The Node.js Scalability Myth

Transcript

1. 1.

   The Node.js Scalability Myth Felix Geisendörfer 15.03.2012 at BerlinJS user

   group
2. 2.

   (@)felixge(.de)

3. 3.

   transloadit.com (using node since v0.0.6)

4. 4.

   core contributor

5. 5.

   I need to build an über-scalable web service. It will

   be the next big thing!
6. 6.

   Easy! I will use node.js and MongoDB!

7. 7.

   Easy! I will use node.js and MongoDB!

8. 8.

   Because node.js is “scalable”, right?

9. 9.

   None
10. 10.

    Audience?

11. 11.

    Backend Developers?

12. 12.

    Are you measuring app performance in production?

13. 13.

    1000 req / s

14. 14.

    100 req / s

15. 15.

    10 req / s

16. 16.

    10 requests / per second = 864000 requests / day

    10 req / s
17. 17.

    Scalability

18. 18.

    None
19. 19.

    The Node.js Scalability Myth • Threads don’t scale • Event

    loops do
20. 20.

    The Node.js Scalability Myth • Threads don’t scale • Event

    loops do (Probably) not true in 2012 (Very likely) not relevant
21. 21.

    More relevant • Understanding scalability • Using the right tool

    for the job
22. 22.

    Vertical Scalability

23. 23.

    Adding more resources to a single node

24. 24.

    None
25. 25.

    Vertical Scalability • CPU • GPUs • Memory • Disk

    • Network
26. 26.

    CPU

27. 27.

    Computers of the future will have hundreds of cores!

28. 28.

    CPU • Node is single threaded (runs on a single

    CPU) • Using multiple CPUs requires multiple node processes (usually via prefork) • No shared memory
29. 29.

    CPU (v8) • V8 compiles JS to Assembly • Just-in-time

    compilation (JIT) • Does reasonably well in those language “benchmarks”
30. 30.

    GPUs

31. 31.

    GPUs • No support in node itself • node-cuda addon

    by Kashif Rasul
32. 32.

    Memory

33. 33.

    Memory • No hard memory limit on 64 bit (since

    node-0.6 / v8-3.6.5) • JS is a garbage collected language (avoid huge heaps) • Buffers do not count towards heap
34. 34.

    Network

35. 35.

    Network • Node’s network stack is a thin layer on

    top of underlaying system calls • Good at fully saturating network interfaces
36. 36.

    Disk

37. 37.

    Disk • Done in thread pool • Unfortunately along with

    DNS at this point • Throughput ok, but not ideal yet
38. 38.

    Horizontal Scalability

39. 39.

    Adding more nodes to a system

40. 40.

    Horizontal Scalability • Node has no horizontal scaling specific features,

    but ... • Fast networking & JSON / Buffers -> DIY
41. 41.

    Horizontal Scalability • Simple stack (requires only build tools, python)

    • Deployable without container server • Good package management via npm
42. 42.

    Horizontal Solution “I have the full solution ...”

43. 43.

    Horizontal Solution “... but unfortunately it did not fit on

    this slide”
44. 44.

    tl;dr • Know your requirements • There are no silver

    bullets
45. 45.

    Questions? Slides will be available at felixge.de

46. 46.

    Thanks