Tradeoffs, Bad Science, and Polar Bears – The World of Java Optimisation

Transcript

1. tradeoffs, bad science, and polar bears:
   
   
   the world of java optimisation
   
   
   Holly Cummins
   
   
   Red Hat
   
   
   
   
   Devoxx Belgium
   
   
   October 5, 2023
   
   
   

   View Slide

2. #Quarkus @[email protected]
   why optimise?
   

   View Slide

3. #Quarkus @[email protected]
   why optimise?
   

   View Slide

4. #Quarkus @[email protected]
   why optimise?
   0.5s extra search
   page time
   

   View Slide

5. #Quarkus @[email protected]
   why optimise?
   0.5s extra search
   page time
   20% drop in
   traffic
   

   View Slide

6. #Quarkus @[email protected]
   why optimise?
   0.5s extra search
   page time
   20% drop in
   traffic
   100 ms latency on page
   load
   

   View Slide

7. #Quarkus @[email protected]
   why optimise?
   0.5s extra search
   page time
   20% drop in
   traffic
   100 ms latency on page
   load
   7% lower
   conversion rate
   

   View Slide

8. #Quarkus @[email protected]
   why optimise?
   0.5s extra search
   page time
   20% drop in
   traffic
   100 ms latency on page
   load
   7% lower
   conversion rate
   

   View Slide

9. #Quarkus @[email protected]
   why optimise?
   0.5s extra search
   page time
   20% drop in
   traffic
   10 ms delay in trading
   platform
   100 ms latency on page
   load
   7% lower
   conversion rate
   

   View Slide

10. #Quarkus @[email protected]
    why optimise?
    0.5s extra search
    page time
    20% drop in
    traffic
    10 ms delay in trading
    platform
    10% drop in
    revenue
    100 ms latency on page
    load
    7% lower
    conversion rate
    

    View Slide

11. #Quarkus @holly_cummins
    what is optimising?
    

    View Slide

12. #Quarkus @[email protected]
    for whom?
    
    
    when?
    
    
    doing what?
    “make it go faster”
    

    View Slide

13. #Quarkus @holly_cummins
    

    View Slide

14. #Quarkus @holly_cummins
    performance can be:
    

    View Slide

15. #Quarkus @holly_cummins
    performance can be:
    throughput
    

    View Slide

16. #Quarkus @holly_cummins
    performance can be:
    throughput
    transactions
    per second
    

    View Slide

17. #Quarkus @holly_cummins
    performance can be:
    throughput
    latency
    transactions
    per second
    

    View Slide

18. #Quarkus @holly_cummins
    performance can be:
    throughput
    latency
    transactions
    per second
    start-up
    time
    

    View Slide

19. #Quarkus @holly_cummins
    performance can be:
    throughput
    latency
    transactions
    per second
    response
    time
    start-up
    time
    

    View Slide

20. #Quarkus @holly_cummins
    performance can be:
    throughput
    latency
    ramp-up
    time
    transactions
    per second
    response
    time
    start-up
    time
    

    View Slide

21. #Quarkus @holly_cummins
    performance can be:
    throughput
    latency
    capacity
    ramp-up
    time
    transactions
    per second
    response
    time
    start-up
    time
    

    View Slide

22. #Quarkus @holly_cummins
    performance can be:
    throughput
    latency
    capacity
    ramp-up
    time
    transactions
    per second
    response
    time
    start-up
    time
    bandwidth
    

    View Slide

23. #Quarkus @holly_cummins
    performance can be:
    throughput
    latency
    capacity
    ramp-up
    time
    transactions
    per second
    response
    time
    start-up
    time
    footprint
    bandwidth
    

    View Slide

24. #Quarkus @holly_cummins
    performance can be:
    throughput
    latency
    capacity
    ramp-up
    time
    transactions
    per second
    response
    time
    start-up
    time
    CPU usage
    footprint
    bandwidth
    

    View Slide

25. #Quarkus @holly_cummins
    performance can be:
    throughput
    latency
    capacity
    utilisation
    ramp-up
    time
    transactions
    per second
    response
    time
    start-up
    time
    CPU usage
    footprint
    bandwidth
    

    View Slide

26. #Quarkus @holly_cummins
    performance can be:
    throughput
    latency
    capacity
    utilisation
    …
    ramp-up
    time
    transactions
    per second
    response
    time
    start-up
    time
    CPU usage
    footprint
    bandwidth
    

    View Slide

27. #Quarkus @holly_cummins
    performance can be:
    throughput
    latency
    capacity
    utilisation
    …
    ramp-up
    time
    transactions
    per second
    response
    time
    start-up
    time
    CPU usage
    footprint
    bandwidth
    

    View Slide

28. #Quarkus @[email protected]
    Never underestimate the bandwidth [throughput] of a
    station wagon full of tapes hurtling down the highway.
    
    
    –Andrew Tanenbaum, 1981
    

    View Slide

29. #Quarkus @[email protected]
    Never underestimate the bandwidth [throughput] of a
    station wagon full of tapes hurtling down the highway.
    
    
    –Andrew Tanenbaum, 1981
    but the latency is terrible …
    
    
    

    View Slide

30. @holly_cummins
    requirements change
    

    View Slide

31. @holly_cummins
    

    View Slide

32. @holly_cummins
    

    View Slide

33. @holly_cummins
    

    View Slide

34. @holly_cummins
    

    View Slide

35. @holly_cummins
    I am not
    designed for this.
    

    View Slide

36. @holly_cummins
    the world changes
    

    View Slide

37. #Quarkus @[email protected]
    

    View Slide

38. #Quarkus @[email protected]
    -Xmx == $
    

    View Slide

39. #Quarkus @[email protected]
    -Xmx == $
    footprint
    

    View Slide

40. #Quarkus @[email protected]
    

    View Slide

41. #Quarkus @[email protected]
    which performs better?
    

    View Slide

42. #Quarkus @[email protected]
    quarkus + graalvm
    trading-off flexibility + throughput
    against startup speed and footprint
    

    View Slide

43. #Quarkus @[email protected]
    quarkus + graalvm
    trading-off flexibility + throughput
    against startup speed and footprint
    uhh … are you supposed
    to shut down applications
    after using them?
    

    View Slide

44. #Quarkus @[email protected]
    aside: quarkus + jvm
    

    View Slide

45. #Quarkus @[email protected]
    aside: quarkus + jvm
    better startup time
    

    View Slide

46. #Quarkus @[email protected]
    aside: quarkus + jvm
    better startup time
    better footprint
    

    View Slide

47. #Quarkus @[email protected]
    aside: quarkus + jvm
    better startup time
    better footprint
    better throughput
    

    View Slide

48. #Quarkus @[email protected]
    aside: quarkus + jvm
    better startup time
    better footprint
    better throughput
    better developer experience
    

    View Slide

49. #Quarkus @[email protected]
    aside: quarkus + jvm
    better startup time
    better footprint
    better throughput
    better developer experience
    

    View Slide

50. #Quarkus @[email protected]
    aside: quarkus + jvm
    better startup time
    better footprint
    better throughput
    better developer experience
    there is no tradeoff
    

    View Slide

51. #Quarkus @[email protected]
    aside: quarkus + jvm
    better startup time
    better footprint
    better throughput
    better developer experience
    there is no tradeoff
    (only elimination of waste)
    

    View Slide

52. #Quarkus @[email protected]
    aside: quarkus + jvm
    better startup time
    better footprint
    better throughput
    better developer experience
    there is no tradeoff
    (only elimination of waste)
    ok, there is a tradeoff: not optimising for dynamism nobody needs in the cloud
    

    View Slide

53. #Quarkus @[email protected]
    “waste”
    
    
    a trade-off where only one side has value
    

    View Slide

54. #Quarkus @holly_cummins
    recap:
    
    
    which is “faster?”
    

    View Slide

55. #Quarkus @holly_cummins
    GraalVM
    Quarkus
    Application
    recap:
    
    
    which is “faster?”
    

    View Slide

56. #Quarkus @holly_cummins
    OpenJDK GraalVM
    Quarkus Quarkus
    Application Application
    recap:
    
    
    which is “faster?”
    

    View Slide

57. #Quarkus @holly_cummins
    ephemeral or serverless
    OpenJDK GraalVM
    Quarkus Quarkus
    Application Application
    recap:
    
    
    which is “faster?”
    

    View Slide

58. #Quarkus @holly_cummins
    ephemeral or serverless
    OpenJDK GraalVM
    Quarkus Quarkus
    Application Application
    running your application
    for a long time
    recap:
    
    
    which is “faster?”
    

    View Slide

59. #Quarkus @holly_cummins
    ephemeral or serverless
    OpenJDK GraalVM
    Quarkus Quarkus
    Application Application
    running your application
    for a long time
    recap:
    
    
    which is “faster?”
    

    View Slide

60. @holly_cummins
    requirements aren’t
    what we think they are
    

    View Slide

61. #Quarkus @[email protected]
    behaviour at idle
    

    View Slide

62. #Quarkus @[email protected]
    “optimise for
    being unused”
    behaviour at idle
    

    View Slide

63. #Quarkus @[email protected]
    “optimise for
    being unused”
    useless
    behaviour at idle
    

    View Slide

64. #Quarkus @[email protected]
    30%
    
    
    of VMs are zombies
    antithesisgroup.com
    

    View Slide

65. #Quarkus @holly_cummins
    how to optimise?
    

    View Slide

66. #Quarkus @[email protected]
    find the bottleneck.
    fix it.
    

    View Slide

67. #IBM @holly_cummins
    what could possibly go wrong?!
    

    View Slide

68. #RedHat @holly_cummins
    intuition
    pitfall 1
    

    View Slide

69. #Quarkus @[email protected]
    this is not the place for ideas
    

    View Slide

70. #Quarkus @[email protected]
    measure, don’t guess.
    

    View Slide

71. #Quarkus @[email protected]
    measure the right thing
    

    View Slide

72. #Quarkus @[email protected]
    measure the right thing
    what do your users care about?
    

    View Slide

73. #RedHat @holly_cummins
    numbers
    pitfall 2
    (The McNamara Fallacy)
    

    View Slide

74. #Quarkus @[email protected]
    

    View Slide

75. #Quarkus @[email protected]
    leading indicators
    

    View Slide

76. #Quarkus @[email protected]
    leading indicators lagging indicators
    

    View Slide

77. #Quarkus @[email protected]
    leading indicators
    we care about them
    lagging indicators
    

    View Slide

78. #Quarkus @[email protected]
    leading indicators
    we care about them
    easy to measure
    lagging indicators
    

    View Slide

79. #Quarkus @[email protected]
    leading indicators
    we care about them
    easy to measure
    hard to change
    lagging indicators
    

    View Slide

80. #Quarkus @[email protected]
    leading indicators
    we care about them
    easy to measure
    hard to change
    lagging indicators
    easy to change
    

    View Slide

81. #Quarkus @[email protected]
    leading indicators
    we care about them
    easy to measure
    hard to change
    lagging indicators
    predictive of a thing we care about
    easy to change
    

    View Slide

82. #Quarkus @[email protected]
    leading indicators
    we care about them
    easy to measure
    hard to change
    lagging indicators
    predictive of a thing we care about
    hard to identify
    easy to change
    

    View Slide

83. #Quarkus @[email protected]
    leading indicators
    we care about them
    easy to measure
    hard to change
    lagging indicators
    predictive of a thing we care about
    hard to identify
    easy to change
    

    View Slide

84. #Quarkus @[email protected]
    caution:
    
    
    performance experiments for entertainment purposes only.
    
    
    do not try these at home.
    

    View Slide

85. #Quarkus @[email protected]
    2007
    

    View Slide

86. #Quarkus @[email protected]
    bad-ish advice: “reduce time spent in garbage collection”
    

    View Slide

87. #Quarkus @[email protected]
    bad-ish advice: “reduce time spent in garbage collection”
    actually, garbage collection can make your application go faster
    

    View Slide

88. #Quarkus @[email protected]
    2007
    

    View Slide

89. #Quarkus @[email protected]
    2007
    compacting every time
    
    
    gc overhead is huge
    

    View Slide

90. #Quarkus @[email protected]
    2007
    compacting every time
    
    
    gc overhead is huge
    

    View Slide

91. #Quarkus @[email protected]
    2007
    compacting every time
    
    
    gc overhead is huge
    throughput is … better?!
    

    View Slide

92. #Quarkus @[email protected]
    2021
    

    View Slide

93. #Quarkus @[email protected]
    2021
    

    View Slide

94. #Quarkus @[email protected]
    -verbose:gc
    -Xverbosegclog:gclog.xml
    -Xcompactgc
    

    View Slide

95. #Quarkus @[email protected]
    -verbose:gc
    -Xverbosegclog:gclog.xml
    -Xgcpolicy:optthruput
    -Xcompactgc
    

    View Slide

96. #Quarkus @[email protected]
    -verbose:gc
    -Xverbosegclog:gclog.xml
    -Xgcpolicy:optthruput
    -Xmx110m
    -Xms110m
    -Xnocompactgc
    

    View Slide

97. #Quarkus @[email protected]
    -verbose:gc
    -Xverbosegclog:gclog.xml
    -Xgcpolicy:optthruput
    -Xmx160m
    -Xms160m
    -Xnocompactgc
    

    View Slide

98. #Quarkus @[email protected]
    -verbose:gc
    -Xverbosegclog:gclog.xml
    -Xgcpolicy:optthruput
    -Xmx300m
    -Xms300m
    -Xcompactgc
    why does the
    performance stay exactly the
    same no matter what gc
    settings I choose?
    

    View Slide

99. #Quarkus @[email protected]
    by the way, this is cheating.
    (remember the ‘bad science’?)
    

    View Slide

100. #Quarkus @[email protected]
     -verbose:gc
     -Xverbosegclog:gclog.xml
     -Xgcpolicy:optthruput
     

     View Slide

101. #Quarkus @[email protected]
     -verbose:gc
     -Xverbosegclog:gclog.xml
     -Xgcpolicy:optthruput
     tool: GCMV
     

     View Slide

102. #Quarkus @[email protected]
     -verbose:gc
     -Xverbosegclog:gclog.xml
     -Xgcpolicy:optthruput
     total GC time: 21.6s
     tool: GCMV
     

     View Slide

103. #Quarkus @[email protected]
     -verbose:gc
     -Xverbosegclog:gclog.xml
     -Xgcpolicy:optthruput
     total GC time: 21.6s
     4.1% of time in GC pause
     tool: GCMV
     

     View Slide

104. #Quarkus @[email protected]
     -verbose:gc
     -Xverbosegclog:gclog.xml
     -Xgcpolicy:optthruput
     total GC time: 21.6s
     4.1% of time in GC pause
     total GC time: 12.0s
     tool: GCMV
     

     View Slide

105. #Quarkus @[email protected]
     -verbose:gc
     -Xverbosegclog:gclog.xml
     -Xgcpolicy:optthruput
     total GC time: 21.6s
     4.1% of time in GC pause
     total GC time: 12.0s
     3.6% of time in GC pause
     tool: GCMV
     

     View Slide

106. #Quarkus @[email protected]
     -verbose:gc
     -Xverbosegclog:gclog.xml
     -Xgcpolicy:optthruput
     total GC time: 21.6s
     4.1% of time in GC pause
     23.9 GB garbage collected
     total GC time: 12.0s
     3.6% of time in GC pause
     13.0 GB garbage collected
     tool: GCMV
     

     View Slide

107. #Quarkus @[email protected]
     -verbose:gc
     -Xverbosegclog:gclog.xml
     -Xgcpolicy:optthruput
     total GC time: 21.6s
     4.1% of time in GC pause
     23.9 GB garbage collected
     493 transactions/s
     total GC time: 12.0s
     3.6% of time in GC pause
     13.0 GB garbage collected
     260 transactions/s
     tool: GCMV
     

     View Slide

108. #Quarkus @[email protected]
     -verbose:gc
     -Xverbosegclog:gclog.xml
     -Xgcpolicy:optthruput
     total GC time: 21.6s
     4.1% of time in GC pause
     23.9 GB garbage collected
     493 transactions/s
     total GC time: 12.0s
     3.6% of time in GC pause
     13.0 GB garbage collected
     260 transactions/s
     tool: GCMV
     

     View Slide

109. #Quarkus @[email protected]
     -verbose:gc
     -Xverbosegclog:gclog.xml
     -Xgcpolicy:optthruput
     total GC time: 21.6s
     4.1% of time in GC pause
     23.9 GB garbage collected
     493 transactions/s
     total GC time: 12.0s
     3.6% of time in GC pause
     13.0 GB garbage collected
     260 transactions/s
     tool: GCMV
     

     View Slide

110. #Quarkus @holly_cummins
     total GC time: 21.6s
     4.1% of time in GC pause
     23.9 GB garbage collected
     493 transactions/s
     total GC time: 12.0s
     3.6% of time in GC pause
     13.0 GB garbage collected
     260 transactions/s
     

     View Slide

111. #Quarkus @holly_cummins
     total GC time: 21.6s
     4.1% of time in GC pause
     23.9 GB garbage collected
     493 transactions/s
     total GC time: 12.0s
     3.6% of time in GC pause
     13.0 GB garbage collected
     260 transactions/s
     

     View Slide

112. #Quarkus @holly_cummins
     total GC time: 21.6s
     4.1% of time in GC pause
     23.9 GB garbage collected
     493 transactions/s
     total GC time: 12.0s
     3.6% of time in GC pause
     13.0 GB garbage collected
     260 transactions/s
     leading indicator
     

     View Slide

113. #Quarkus @holly_cummins
     total GC time: 21.6s
     4.1% of time in GC pause
     23.9 GB garbage collected
     493 transactions/s
     total GC time: 12.0s
     3.6% of time in GC pause
     13.0 GB garbage collected
     260 transactions/s
     leading indicator
     

     View Slide

114. #Quarkus @holly_cummins
     total GC time: 21.6s
     4.1% of time in GC pause
     23.9 GB garbage collected
     493 transactions/s
     total GC time: 12.0s
     3.6% of time in GC pause
     13.0 GB garbage collected
     260 transactions/s
     leading indicator
     lagging indicator
     

     View Slide

115. #Quarkus @holly_cummins
     total GC time: 21.6s
     4.1% of time in GC pause
     23.9 GB garbage collected
     493 transactions/s
     total GC time: 12.0s
     3.6% of time in GC pause
     13.0 GB garbage collected
     260 transactions/s
     leading indicator
     lagging indicator
     ?
     

     View Slide

116. #Quarkus @holly_cummins
     total GC time: 21.6s
     4.1% of time in GC pause
     23.9 GB garbage collected
     493 transactions/s
     total GC time: 12.0s
     3.6% of time in GC pause
     13.0 GB garbage collected
     260 transactions/s
     leading indicator
     lagging indicator
     ?
     ?
     

     View Slide

117. #Quarkus @[email protected]
     so wait, what changed to make the app faster?
     running jmeter on the same machine as the app gives a big speedup!
     

     View Slide

118. #Quarkus @[email protected]
     “Any improvements made anywhere
     besides the bottleneck are an illusion.”
     
     
     – Gene Kim
     

     View Slide

119. #Quarkus @[email protected]
     time kills all performance advice
     (even mine)
     

     View Slide

120. #Quarkus @[email protected]
     zgc Brian Goetz, yesterday
     

     View Slide

121. #Quarkus @[email protected]
     zgc Brian Goetz, yesterday
     

     View Slide

122. #Quarkus @[email protected]
     zgc Brian Goetz, yesterday
     

     View Slide

123. #Quarkus @[email protected]
     zgc Brian Goetz, yesterday
     

     View Slide

124. #Quarkus @[email protected]
     zgc Brian Goetz, yesterday
     tradeoff: throughput
     is ~2% lower
     

     View Slide

125. #Quarkus @[email protected]
     zgc Brian Goetz, yesterday
     tradeoff: throughput
     is ~2% lower
     tradeoff: memory is
     higher
     

     View Slide

126. #Quarkus @[email protected]
     zgc Brian Goetz, yesterday
     tradeoff: throughput
     is ~2% lower
     tradeoff: memory is
     higher
     java 21 adds generational zgc
     
     
     reduces zgc footprint by 75%
     

     View Slide

127. #Quarkus @[email protected]
     the takeaways:
     
     
     gc can improve performance by rearranging the heap
     
     
     find the bottleneck
     
     
     validate advice independently
     
     
     optimise the right thing … for you
     

     View Slide

128. #RedHat @holly_cummins
     advice
     pitfall 3
     

     View Slide

129. #Quarkus @[email protected]
     I read it on the
     internet!
     

     View Slide

130. #IBM @holly_cummins
     noooooo!
     “make one big method because method dispatching is slow”
     

     View Slide

131. #IBM @holly_cummins
     noooooo!
     “re-use your objects to help the garbage collector”
     

     View Slide

132. #IBM @holly_cummins
     noooooo!
     “to tune your JVM, use this command-line:”
     
     
     -server -Xms1g -Xmx1g -XX:PermSize=1g -XX:MaxPermSize=256m -Xmn256m
     -Xss64k -XX:SurvivorRatio=30 -XX:+UseConcMarkSweepGC -XX:
     +CMSParallelRemarkEnabled -XX:+UseCMSInitiatingOccupancyOnly
     -XX:CMSInitiatingOccupancyFraction=10 -XX:+ScavengeBeforeFullGC -XX:
     +CMSScavengeBeforeRemark -XX:+PrintGCDateStamps -verbose:gc -XX:
     +PrintGCDetails -Dsun.net.inetaddr.ttl=5 -XX:+HeapDumpOnOutOfMemoryError
     -XX:HeapDumpPath=`date`.hprof -Dcom.sun.management.jmxremote.port=5616
     -Dcom.sun.management.jmxremote.authenticate=false
     -Dcom.sun.management.jmxremote.ssl=false -server -Xms2g -Xmx2g
     -XX:MaxPermSize=256m -XX:NewRatio=1 -XX:+UseConcMarkSweepGC
     

     View Slide

133. #IBM @holly_cummins
     noooooo!
     use StringBuilder, never concatenate strings with +=
     

     View Slide

134. #IBM @holly_cummins
     noooooo!
     use StringBuilder, never concatenate strings with +=
     wait, what?
     
     
     yes, right?
     

     View Slide

135. #Quarkus @[email protected]
     remember, measure, don’t guess.
     

     View Slide

136. #Quarkus @[email protected]
     2 things ruin advice:
     
     
     • context
     
     
     • time
     

     View Slide

137. #RedHat @holly_cummins
     micro-optimisation theatre
     pitfall 4
     

     View Slide

138. #Quarkus @[email protected]
     static string beSlow()
     {
     string result = "";
     for (int i = 0; i < 314159; i++)
     {
     result += getStringData(i);
     }
     return result;
     }
     

     View Slide

139. #Quarkus @[email protected]
     @Override
     public String toString() {
     String ret = "\n\tMarket Summary at: " + getSummaryDate() + “\n\t\t TSIA:" + getTSIA() + “\n\t\t openTSIA:" + getOpenTSIA()
     + “\n\t\t gain:" + getGainPercent() + “\n\t\t volume:" + getVolume();
     if ((getTopGainers() == null) || (getTopLosers() == null)) {
     return ret;
     }
     ret += "\n\t\t Current Top Gainers:";
     Iterator it = getTopGainers().iterator();
     while (it.hasNext()) {
     QuoteDataBean quoteData = it.next();
     ret += ("\n\t\t\t" + quoteData.toString());
     }
     ret += "\n\t\t Current Top Losers:";
     it = getTopLosers().iterator();
     while (it.hasNext()) {
     QuoteDataBean quoteData = it.next();
     ret += ("\n\t\t\t" + quoteData.toString());
     }
     return ret;
     }
     

     View Slide

140. #Quarkus @[email protected]
     @Override
     public String toString() {
     String ret = "\n\tMarket Summary at: " + getSummaryDate() + “\n\t\t TSIA:" + getTSIA() + “\n\t\t openTSIA:" + getOpenTSIA()
     + “\n\t\t gain:" + getGainPercent() + “\n\t\t volume:" + getVolume();
     if ((getTopGainers() == null) || (getTopLosers() == null)) {
     return ret;
     }
     ret += "\n\t\t Current Top Gainers:";
     Iterator it = getTopGainers().iterator();
     while (it.hasNext()) {
     QuoteDataBean quoteData = it.next();
     ret += ("\n\t\t\t" + quoteData.toString());
     }
     ret += "\n\t\t Current Top Losers:";
     it = getTopLosers().iterator();
     while (it.hasNext()) {
     QuoteDataBean quoteData = it.next();
     ret += ("\n\t\t\t" + quoteData.toString());
     }
     return ret;
     }
     

     View Slide

141. #Quarkus @[email protected]
     @Override
     public String toString() {
     String ret = "\n\tMarket Summary at: " + getSummaryDate() + “\n\t\t TSIA:" + getTSIA() + “\n\t\t openTSIA:" + getOpenTSIA()
     + “\n\t\t gain:" + getGainPercent() + “\n\t\t volume:" + getVolume();
     if ((getTopGainers() == null) || (getTopLosers() == null)) {
     return ret;
     }
     ret += "\n\t\t Current Top Gainers:";
     Iterator it = getTopGainers().iterator();
     while (it.hasNext()) {
     QuoteDataBean quoteData = it.next();
     ret += ("\n\t\t\t" + quoteData.toString());
     }
     ret += "\n\t\t Current Top Losers:";
     it = getTopLosers().iterator();
     while (it.hasNext()) {
     QuoteDataBean quoteData = it.next();
     ret += ("\n\t\t\t" + quoteData.toString());
     }
     return ret;
     }
     

     View Slide

142. #Quarkus @[email protected]
     @Override
     public String toString() {
     String ret = "\n\tMarket Summary at: " + getSummaryDate() + “\n\t\t TSIA:" + getTSIA() + “\n\t\t openTSIA:" + getOpenTSIA()
     + “\n\t\t gain:" + getGainPercent() + “\n\t\t volume:" + getVolume();
     if ((getTopGainers() == null) || (getTopLosers() == null)) {
     return ret;
     }
     ret += "\n\t\t Current Top Gainers:";
     Iterator it = getTopGainers().iterator();
     while (it.hasNext()) {
     QuoteDataBean quoteData = it.next();
     ret += ("\n\t\t\t" + quoteData.toString());
     }
     ret += "\n\t\t Current Top Losers:";
     it = getTopLosers().iterator();
     while (it.hasNext()) {
     QuoteDataBean quoteData = it.next();
     ret += ("\n\t\t\t" + quoteData.toString());
     }
     return ret;
     }
     this never gets called
     

     View Slide

143. #Quarkus @[email protected]
     let’s make my travel energy-efficient
     

     View Slide

144. #Quarkus @[email protected]
     every little helps?
     

     View Slide

145. #Quarkus @[email protected]
     every little helps?
     every optimisation is another optimisation you aren’t doing
     

     View Slide

146. #Quarkus @[email protected]
     our platforms help
     

     View Slide

147. #Quarkus @[email protected]
     static string beSlow()
     {
     string result = "";
     for (int i = 0; i < 314159; i++)
     {
     result += getStringData(i);
     }
     return result;
     }
     

     View Slide

148. #Quarkus @[email protected]
     static string beSlow()
     {
     string result = “”;
     result += getStringData(1);
     result += getStringData(2);
     result += getStringData(3);
     return result;
     }
     

     View Slide

149. #Quarkus @[email protected]
     static string beSlow()
     {
     string result = “”;
     result += getStringData(1);
     result += getStringData(2);
     result += getStringData(3);
     return result;
     }
     this is fine
     

     View Slide

150. #Quarkus @[email protected]
     the JVM writers have far more time for optimising than you do
     clean, typical, code runs best
     

     View Slide

151. #Quarkus @[email protected]
     but …
     

     View Slide

152. #Quarkus @[email protected]
     why did the quarkus 2.0.0 codebase use stringbuilder so much?
     but …
     

     View Slide

153. #Quarkus @[email protected]
     why did the quarkus 2.0.0 codebase use stringbuilder so much?
     remember, the outside world can change
     but …
     

     View Slide

154. #Quarkus @[email protected]
     remember, measure, don’t guess.
     

     View Slide

155. #RedHat @holly_cummins
     the shiny stuff
     pitfall 5
     

     View Slide

156. #Quarkus @[email protected]
     loom
     

     View Slide

157. #Quarkus @[email protected]
     loom
     very good at waiting
     

     View Slide

158. #Quarkus @[email protected]
     loom
     very good at waiting
     not so good at cpu-bound
     tasks
     

     View Slide

159. #Quarkus @[email protected]
     loom
     very good at waiting
     not so good at cpu-bound
     tasks
     can interact badly with libraries
     

     View Slide

160. #Quarkus @[email protected]
     loom
     very good at waiting
     not so good at cpu-bound
     tasks
     can interact badly with libraries
     if a virtual thread gets pinned or
     does a long cpu process all its
     friends grind to a halt
     

     View Slide

161. #Quarkus @[email protected]
     is thread management
     really your bottleneck?
     loom is not a magic “faster thread”
     
     
     

     View Slide

162. #Quarkus @[email protected]
     remember, measure, don’t guess.
     
     
     

     View Slide

163. #Quarkus @[email protected]
     remember, measure, don’t guess.
     
     
     -Djdk.tracePinnedThreads
     

     View Slide

164. #Quarkus @[email protected]
     remember, measure, don’t guess.
     
     
     -Djdk.tracePinnedThreads
     @ShouldNotPin (in quarkus tests)
     

     View Slide

165. #Quarkus @[email protected]
     remember, measure, don’t guess.
     
     
     -Djdk.tracePinnedThreads
     @ShouldNotPin (in quarkus tests)
     

     View Slide

166. #Quarkus @[email protected]
     learn more
     

     View Slide

167. #Quarkus @holly_cummins
     ok, but how to optimise?
     

     View Slide

168. #Quarkus @[email protected]
     tools
     

     View Slide

169. #Quarkus @[email protected]
     “What you can optimize is limited to what
     you can observe.”
     
     
     -Susie Xia, Netflix
     

     View Slide

170. #Quarkus @[email protected]
     observability
     

     View Slide

171. #Quarkus @[email protected]
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     

     View Slide

172. #Quarkus @[email protected]
     method profiler
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     

     View Slide

173. #Quarkus @[email protected]
     method profiler
     VisualVM
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     

     View Slide

174. #Quarkus @[email protected]
     method profiler
     Mission Control
     VisualVM
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     

     View Slide

175. #Quarkus @[email protected]
     method profiler
     Mission Control
     VisualVM
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     IntelliJ Profiler
     

     View Slide

176. #Quarkus @[email protected]
     method profiler
     Mission Control
     VisualVM
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     IBM Health Center
     (for OpenJ9)
     IntelliJ Profiler
     

     View Slide

177. #Quarkus @[email protected]
     method profiler
     Mission Control
     flame graphs
     VisualVM
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     IBM Health Center
     (for OpenJ9)
     IntelliJ Profiler
     

     View Slide

178. #Quarkus @[email protected]
     method profiler
     GC analysis
     Mission Control
     flame graphs
     VisualVM
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     IBM Health Center
     (for OpenJ9)
     IntelliJ Profiler
     

     View Slide

179. #Quarkus @[email protected]
     method profiler
     GC analysis
     Mission Control
     flame graphs
     GCMV
     VisualVM
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     IBM Health Center
     (for OpenJ9)
     IntelliJ Profiler
     

     View Slide

180. #Quarkus @[email protected]
     method profiler
     GC analysis
     heap analysis
     Mission Control
     flame graphs
     GCMV
     VisualVM
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     IBM Health Center
     (for OpenJ9)
     IntelliJ Profiler
     

     View Slide

181. #Quarkus @[email protected]
     method profiler
     GC analysis
     heap analysis
     Mission Control
     flame graphs
     GCMV
     VisualVM
     Eclipse MAT
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     IBM Health Center
     (for OpenJ9)
     IntelliJ Profiler
     

     View Slide

182. #Quarkus @[email protected]
     method profiler
     GC analysis
     heap analysis
     APM
     Mission Control
     flame graphs
     GCMV
     VisualVM
     Eclipse MAT
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     IBM Health Center
     (for OpenJ9)
     IntelliJ Profiler
     

     View Slide

183. #Quarkus @[email protected]
     method profiler
     GC analysis
     heap analysis
     APM
     Mission Control
     flame graphs
     GCMV
     VisualVM
     Eclipse MAT
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     GlowRoot
     IBM Health Center
     (for OpenJ9)
     IntelliJ Profiler
     

     View Slide

184. #Quarkus @[email protected]
     method profiler
     GC analysis
     heap analysis
     APM
     Mission Control
     flame graphs
     GCMV
     New Relic*
     VisualVM
     Eclipse MAT
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     GlowRoot
     IBM Health Center
     (for OpenJ9)
     IntelliJ Profiler
     

     View Slide

185. #Quarkus @[email protected]
     method profiler
     GC analysis
     heap analysis
     APM
     Mission Control
     flame graphs
     GCMV
     New Relic*
     AppDynamics*
     VisualVM
     Eclipse MAT
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     GlowRoot
     IBM Health Center
     (for OpenJ9)
     IntelliJ Profiler
     

     View Slide

186. #Quarkus @[email protected]
     method profiler
     GC analysis
     heap analysis
     APM
     Mission Control
     flame graphs
     GCMV
     New Relic*
     AppDynamics*
     VisualVM
     Dynatrace*
     Eclipse MAT
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     GlowRoot
     IBM Health Center
     (for OpenJ9)
     IntelliJ Profiler
     

     View Slide

187. #Quarkus @[email protected]
     method profiler
     GC analysis
     heap analysis
     APM
     Mission Control
     flame graphs
     GCMV
     New Relic*
     AppDynamics*
     VisualVM
     Dynatrace*
     Eclipse MAT
     * not free
     
     
     this is an incomplete list, because there are a lot of tools out there, and many cost money
     GlowRoot
     IBM Health Center
     (for OpenJ9)
     IntelliJ Profiler
     

     View Slide

188. #Quarkus @[email protected]
     method profiler
     
     
     

     View Slide

189. #Quarkus @[email protected]
     method profiler
     
     
     

     View Slide

190. #Quarkus @[email protected]
     method profiler
     
     
     

     View Slide

191. #Quarkus @[email protected]
     method profiler
     
     
     

     View Slide

192. #Quarkus @[email protected]
     method profiler
     
     
     Open Telemetry
     

     View Slide

193. @holly_cummins
     you may need to know the whole system
     context to know what to optimise
     

     View Slide

194. #Quarkus @[email protected]
     “Nines don’t matter if your
     users aren’t happy.”
     
     
     – Charity Majors
     

     View Slide

195. #Quarkus @[email protected]
     don’t forget the edges
     queueing theory helps us understand
     where the disasters happen
     

     View Slide

196. #Quarkus @[email protected]
     “When it comes to IT performance, amateurs look
     at averages. Professionals look at distributions.”
     
     
     – Avishai Ish-Shalom
     

     View Slide

197. #Quarkus @[email protected]
     slow performance can turn into big cloud bills
     make cloud costs visible to engineers
     

     View Slide

198. #Quarkus @holly_cummins
     ok, but you promised bears
     

     View Slide

199. #Quarkus @[email protected]
     if you leave the TV on when
     you’re not using it, you’re a
     polar bear murderer
     

     View Slide

200. #Quarkus @[email protected]
     data centres use 1-2% of
     the world’s electricity
     

     View Slide

201. #Quarkus @[email protected]
     performance optimisation is carbon optimisation
     

     View Slide

202. #Quarkus @[email protected]
     cost: leading indicator for carbon
     
     
     performance: leading indicator for carbon
     

     View Slide

203. #Quarkus @[email protected]
     example:
     
     
     what our quarkus experiments showed
     

     View Slide

204. @holly_cummins #RedHat
     Source: Clement Escoffier
     cost impact of framework choice
     Setup:
     
     
     • 800 requests/second, over 20 days
     
     
     • SLA > 99%
     
     
     • AWS instances
     
     
     Assumptions:
     
     
     • Costs are for us-east-1 data centre
     

     View Slide

205. @holly_cummins #RedHat
     Setup:
     
     
     • 800 requests/second, over 20 days
     
     
     • SLA > 99%
     
     
     Assumptions:
     
     
     • 50% load
     
     
     • us-east-1 data centre
     
     
     • Teads dataset
     Source: Clement Escoffier x Teads
     cloud carbon impact of framework choice
     carbon impact of framework choice
     

     View Slide

206. @holly_cummins #RedHat
     Setup:
     
     
     • 800 requests/second, over 20 days
     
     
     • SLA > 99%
     
     
     Assumptions:
     
     
     • 50% load
     
     
     • us-east-1 data centre
     
     
     • Teads dataset
     Source: Clement Escoffier x Teads
     cloud carbon impact of framework choice
     carbon impact of framework choice
     economic model in action:
     
     
     the cost and carbon metrics are
     (roughly) the same
     

     View Slide

207. @holly_cummins #RedHat
     capacity
     Source: John O’Hara
     Setup:
     
     
     • REST + CRUD
     
     
     • large heap
     
     
     • RAPL energy measurement
     
     
     Assumptions:
     
     
     • US energy mix
     climate impact of framework choice
     

     View Slide

208. @holly_cummins #RedHat
     capacity
     Source: John O’Hara
     Setup:
     
     
     • REST + CRUD
     
     
     • large heap
     
     
     • RAPL energy measurement
     
     
     Assumptions:
     
     
     • US energy mix
     climate impact of framework choice
     shorter line means
     lower max throughput
     

     View Slide

209. @holly_cummins #RedHat
     capacity
     Source: John O’Hara
     Setup:
     
     
     • REST + CRUD
     
     
     • large heap
     
     
     • RAPL energy measurement
     
     
     Assumptions:
     
     
     • US energy mix
     climate impact of framework choice
     shorter line means
     lower max throughput
     higher line means worse
     carbon footprint
     

     View Slide

210. @holly_cummins #RedHat
     capacity
     Source: John O’Hara
     Setup:
     
     
     • REST + CRUD
     
     
     • large heap
     
     
     • RAPL energy measurement
     
     
     Assumptions:
     
     
     • US energy mix
     climate impact of framework choice
     vrrrooom model in action:
     
     
     quarkus on JVM has the smallest footprint …
     because it has the highest throughput
     shorter line means
     lower max throughput
     higher line means worse
     carbon footprint
     

     View Slide

211. #Quarkus @[email protected]
     

     View Slide

212. #Quarkus @[email protected]
     we should not waste
     
     
     • electricity
     
     
     • hardware
     

     View Slide

213. #Quarkus @[email protected]
     good performance makes us good.
     we should not waste
     
     
     • electricity
     
     
     • hardware
     

     View Slide

214. #Quarkus @[email protected]
     sooo …
     
     
     you can optimise, and it can be fun
     
     
     measure, don’t guess
     
     
     only optimise what matters
     @holly_cummins
     #Quarkus Slides
     

     View Slide