    ● ●
    ● ●
    ● ● ●
    ●
    ● ● ●
    ● ●
    ● ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ● ● ●
    ● ●
    ●
    ●
    ●
    ●
    ●
    ●
    ● ●
    ● ●
    ●
    ● ●
    ● ●
    ●
    ●
    ●
    ● ● ● ●
    ●
    ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ● ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ●
    ●●
    ●
    ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ● ●
    ● ● ●
    ● ●
    ●
    ●
    ●
    ● ● ●
    ● ●●
    ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ●
    ●
    ●
    ●
    ● ●●
    ●
    ●
    ●
    ● ●
    ● ●
    ●
    ● ●
    ●
    ●
    ●
    ●
    ● ●
    ● ●
    ●
    ● ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ● ● ●
    ●
    ●
    ●
    ●
    ●
    ● ●
    ● ● ●
    ●
    ●
    ●
    ●
    ●
    ●
    ● ●
    ●
    ● ●
    ●
    ●
    ●
    ● ● ● ●
    ●
    ● ●
    ●
    ● ●
    ● ● ● ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ●
    ● ●
    ●
    ●
    ●
    ●
    ● ● ●
    ●
    ●
    ●
    ●
    ●
    ● ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ●
    ●
    ● ● ●
    ● ● ●
    ●
    ● ●
    ● ●
    ● ● ●
    ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ● ●
    ●
    ●
    ●
    ●
    ●
    ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ●
    ●
    ●
    ●
    ● ● ●
    ●
    ●
    ●
    ●
    ● ● ● ●
    ●
    ● ●
    ●
    ●
    ● ●
    ●
    ● ●●
    ● ● ●
    ● ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ●
    ● ● ● ●
    ●
    ● ●
    ● ●
    ●
    ●
    ● ●
    ● ●
    ● ●
    ● ●
    ●
    ● ● ●
    ●
    ●
    ●
    ●
    ●
    ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ●
    ● ●
    3,000
    4,000
    5,000
    6,000
    mysql2 new−parser
    benchmark
    mbit
    benchmark
    ●
    ●
    mysql2
    new−parser
    

    View Slide

88. ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ● ●
    ●
    ● ● ●
    ● ● ●
    ● ●
    ● ● ●
    ●
    ●
    ● ●
    ● ● ●●
    ● ● ●
    ● ● ● ●
    ● ●
    ● ●
    ● ●
    ● ●
    ● ●
    ● ●
    ● ●
    ●
    ●
    ● ● ●
    ● ●
    ● ● ● ●
    ●
    ● ●
    ● ● ●
    ● ● ●
    ● ●
    ●
    ● ●
    ● ● ●
    ● ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ● ●
    ● ●
    ●
    ● ● ●
    ● ●
    ●
    ●
    ●
    ●
    ●
    ● ●
    ● ●
    ● ●
    ● ● ●
    ●
    ● ● ●
    ●
    ● ●
    ●
    ● ● ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ● ●
    ●
    ● ●
    ● ● ●
    ●
    ●
    ● ●
    ● ●
    ● ●
    ● ● ●●
    ● ●
    ●
    ● ●
    ● ●
    ● ● ●
    ●● ● ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●● ● ●
    ●
    ● ● ●
    ● ●
    ●
    ●
    ●
    ●
    ● ● ● ●
    ●
    ●
    ●
    ●
    ●
    ●
    ● ●
    ● ●
    ● ● ●
    ●
    ● ● ●
    ● ●
    ● ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ● ● ●
    ● ●
    ●
    ●
    ●
    ●
    ●
    ●
    ● ●
    ● ●
    ●
    ● ●
    ● ●
    ●
    ●
    ●
    ● ● ● ●
    ●
    ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ● ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ●
    ●●
    ●
    ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ● ●
    ● ● ●
    ● ●
    ●
    ●
    ●
    ● ● ●
    ● ●●
    ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ●
    ●
    ●
    ●
    ● ●●
    ●
    ●
    ●
    ● ●
    ● ●
    ●
    ● ●
    ●
    ●
    ●
    ●
    ● ●
    ● ●
    ●
    ● ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ● ● ●
    ●
    ●
    ●
    ●
    ●
    ● ●
    ● ● ●
    ●
    ●
    ●
    ●
    ●
    ●
    ● ●
    ●
    ● ●
    ●
    ●
    ●
    ● ● ● ●
    ●
    ● ●
    ●
    ● ●
    ● ● ● ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ●
    ● ●
    ●
    ●
    ●
    ●
    ● ● ●
    ●
    ●
    ●
    ●
    ●
    ● ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ●
    ●
    ● ● ●
    ● ● ●
    ●
    ● ●
    ● ●
    ● ● ●
    ●
    ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ● ●
    ●
    ●
    ●
    ●
    ●
    ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ●
    ●
    ●
    ●
    ● ● ●
    ●
    ●
    ●
    ●
    ● ● ● ●
    ●
    ● ●
    ●
    ●
    ● ●
    ●
    ● ●●
    ● ● ●
    ● ●
    ● ●
    ●
    ● ●
    ●
    ● ●
    ●
    ●
    ● ● ● ●
    ●
    ● ●
    ● ●
    ●
    ●
    ● ●
    ● ●
    ● ●
    ● ●
    ●
    ● ● ●
    ●
    ●
    ●
    ●
    ●
    ●
    ●
    ●
    ●
    ● ●
    ●
    ●
    ●
    ● ●
    3,000
    4,000
    5,000
    6,000
    mysql2 new−parser
    benchmark
    mbit
    benchmark
    ●
    ●
    mysql2
    new−parser
    Dafuq?
    Dafuq?
    

    View Slide

89. 3,000
    4,000
    5,000
    6,000
    0 100 200 300
    number
    mbit
    benchmark
    mysql2
    new−parser
    

    View Slide

90. 10
    20
    30
    0 100 200 300
    number
    Heap Total (MB)
    benchmark
    mysql2
    new−parser
    

    View Slide

91. 5
    10
    15
    0 100 200 300
    number
    Heap Used (MB)
    benchmark
    mysql2
    new−parser
    

    View Slide

92. tl;dr
    

    View Slide

93. 1. Write a benchmark
    2. Write/change a little code
    3. Collect data
    4. Find problems
    5. Goto 2
    

    View Slide

94. Thank you
    

    View Slide

95. github.com/felixge/faster-than-c
    All benchmarks, results and analysis scripts
    Thank you !
    

    View Slide

96. View Slide

97. View Slide

98. nodecopter.com
    Dublin, Oct 20
    (Saturday)
    Brighton, Nov 10
    http://tinyurl.com/brcopter
    

    View Slide