10 things I learned making the fastest js server runtime in the world

This presentation is about server performance, which means that no time in the world would be enough to cover it all. Hopefully, I can share with you the top #10 things I've learned while putting JavaScript on the top of the server side benchmarks. You will learn about runtimes and engines, how some are more capable than others, and sometimes the obvious choice is not always the right one... This talk is about thinking outside of the box, being creative and don't take anything for granted. We will debunk myths about native code vs script or RAM usage, it's going to be fast! I promise!

Transcript

1. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
2. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
3. #1: #1: DEFINE FAST SERVER DEFINE FAST SERVER @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
4. server noun serv· er | \ ˈsər-vər\ a computer in a network that is used to provide services (such as access to files or shared peripherals or the routing of e-mail) to other computers in the network @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
5. Site Reliability Engineering is a discipline that incorporates aspects of software engineering and applies them to infrastructure and operations problems. The main goals are to create ultra-scalable and highly reliable software systems https://landing.google.com/sre/books/ @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
6. Site Reliability Engineering metrics Rate Errors  Latency  Saturation Utilization @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
7. Site Reliability Engineering metrics Rate —  Request rate, in requests/sec Errors  Latency  Saturation Utilization @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
8. Site Reliability Engineering metrics Rate —  Request rate, in requests/sec Errors —  Error rate, in errors/sec Latency  Saturation Utilization @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
9. Site Reliability Engineering metrics Rate —  Request rate, in requests/sec Errors —  Error rate, in errors/sec Latency —  Response time, including queue/wait time Saturation Utilization @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
10. #2: #2: UNDERSTAND THE APPLICATION LIFECYCLE UNDERSTAND THE APPLICATION LIFECYCLE @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
11. Server Application long running process deployed on cloud or bare metal attached to a fast network "enough" CPU/Memory @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
12. #3: #3: STATISTICS STATISTICS @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
13. mean - median @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
14. mean - median 99 percentile @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
15. #4: #4: JAVASCRIPT RUNTIMES JAVASCRIPT RUNTIMES @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
16. V8 JavaScript engine Speed up real-world performance for modern JavaScript, and enable developers to build a faster future web. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
17. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
18. #5: #5: BENCHMARK BENCHMARK @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
19. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
20. TECHEMPOWER TECHEMPOWER FRAMEWORK BENCHMARKS FRAMEWORK BENCHMARKS COMMITS: COMMITS: CONTRIBUTORS: CONTRIBUTORS: https://github.com/TechEmpower/FrameworkBenchmarks/ @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
21. TECHEMPOWER TECHEMPOWER FRAMEWORK BENCHMARKS FRAMEWORK BENCHMARKS COMMITS: COMMITS: 10715 10715 CONTRIBUTORS: CONTRIBUTORS: https://github.com/TechEmpower/FrameworkBenchmarks/ @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
22. TECHEMPOWER TECHEMPOWER FRAMEWORK BENCHMARKS FRAMEWORK BENCHMARKS COMMITS: COMMITS: 10715 10715 CONTRIBUTORS: CONTRIBUTORS: 504 504 https://github.com/TechEmpower/FrameworkBenchmarks/ @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
23. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
24. #6: #6: QUESTION THE STATUS QUO QUESTION THE STATUS QUO @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
25. Source (2014): http://blog.kazuhooku.com @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
26. Recap 1. we want fast as in fast server @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
27. Recap 1. we want fast as in fast server 2. server as in a long-running process @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
28. Recap 1. we want fast as in fast server 2. server as in a long-running process 3. monitor rates, latency (and errors) @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
29. Recap 1. we want fast as in fast server 2. server as in a long-running process 3. monitor rates, latency (and errors) 4. JS is more than V8/node @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
30. Recap 1. we want fast as in fast server 2. server as in a long-running process 3. monitor rates, latency (and errors) 4. JS is more than V8/node 5. results must be peer-reviewed and verifiable @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
31. #7: #7: LOOK UNDER THE HOOD LOOK UNDER THE HOOD @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
32. const http = require('http'); http .createServer((request, response) => { response.end("Hello World\n"); }) .listen(8000); https://nodejs.org/en/blog/uncategorized/profiling-node-js @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
33. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
34. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
35. #8: #8: TRY OTHER ENGINES TRY OTHER ENGINES @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
36. Graal https://graalvm.org @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
37. GRAAL GRAAL VS VS GRAAL.JS GRAAL.JS VS VS NODE.JS NODE.JS https://stefan-marr.de/papers/dls-marr-et-al-cross-language-compiler-benchmarking- are-we-fast-yet/ @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
38. Eclipse Vert.x Eclipse Vert.x is a polyglot reactive application framework that runs on the Java Virtual Machine https://vertx.io @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
39. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
40. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
41. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
42. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
43. #9: #9: UNIX PRINCIPLE UNIX PRINCIPLE @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
44. GRAAL.JS GRAAL.JS @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
45. GRAAL.JS GRAAL.JS + VERT.X + VERT.X @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
46. GRAAL.JS GRAAL.JS + VERT.X + VERT.X = ES4X = ES4X @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
47. GRAAL.JS GRAAL.JS + VERT.X + VERT.X = ES4X = ES4X https://github.com/reactiverse/es4x @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
48. ES4X Features commonjs loader npm compatibility (no native modules for now) .d.ts definitions for the java APIs debug/profile using chrome-devtools https://reactiverse.io/es4x @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
49. CLOC stats Language files blank comment code Java 21 318 584 1393 JavaScript 10 148 253 887 JSON 1 1 0 51 SUM: 32 467 837 2331 http://github.io/reactiverse/es4x @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
50. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
51. #10: #10: VERIFY THE RESULTS VERIFY THE RESULTS @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
52. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
53. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
54. HOW DOES IT COMPARE? HOW DOES IT COMPARE? @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
55. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
56. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
57. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
58. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
59. CONCLUSION CONCLUSION There's nothing wrong with JavaScript. @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
60. CONCLUSION CONCLUSION There's nothing wrong with JavaScript. Yes JavaScript is fast @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
61. CONCLUSION CONCLUSION There's nothing wrong with JavaScript. Yes JavaScript is fast You don't need to switch to Go/Rust/etc... @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
62. CONCLUSION CONCLUSION There's nothing wrong with JavaScript. Yes JavaScript is fast You don't need to switch to Go/Rust/etc... Dare to Experiment @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes
63. THANK YOU! THANK YOU! @pml0pes pmlopes https://reactiverse.io/es4x https://vertx.io https://graalvm.org @pml0pes @pml0pes @pml0pes @pml0pes @pml0pes