These Five Tricks Can Make Your Apps Greener, Cheaper, & Nicer

The code we write has a climate impact. But how big is that impact? How do we measure it? How do we reduce it? Is the cloud helping? What’s going on with Virginia? Are we still allowed to do CI/CD? Will native compilation save us? Is Java even a good choice anymore? This talk discusses some of the trade-offs for a modern software developer, and provides a roadmap to figuring out the right thing.

Disclosure: Holly works on Quarkus. Along the way, she will talk about Quarkus sustainability measurements we’ve been doing … but it’s ok, because she promises that the Quarkus carbon data is exciting and interesting.

Transcript

1. Holly Cummins
   
   
   Senior Principal Software Engineer, Quarkus
   
   
   J-Spring
   
   
   June 21, 2023
   
   
   @holly_cummins@hachyderm.io
   these five tricks can
   make your Java apps
   
   
   greener
   
   
   cheaper &
   
   
   nicer
   

   View Slide

2. @holly_cummins #RedHat
   Sources:
   
   
   https://www.greenit.fr/wp-content/uploads/2019/11/GREENIT_EENM_etude_EN_accessible.pdf
   
   
   https://ourworldindata.org/ghg-emissions-by-sector
   the digital world creates
   more carbon emissions
   than aviation
   

   View Slide

3. @holly_cummins #RedHat
   Lighter area represents high and low estimates, where available.
   
   
   Sources:
   
   
   https://www.iea.org/fuels-and-technologies/data-centres-networks
   
   
   https://ourworldindata.org/grapher/electricity-demand?tab=table&country=USA~GBR~FRA~DEU~IND~BRA
   data centres use as
   much electricity as a
   medium country
   

   View Slide

4. @holly_cummins #RedHat
   it’s not just artificial intelligence
   

   View Slide

5. @holly_cummins #RedHat
   it’s not just artificial intelligence
   it’s not just cryptocurrency mining
   

   View Slide

6. @holly_cummins #RedHat
   it’s not just artificial intelligence
   it’s not just cryptocurrency mining
   it’s all of us
   

   View Slide

7. @holly_cummins #RedHat
   aaaaaaaargh
   

   View Slide

8. @holly_cummins #RedHat
   aaaaaaaargh?
   

   View Slide

9. @holly_cummins #RedHat
   be a solutionist
   

   View Slide

10. @holly_cummins #RedHat
    how do we do make solutions?
    

    View Slide

11. @holly_cummins #RedHat
    green software foundation: principles
    

    View Slide

12. @holly_cummins #RedHat
    carbon
    awareness
    green software foundation: principles
    

    View Slide

13. @holly_cummins #RedHat
    carbon
    awareness
    green software foundation: principles
    where
    

    View Slide

14. @holly_cummins #RedHat
    carbon
    awareness
    green software foundation: principles
    where
    when
    

    View Slide

15. @holly_cummins #RedHat
    carbon
    awareness
    green software foundation: principles
    hardware
    efficiency
    where
    when
    

    View Slide

16. @holly_cummins #RedHat
    carbon
    awareness
    green software foundation: principles
    hardware
    efficiency
    where
    when
    elasticity
    

    View Slide

17. @holly_cummins #RedHat
    carbon
    awareness
    green software foundation: principles
    hardware
    efficiency
    where
    when
    elasticity
    utilisation
    

    View Slide

18. @holly_cummins #RedHat
    carbon
    awareness
    green software foundation: principles
    hardware
    efficiency
    electricity
    efficiency
    where
    when
    elasticity
    utilisation
    

    View Slide

19. @holly_cummins #RedHat
    algorithms
    carbon
    awareness
    green software foundation: principles
    hardware
    efficiency
    electricity
    efficiency
    where
    when
    elasticity
    utilisation
    

    View Slide

20. @holly_cummins #RedHat
    algorithms
    stack
    carbon
    awareness
    green software foundation: principles
    hardware
    efficiency
    electricity
    efficiency
    where
    when
    elasticity
    utilisation
    

    View Slide

21. @holly_cummins #RedHat
    algorithms
    
    
    stack
    carbon
    awareness
    hardware
    efficiency
    electricity
    efficiency
    where
    
    
    when
    elasticity
    
    
    utilisation
    green software foundation: principles
    

    View Slide

22. @holly_cummins
    #RedHat
    efficiency
    

    View Slide

23. @holly_cummins
    #RedHat
    what programming languages
    use the least energy?
    

    View Slide

24. @holly_cummins
    #RedHat
    what programming languages
    use the most energy?
    

    View Slide

25. View Slide

26. energy efficiency
    of programming
    languages
    

    View Slide

27. @holly_cummins
    #RedHat
    

    View Slide

28. @holly_cummins
    #RedHat
    

    View Slide

29. @holly_cummins
    #RedHat
    what’s the most carbon-efficient java?
    

    View Slide

30. @holly_cummins #RedHat
    

    View Slide

31. @holly_cummins #RedHat
    

    View Slide

32. @holly_cummins #RedHat
    does being small and fast reduce carbon footprint?
    

    View Slide

33. #Quarkus @holly_cummins
    measure, don’t guess.
    

    View Slide

34. digression:
    
    
    measuring carbon is hard
    

    View Slide

35. @holly_cummins #RedHat
    step 1: measure power usage
    

    View Slide

36. @holly_cummins #RedHat
    step 1: measure power usage
    wall power measurement
    

    View Slide

37. @holly_cummins #RedHat
    step 1: measure power usage
    wall power measurement
    more complete
    

    View Slide

38. @holly_cummins #RedHat
    step 1: measure power usage
    wall power measurement
    more complete
    needs access to the wall
    

    View Slide

39. @holly_cummins #RedHat
    step 1: measure power usage
    wall power measurement
    more complete
    needs access to the wall
    RAPL
    

    View Slide

40. @holly_cummins #RedHat
    step 1: measure power usage
    wall power measurement
    more complete
    needs access to the wall
    RAPL
    programmatically accessible
    

    View Slide

41. @holly_cummins #RedHat
    step 1: measure power usage
    wall power measurement
    more complete
    needs access to the wall
    RAPL
    programmatically accessible
    misses some components
    

    View Slide

42. @holly_cummins #RedHat
    step 1: measure power usage
    wall power measurement
    more complete
    needs access to the wall
    data costs carbon
    RAPL
    programmatically accessible
    misses some components
    

    View Slide

43. @holly_cummins #RedHat
    

    View Slide

44. @holly_cummins #RedHat
    

    View Slide

45. @holly_cummins #RedHat
    

    View Slide

46. @holly_cummins #RedHat
    load
    

    View Slide

47. @holly_cummins #RedHat
    Source: Teads EC2 instances carbon dataset
    

    View Slide

48. @holly_cummins #RedHat
    coal wind
    step 2: convert power usage to carbon
    solar
    

    View Slide

49. @holly_cummins #RedHat
    published energy mixes
    (these are made-up energy mixes)
    

    View Slide

50. @holly_cummins #RedHat
    published energy mixes … but methodologies are not open
    (these are made-up energy mixes)
    

    View Slide

51. @holly_cummins #RedHat
    published energy mixes … but methodologies are not open
    … or consistent
    (these are made-up energy mixes)
    

    View Slide

52. @holly_cummins #RedHat
    step 3: embedded carbon
    (manufacturing has costs)
    

    View Slide

53. @holly_cummins #RedHat
    simpler models
    

    View Slide

54. @holly_cummins #RedHat
    all models are wrong, some are useful
    

    View Slide

55. @holly_cummins #RedHat
    trick 1: the vrrrrrooooooooom model*
    * a made-up name
    

    View Slide

56. @holly_cummins
    #RedHat
    

    View Slide

57. @holly_cummins
    #RedHat
    these two columns
    are almost the same
    

    View Slide

58. @holly_cummins #RedHat
    energy consumption (sort of, mostly)
    is proportional to execution time
    

    View Slide

59. @holly_cummins #RedHat
    trick 2: the economic model*
    

    View Slide

60. @holly_cummins #RedHat
    trick 2: the economic model*
    * "economic input-output life cycle assessment"
    

    View Slide

61. @holly_cummins #RedHat
    reducing your cloud spend
    (probably) reducing your carbon footprint*
    hardware spend
    electricity bill
    * if you keep other factors the same
    

    View Slide

62. so … quarkus?
    

    View Slide

63. @holly_cummins #RedHat
    density
    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

64. @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
    

    View Slide

65. @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
    economic model in action:
    
    
    the cost and carbon metrics are
    (roughly) the same
    

    View Slide

66. @holly_cummins #RedHat
    Setup:
    
    
    • REST + CRUD
    
    
    • large heap
    
    
    • RAPL energy measurement
    
    
    • multiple instances to support high load
    

    Assumptions:
    
    
    • US energy mix
    Source: John O’Hara
    climate impact as a function of load
    

    View Slide

67. @holly_cummins #RedHat
    Setup:
    
    
    • REST + CRUD
    
    
    • large heap
    
    
    • RAPL energy measurement
    
    
    • multiple instances to support high load
    

    Assumptions:
    
    
    • US energy mix
    Source: John O’Hara
    climate impact as a function of load
    vrrrooom model in action:
    
    
    quarkus on JVM has the smallest footprint …
    because it has the highest throughput
    

    View Slide

68. @holly_cummins #RedHat
    trick 3: use quarkus
    

    View Slide

69. @holly_cummins #RedHat
    trick 3: use quarkus
    quarkus ‘automatically’ saves
    

    View Slide

70. @holly_cummins #RedHat
    trick 3: use quarkus
    quarkus ‘automatically’ saves
    • time
    

    View Slide

71. @holly_cummins #RedHat
    trick 3: use quarkus
    quarkus ‘automatically’ saves
    • time
    • money
    

    View Slide

72. @holly_cummins #RedHat
    trick 3: use quarkus
    quarkus ‘automatically’ saves
    • time
    • money
    • carbon (~2x)
    

    View Slide

73. @holly_cummins #RedHat
    trick 3: use quarkus
    quarkus ‘automatically’ saves
    • time
    • money
    • carbon (~2x)
    • … even when Spring compatibility libraries are used
    (almost no code changes except dependencies and tests)
    

    View Slide

74. @holly_cummins #RedHat
    trick 3: use quarkus
    quarkus ‘automatically’ saves
    • time
    • money
    • carbon (~2x)
    • … even when Spring compatibility libraries are used
    (almost no code changes except dependencies and tests)
    

    View Slide

75. @holly_cummins #RedHat
    carbon
    awareness
    green software foundation: principles
    hardware
    efficiency
    electricity
    efficiency
    

    View Slide

76. @holly_cummins #RedHat
    carbon
    awareness
    green software foundation: principles
    hardware
    efficiency
    electricity
    efficiency
    elasticity
    

    View Slide

77. @holly_cummins #RedHat
    carbon
    awareness
    green software foundation: principles
    hardware
    efficiency
    electricity
    efficiency
    elasticity
    utilisation
    

    View Slide

78. @holly_cummins #RedHat
    application
    utilisation
    

    View Slide

79. @holly_cummins #RedHat
    application
    utilisation
    high utilisation
    
    
    good case
    

    View Slide

80. @holly_cummins #RedHat
    application
    utilisation
    over-utilisation
    
    
    very bad case
    

    View Slide

81. @holly_cummins #RedHat
    application
    utilisation
    over-utilisation
    
    
    very bad case
    under-utilisation
    
    
    wasteful case
    

    View Slide

82. @holly_cummins #RedHat
    application
    elasticity
    high utilisation
    
    
    good case
    @holly_cummins
    

    View Slide

83. @holly_cummins #RedHat
    application
    elasticity
    scale-up
    
    
    good utilisation
    @holly_cummins
    

    View Slide

84. @holly_cummins #RedHat
    application
    elasticity
    scale-down
    
    
    good utilisation
    @holly_cummins
    

    View Slide

85. #RedHat
    @holly_cummins
    2017 server-survey
    
    
    25%
    
    
    doing no useful work
    
    
    (16,000 sampled)
    

    View Slide

86. #RedHat
    @holly_cummins
    2017 server-survey
    
    
    25%
    
    
    doing no useful work
    
    
    (16,000 sampled)
    “perhaps someone
    forgot to turn them off”
    

    View Slide

87. #RedHat
    @holly_cummins@hachyderm.io
    https://www.anthesisgroup.com/wp-content/uploads/2019/11/Comatose-Servers-Redux-2017.pdf
    2014 server-survey
    
    
    29%
    
    
    active less than 5% of the time
    
    
    (4,000 sampled)
    

    View Slide

88. @holly_cummins #RedHat
    cloud elasticity?
    https://www.business2community.com/cloud-computing/overprovisioning-always-on-resources-lead-to-26-6-billion-in-public-cloud-waste-expected-in-2021-02381033
    

    View Slide

89. @holly_cummins #RedHat
    cloud elasticity?
    2021:
    https://www.business2community.com/cloud-computing/overprovisioning-always-on-resources-lead-to-26-6-billion-in-public-cloud-waste-expected-in-2021-02381033
    

    View Slide

90. @holly_cummins #RedHat
    cloud elasticity?
    2021:
    $26.6 billion wasted
    https://www.business2community.com/cloud-computing/overprovisioning-always-on-resources-lead-to-26-6-billion-in-public-cloud-waste-expected-in-2021-02381033
    

    View Slide

91. @holly_cummins #RedHat
    cloud elasticity?
    2021:
    $26.6 billion wasted
    by always-on cloud instances
    https://www.business2community.com/cloud-computing/overprovisioning-always-on-resources-lead-to-26-6-billion-in-public-cloud-waste-expected-in-2021-02381033
    

    View Slide

92. @holly_cummins #RedHat
    it’s not just money
    

    View Slide

93. @holly_cummins #RedHat
    it’s not just electricity
    it’s not just money
    

    View Slide

94. @holly_cummins #RedHat
    it’s not just electricity
    it’s water
    it’s not just money
    

    View Slide

95. @holly_cummins #RedHat
    it’s not just electricity
    it’s water
    it’s e-waste
    it’s not just money
    

    View Slide

96. @holly_cummins
    #RedHat
    achieving elasticity
    we used to leave
    our applications
    running all the time
    when we
    scripted turning
    them off at night,
    we reduced our
    cloud bill by
    30%
    @darkandnerdy, Chicago DevOpsDays
    

    View Slide

97. @holly_cummins
    #RedHat
    yes, turning applications off is scary
    

    View Slide

98. @holly_cummins
    #RedHat
    what if … turning
    applications off was
    no more scary than
    turning the lights off?
    ultimate elasticity
    
    
    

    View Slide

99. @holly_cummins
    #RedHat
    

    View Slide

100. @holly_cummins
     #RedHat
     

     View Slide

101. @holly_cummins
     #RedHat
     turning it off and on again must
     

     View Slide

102. @holly_cummins
     #RedHat
     turning it off and on again must
     • be fast
     

     View Slide

103. @holly_cummins
     #RedHat
     turning it off and on again must
     • be fast
     • actually work
     

     View Slide

104. @holly_cummins
     #RedHat
     turning it off and on again must
     • be fast
     • actually work
     • idempotency
     

     View Slide

105. @holly_cummins
     #RedHat
     turning it off and on again must
     • be fast
     • actually work
     • idempotency
     • resiliency
     

     View Slide

106. @holly_cummins
     #RedHat
     architect things to be turned off and on often
     

     View Slide

107. @holly_cummins
     #RedHat
     trick 4:
     LightSwitchOps
     architect things to be turned off and on often
     

     View Slide

108. @holly_cummins
     #RedHat
     trick 4:
     LightSwitchOps
     architect things to be turned off and on often
     

     View Slide

109. #Quarkus @holly_cummins
     you can’t optimise what you can’t measure
     

     View Slide

110. @holly_cummins #RedHat
     FinOps
     figuring out who in your company forgot to turn off their cloud
     

     View Slide

111. @holly_cummins #RedHat
     

     View Slide

112. @holly_cummins #RedHat
     backstage.io
     

     View Slide

113. @holly_cummins #RedHat
     backstage.io
     •cost insights plugin
     

     View Slide

114. @holly_cummins #RedHat
     backstage.io
     •cost insights plugin
     •cloud carbon footprint plugin
     

     View Slide

115. @holly_cummins #RedHat
     algorithms
     
     
     stack
     carbon
     awareness
     hardware
     efficiency
     electricity
     efficiency
     where
     
     
     when
     elasticity
     
     
     utilisation
     green software foundation: principles
     

     View Slide

116. trick 5: electricity source
     

     View Slide

117. @holly_cummins #RedHat
     data center location matters
     

     View Slide

118. @holly_cummins #RedHat
     

     View Slide

119. @holly_cummins #RedHat
     

     View Slide

120. @holly_cummins #RedHat
     we need to talk
     about virginia
     

     View Slide

121. @holly_cummins #RedHat
     look at the sustainability information
     before choosing a hosting region
     

     View Slide

122. @holly_cummins #RedHat
     look at the sustainability information
     before choosing a hosting region
     choose a cloud provider who make this easy
     

     View Slide

123. @holly_cummins #RedHat
     time of day matters
     • (most) renewables are intermittent
     
     
     • if grid load is high, shortfalls are filled
     by fossil fuels
     

     View Slide

124. @holly_cummins #RedHat
     aaaaaaaargh?
     

     View Slide

125. @holly_cummins #RedHat
     “no-regrets” solutions
     

     View Slide

126. @holly_cummins
     #RedHat
     

     View Slide

127. @holly_cummins
     #RedHat
     co-benefits
     

     View Slide

128. @holly_cummins
     #RedHat
     co-benefits
     the double win
     

     View Slide

129. @holly_cummins
     #RedHat
     co-benefits
     the double win
     win-win
     

     View Slide

130. @holly_cummins
     #RedHat
     co-benefits
     the double win
     win-win
     twofer
     

     View Slide

131. @holly_cummins
     #RedHat
     co-benefits
     the double win
     win-win
     twofer
     überwinden
     

     View Slide

132. @holly_cummins
     #RedHat
     co-benefits
     the double win
     win-win
     twofer
     überwinden
     

     View Slide

133. @holly_cummins #RedHat
     climate solutions can
     make everything better
     

     View Slide

134. @holly_cummins #RedHat
     remember the zombie servers?
     https://www.business2community.com/cloud-computing/overprovisioning-always-on-resources-lead-to-26-6-billion-in-public-cloud-waste-expected-in-2021-02381033
     

     View Slide

135. @holly_cummins #RedHat
     remember the zombie servers?
     what else could we do
     with that $26.6 billion
     
     
     of wasted cloud spend?
     https://www.business2community.com/cloud-computing/overprovisioning-always-on-resources-lead-to-26-6-billion-in-public-cloud-waste-expected-in-2021-02381033
     

     View Slide

136. @holly_cummins #RedHat
     the double-win
     turning things off saves a lot of money
     

     View Slide

137. @holly_cummins #RedHat
     the double-win
     renewable electricity is 9x cheaper
     

     View Slide

138. @holly_cummins #RedHat
     the double-win
     renewable electricity is 9x cheaper
     hosting in Montreal:
     

     View Slide

139. @holly_cummins #RedHat
     the double-win
     renewable electricity is 9x cheaper
     hosting in Montreal:
     88% less carbon than the same workload in London
     

     View Slide

140. @holly_cummins #RedHat
     the double-win
     renewable electricity is 9x cheaper
     hosting in Montreal:
     88% less carbon than the same workload in London
     and it’s 15% cheaper
     

     View Slide

141. @holly_cummins #RedHat
     remember the vrrrrrooooooooom model?
     (probably not, it was a made-up name)
     

     View Slide

142. @holly_cummins #RedHat
     the double-win
     

     View Slide

143. @holly_cummins #RedHat
     the double-win
     

     View Slide

144. @holly_cummins #RedHat
     “this is not sacrifice. it’s advancement.”
     
     
     – Dr. Jonathan Foley
     

     View Slide

145. @holly_cummins #RedHat
     tl;dpa
     
     
     (too long; didn’t pay attention)
     trick 1: the vrrrooooom model says faster is greener
     
     
     trick 2: the economic model says cheaper is greener
     
     
     trick 3: choose a fast and energy-efficient framework
     (such as quarkus)
     
     
     trick 4: architect to be able to turn stuff off
     (LightSwitchOps)
     
     
     trick 5: choose your hosting wisely
     

     View Slide

146. @holly_cummins #RedHat
     we all make a difference
     

     View Slide

147. slides
     thank you
     
     
     and don’t forget to vote in the J-Spring
     app, which is running on Quarkus :)
     
     
     @holly_cummins@hachyderm.io
     

     View Slide