Efficient Software: A Developer’s Manual for Saving The World

None of us actually like waste, but many of us tolerate it. This is a shame, because waste is really really bad.
It makes our software more expensive to develop, and more expensive to run. It contributes to climate change. It means sometimes, people who’d like to use our software, can’t. It slows us down.

In this talk, Holly will present a range of practical waste-reduction techniques, including:

- LightSwitchOps
- Moving computational work to where it hurts least
- Measuring the right thing, instead of measuring the wrong thing (harder than it seems!)
- Performance profiling basics
- Doing less

Transcript

1. Holly Cummins Senior Principal Software Engineer, Quarkus Devoxx UK May 9, 2025 Efficient Software: A Developer’s Manual for Saving The World
2. @holly_cummins #RedHat software that does less efficient software?
3. @holly_cummins #RedHat
4. @holly_cummins #RedHat software that does what you need needed with fewer resources efficient software:
5. @holly_cummins #RedHat software that does less of what you don’t need
6. @holly_cummins #RedHat “Half the money I spend on advertising is wasted; the trouble is I don't know which half.” – John Wanamaker (1838-1922)
7. @holly_cummins #RedHat half the CPU cycles of our apps are wasted; the trouble is we don’t know which half.
8. @holly_cummins #RedHat half the footprint of our apps is wasted; the trouble is we don’t know which half.
9. @holly_cummins #RedHat no compromise improved efficiency
10. @holly_cummins #RedHat no regrets improved efficiency
11. @holly_cummins #RedHat intel mac vs arm mac example
12. @holly_cummins #RedHat memory cpu network time hardware mone carbon
13. #RedHat @hollycummins.com cost benefits of efficiency user experience saving the world
14. #RedHat @hollycummins.com 100 ms latency on page load 7% lower conversion rate
15. @holly_cummins #RedHat https://blog.linkedin.com/2017/august/3/making-linkedin-more-accessible-via-linkedin-lite modern web is so inefficient it is useless for part of its audience “my heart sank … our new feature failed to load because of poor internet connectivity” Nashik
16. #RedHat @hollycummins.com €12,000 saved by one script for shutting down a school’s computers overnight
17. #RedHat @hollycummins.com ⅔ fewer cloud instances by switching to quarkus
18. #RedHat @hollycummins.com
19. @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
20. @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
21. #RedHat @hollycummins.com
22. #RedHat @hollycummins.com it’s not just electricity it’s water it’s e-waste it’s embodied carbon
23. #RedHat @hollycummins.com why not fix all this?
24. #RedHat @hollycummins.com yes plz, how?
25. @holly_cummins #RedHat performance engineering finops code ops how it’s used
26. @holly_cummins #RedHat stack code zombies ops algorithms elasticity + provisioning
27. @holly_cummins #RedHat stack code zombies ops algorithms elasticity + provisioning
29. @holly_cummins #RedHat
30. energy efficiency of programming languages
31. @holly_cummins rust is hard no compromise
32. quarkus no compromise
33. @holly_cummins #RedHat java application frameworks needed dynamic dependencies the distant past
34. @holly_cummins #RedHat cloud apps are immutable now
35. @holly_cummins #RedHat a highly dynamic runtime in a container is pointless
36. @holly_cummins Java dynamism @ @ </> packaging (maven, gradle…) build time runtime ready to do work! load and parse • config files • properties • yaml • xml • etc. • classpath scanning and annotation discovery • attempt to load class to enable/disable features build a metamodel of the world start • thread pools • I/O • etc.
37. @holly_cummins what if we start the application more than once? @ @ </> @ @ </> @ @ </> @ @ </> so much work gets redone every time
38. @holly_cummins Hibernate speed example: JTA auto-wiring Class.forName(“LikelyJTAImplementation”); Class.forName(“APossibleJTAImplementation”); Class.forName(“AnotherJTAImplementation”); Class.forName(“NicheJTAImplementation”); Class.forName(“VeryNicheJTAImplementation”); … ~129 auto-wiring attempts every single start. ClassNotFoundException
39. @holly_cummins unused implementation the one we want interface megamorphic call slow dispatching unused implementation unused implementation the true cost of loaded classes isn’t just memory + start time method dispatching:
40. @holly_cummins the true cost of loaded classes isn’t just memory + start time the one we want monomorphic call fast dispatching interface
41. @holly_cummins how do we fix all this?
42. @holly_cummins @ @ </> build time runtime ready to do work! start • thread pools • I/O • etc. what if we initialize at build time?
43. @holly_cummins @ @ </> repeated starts are now efficient less wasted work
44. @holly_cummins #RedHat https://medium.com/arconsis/spring-boot-vs-quarkus-part-2-jvm-runtime-performance-af45d0db116e spring boot quarkus response time 1901 ms 294 ms throughput 523 req/s 3374 req/s
45. @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
46. @holly_cummins #RedHat stack code zombies ops algorithms elasticity + provisioning
47. @holly_cummins #RedHat measure, don’t guess. first step of performance optimisation
48. @holly_cummins #RedHat but measuring can be expensive - measure more often - measure proxies (with caution)
49. @holly_cummins #RedHat but measuring can be scary - make sure you’re measuring the right thing - be cautious with micro-benchmarks
50. @holly-cummins.bsky.social make sure to measure the big picture before micro-optimizing the small picture
51. @holly_cummins #RedHat wrong measurement de-optimising doing nothing
52. @holly_cummins #RedHat monitor profile analyze tune + apply source: Monica Beckwith
53. @holly-cummins.bsky.social profile shave shave profile shave profile shave profile shave profile shave profile
54. #Quarkus @holly_cummins method profiler GC analysis heap analysis APM distributed tracing Mission Control flame graphs GCMV New Relic* AppDynamics* Micrometer OpenTelemetry 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)
55. #Quarkus @holly_cummins is there a shortcut?
56. @holly_cummins #RedHat
57. @holly_cummins #RedHat
58. @holly_cummins the vrroooom model but what about saving the world?
59. @holly_cummins naming is the hardest problem in computer science invented by Dr. Vroom (really!)
60. @holly_cummins naming is the hardest problem in computer science
61. @holly_cummins my vrroooom model
63. @holly_cummins #RedHat these two columns are almost the same
64. @holly_cummins #RedHat Energy 1 10 100 Time 1 10 100 the trend line is more or less straight energy efficiency across programming languages Python Rust Java Go
65. case study quarkus
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 wait, what?
67. @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
68. @holly_cummins #RedHat stack code zombies ops algorithms elasticity + provisioning
69. #RedHat @hollycummins.com all the —opses
70. #RedHat @hollycummins.com GreenOps greenops is a mid-sized trilobite (really)
71. #RedHat @hollycummins.com FinOps figuring out who in your company forgot to turn off their cloud
72. @holly_cummins #RedHat application utilisation high utilisation good case over-utilisation very bad case under-utilisation wasteful case
73. @holly_cummins #RedHat application elasticity high utilisation good case scale-up good utilisation scale-down good utilisation @holly_cummins
74. #RedHat @hollycummins.com 29% of servers active <5% of the time https://www.anthesisgroup.com/wp-content/uploads/2019/11/Comatose-Servers-Redux-2017.pdf
75. #RedHat @hollycummins.com the average server: 12 - 18% of capacity 30 - 60 % of maximum power https://www.nrdc.org/sites/default/files/data-center-efficiency-assessment-IB.pdf
76. #RedHat @hollycummins.com backstage.io •cost insights plugin •cloud carbon footprint plugin image credit: RedMonk
77. @holly_cummins #RedHat stack code zombies ops algorithms elasticity + provisioning
78. #RedHat @hollycummins.com Hey boss, I created a Kubernetes cluster. I forgot it for 2 months. … and it’s €1000 a month. 2018
79. #RedHat @hollycummins.com
80. #RedHat @hollycummins.com 30% of servers doing no useful work
81. #RedHat @hollycummins.com zombie “they haven't delivered any information or computing services for six months or more”
82. #RedHat @hollycummins.com “we run this as a batch job on weekends, but the servers stay up all week” “
83. #RedHat @hollycummins.com “we only use this system in UK working hours, but we leave it running 24/7 ” “
84. #RedHat @hollycummins.com $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 2021 study
85. @hollycummins.com #RedHat ultimate elasticity
86. #RedHat @hollycummins.com we don’t switch the light off because we’re not sure if it will come back on never happens
87. #RedHat @hollycummins.com we don’t switch the server off because we’re not sure if it will come back on happens all the time
88. #RedHat @hollycummins.com we don’t switch the server off because it would be too much work to recreate it happens all the time
89. @hollycummins.com #RedHat turning it off and on again must • be fast • actually work • idempotency • resiliency
90. @hollycummins.com #RedHat LightSwitchOps making turning servers off as safe and easy as turning lights off
91. @hollycummins.com #RedHat step 1: no more scary state
92. #RedHat @hollycummins.com GitOps (infrastructure as code)
93. #RedHat @hollycummins.com kubectl apply -f all-my-cluster/ ansible-playbook stuff.yml spin it down spin it up
94. @hollycummins.com #RedHat step 1: no more scary state step 2: automate, automate
95. zombie reduction does not need to be fancy
96. #RedHat @hollycummins.com scream test
97. #RedHat @hollycummins.com “eco-monkey”
98. @hollycummins.com #RedHat the scream is real this internal server doesn’t seem to have a purpose uh … why did the backbone of a client’s network just vanish? let’s turn it off! oops.
99. #RedHat @hollycummins.com large UK bank, 2013 50% reduction in CPUs with a lease system self-destructing instances
100. @hollycummins.com #RedHat timed shutoff 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
101. @hollycummins.com #RedHat fancier scripts … with a frontend and a backend
102. @hollycummins.com #RedHat quarkus
103. @holly_cummins #RedHat what problem are we trying to solve?
104. @holly_cummins #RedHat define SLA performance engineering basics define success
105. @holly_cummins #RedHat resources are always limited we all have too much to do
106. @holly_cummins #RedHat - 99% of requests under 100ms - can run on t2.micro instance - cost per transaction < £0.02 - save the world example SLAs ?
107. @holly_cummins #RedHat diminishing returns effort amount of world-saving low-hanging fruit wringing performance out of a stone hard work stop here?
108. @holly_cummins #RedHat return on investment time money saved by the fix less money spent on the fix winning financial ruin stop here?
109. @holly_cummins #RedHat tl;dpa (too long; didn’t pay attention) efficient code is good for users, good for business, and good for the world choose an efficient stack (such as quarkus!) measure, profile, and optimise your code to get rid of waste think about how you provision: is it elastic? have you got zombies? flamethrower them!
110. slides @hollycummins.com