Don't Forget The Network: Your App Is Slower Than You Think (RailsConf 2016)

When you look at your response times, satisfied that they are "fast enough", you're forgetting an important thing: your users are on the other side of a network connection, and their browser has to process and render the data that you sent so quickly. This talk examines some often overlooked parts of web applications that can destroy your user experience even when your response times seem fantastic. We'll talk about networks, routing, client and server-side VMs, and how to measure and mitigate their issues.

Transcript

1. don't forget
   the network
   your app is slower
   than you think
   

   View Slide

2. André Arko
   @indirect
   

   View Slide

3. The Ruby Way
   therubyway.io
   

   View Slide

4. View Slide

5. View Slide

6. View Slide

7. Routing
   

   View Slide

8. Routing
   your app has this
   

   View Slide

9. Routing
   how does it work?
   

   View Slide

10. Development
    App
    You
    

    View Slide

11. Production
    People Router
    Server
    App
    App
    Router
    Server
    App
    App
    Router
    

    View Slide

12. Routing
    how slow is it?
    

    View Slide

13. Routing
    does it back up?
    

    View Slide

14. Request time
    

    View Slide

15. Request time
    not the time
    you measure
    

    View Slide

16. Request time
    wall-clock time
    from real clients
    

    View Slide

17. Request time
    make requests from
    around the world
    

    View Slide

18. Servers
    

    View Slide

19. Servers
    you have no idea
    what is going on
    

    View Slide

20. really.
    

    View Slide

21. Runtime lag
    

    View Slide

22. Runtime lag
    how do you tell you
    lost consciousness?
    

    View Slide

23. Runtime lag
    you have it.
    

    View Slide

24. Runtime lag
    you have it.
    how bad is it?
    

    View Slide

25. VM lag
    

    View Slide

26. VM lag
    do you have it?
    

    View Slide

27. VM lag
    do you check for it?
    

    View Slide

28. VM lag
    do you know how
    to check for it?
    

    View Slide

29. Metrics
    

    View Slide

30. Metrics
    are important
    

    View Slide

31. Metrics
    tell you what
    is happening
    

    View Slide

32. 
    you rn →
    

    View Slide

33. Metrics
    convince you
    you understand
    

    View Slide

34. 
    you later →
    

    View Slide

35. Averages
    convince you
    you understand
    

    View Slide

36. Averages
    are lie-candy
    for your brain
    

    View Slide

37. “Normal”
    5
    -5 -4 -3 -2 -1 0 1 2 3 4
    0
    0.1
    0.2
    0.3
    0.4
    

    View Slide

38. “Normal”
    5
    -5 -4 -3 -2 -1 0 1 2 3 4
    0
    0.1
    0.2
    0.3
    0.4
    

    View Slide

39. Real Life
    5
    -5 -4 -3 -2 -1 0 1 2 3 4
    0
    0.1
    0.2
    0.3
    0.4
    

    View Slide

40. brendangregg.com
    

    View Slide

41. brendangregg.com
    

    View Slide

42. 
    just heard
    “w
    e
    have
    a
    great average” →
    

    View Slide

43. The problem with averages:
    If you put one hand in a bucket of ice
    and the other in a bucket of hot coals,
    on average, you’re comfortable.
    Erik Michaels-Ober
    @sferik
    

    View Slide

44. Averages
    mask problems
    

    View Slide

45. 10
    0 1 2 3 4 5 6 7 8 9
    250
    0
    50
    100
    150
    200
    

    View Slide

46. Graph
    the median
    

    View Slide

47. 10
    0 1 2 3 4 5 6 7 8 9
    250
    0
    50
    100
    150
    200
    

    View Slide

48. Graph
    95th percentile
    

    View Slide

49. 10
    0 1 2 3 4 5 6 7 8 9
    250
    0
    50
    100
    150
    200
    

    View Slide

50. Graph
    99th percentile
    

    View Slide

51. 10
    0 1 2 3 4 5 6 7 8 9
    1000
    0
    250
    500
    750
    

    View Slide

52. Aggregate graphs
    another average
    

    View Slide

53. View Slide

54. Breakout graphs
    show each source
    

    View Slide

55. View Slide

56. Seriously, do it
    Visualize your data
    

    View Slide

57. graphic by Schutz and Avenue, CC-Attribution-ShareAlike, taken from from the Wikipedia article on Anscombe's quartet
    

    View Slide

58. Average of X: 9 Average of X: 9
    Average of X: 9 Average of X: 9
    

    View Slide

59. Average of Y: 7.50 Average of Y: 7.50
    Average of Y: 7.50 Average of Y: 7.50
    

    View Slide

60. Average of X
    Average of Y
    Variance of X
    Variance of Y
    Correlation of X and Y
    Linear regression
    All four data sets have the same
    

    View Slide

61. Aggregate alerts
    more dead servers
    than alive servers
    

    View Slide

62. 
    site’s up if any
    servers are up!
    

    View Slide

63. Breakout alerts
    first dead server
    not all the servers
    

    View Slide

64. So, in the end
    

    View Slide

65. don’t forget
    the network
    

    View Slide

66. user experience
    is all that matters
    after deploying
    

    View Slide

67. @indirect
    andre@arko.net
    Questions?
    

    View Slide