Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

of

Scaling Uber Slide 1 Scaling Uber Slide 2 Scaling Uber Slide 3 Scaling Uber Slide 4 Scaling Uber Slide 5 Scaling Uber Slide 6 Scaling Uber Slide 7 Scaling Uber Slide 8 Scaling Uber Slide 9 Scaling Uber Slide 10 Scaling Uber Slide 11 Scaling Uber Slide 12 Scaling Uber Slide 13 Scaling Uber Slide 14 Scaling Uber Slide 15 Scaling Uber Slide 16 Scaling Uber Slide 17 Scaling Uber Slide 18 Scaling Uber Slide 19 Scaling Uber Slide 20 Scaling Uber Slide 21 Scaling Uber Slide 22 Scaling Uber Slide 23 Scaling Uber Slide 24 Scaling Uber Slide 25 Scaling Uber Slide 26 Scaling Uber Slide 27 Scaling Uber Slide 28 Scaling Uber Slide 29 Scaling Uber Slide 30 Scaling Uber Slide 31 Scaling Uber Slide 32 Scaling Uber Slide 33 Scaling Uber Slide 34 Scaling Uber Slide 35 Scaling Uber Slide 36 Scaling Uber Slide 37 Scaling Uber Slide 38 Scaling Uber Slide 39 Scaling Uber Slide 40 Scaling Uber Slide 41 Scaling Uber Slide 42 Scaling Uber Slide 43 Scaling Uber Slide 44 Scaling Uber Slide 45 Scaling Uber Slide 46 Scaling Uber Slide 47 Scaling Uber Slide 48 Scaling Uber Slide 49 Scaling Uber Slide 50 Scaling Uber Slide 51 Scaling Uber Slide 52 Scaling Uber Slide 53 Scaling Uber Slide 54 Scaling Uber Slide 55 Scaling Uber Slide 56 Scaling Uber Slide 57 Scaling Uber Slide 58 Scaling Uber Slide 59 Scaling Uber Slide 60 Scaling Uber Slide 61 Scaling Uber Slide 62 Scaling Uber Slide 63 Scaling Uber Slide 64 Scaling Uber Slide 65 Scaling Uber Slide 66
Upcoming SlideShare
Visualising Data with Code
Next

166 Likes

Share

Scaling Uber

Video and slides synchronized, mp3 and slide download available at URL http://bit.ly/1ncT8iO.

From its simple roots as a PHP program, Uber has grown into a complex distributed system deployed across multiple datacenters using multiple databases and programming languages. Matt Ranney covers the evolution of Uber's architecture and some of the systems they built to handle the current scaling challenges. Filmed at qconsf.com.

Matt Ranney is the Chief Systems Architect at Uber. He has a computer science degree which has come in handy over a career of mostly network engineering, operations, and analytics.

Related Books

Free with a 30 day trial from Scribd

See all

Scaling Uber

  1. SCALINGUBERMATT RANNEY
  2. InfoQ.com: News & Community Site • 750,000 unique visitors/month • Published in 4 languages (English, Chinese, Japanese and Brazilian Portuguese) • Post content from our QCon conferences • News 15-20 / week • Articles 3-4 / week • Presentations (videos) 12-15 / week • Interviews 2-3 / week • Books 1 / month Watch the video with slide synchronization on InfoQ.com! http://www.infoq.com/presentations /uber-scalability-arch
  3. Purpose of QCon - to empower software development by facilitating the spread of knowledge and innovation Strategy - practitioner-driven conference designed for YOU: influencers of change and innovation in your teams - speakers and topics driving the evolution and innovation - connecting and catalyzing the influencers and innovators Highlights - attended by more than 12,000 delegates since 2007 - held in 9 cities worldwide Presented at QCon San Francisco www.qconsf.com
  4. As of November 2015: Uber Cities Worldwide: 356 Countries: 65 Employees: 4,800 Engineers: 1,500 US Driver Payments in 2015: $3.5B
  5. UBER ENGINEERING HISTORY
  6. 2009-2010 Outsourced PHP + MySQL Jan 2011 "dispatch" - Node.JS/MongoDB Jan 2011 “API” - Python/SQLAlchemy/MySQL Feb 2012 Dispatch swaps MongoDB for Redis May 2012 Dispatch adds ON fallback Jan 2013 First non-API Python services Feb 2013 API switched to Postgres Mar 2014 New Python services use MySQL Mar 2014 Schemaless begins, must finish before pg collapse Sep 2014 First Schemaless - trips out of Postgres Aug 2015 Dispatch X.0 / Ringpop / Riak
  7. TECHNICAL DEBT
  8. MICROSERVICES
  9. MICROSERVICES Immutable? Append Only?
  10. Node.JS Python Go Java
  11. SCALING NODE
  12. Getting out of the HTTP+JSON business HTTP is slow, complex, and inconsistent JSON is hard to validate, awkward in non-node Thrift is OK, but generated code is bad
  13. SERVICE DISCOVERY
  14. Lots of services, lots of instances Mostly Node.JS and Python Call graph unknowable Self-inflicted DoS Cascading failures
  15. load balancerservice A service B service B
  16. load balancer service A service B service B
  17. horizontal scalability zipkin tracing circuit breaking rate limiting failure testable almost no configuration as available as possible
  18. overall latency ≥ latency of slowest component 1ms avg, 1000ms p99 use 1: 1% at least 1000ms use 100: 63% at least 1000ms 1.0 - 0.99^100 = 0.634 = 63.4% LATENCY
  19. requeststhatareslow 0% 25% 50% 75% 100% Processes Used 1 2 4 8 16 32 64 128 256 512 1024 p95 p99 p99.9
  20. CULTURAL CHANGES
  21. FAILURE TESTING
  22. RETRIES
  23. partner app dispatch DC1 Location Updates State Digest dispatch DC2 Location Updates State Request
  24. http://www.principlesofchaos.org/
  25. EMBRACE THE CHAOS
  26. THANKS
  27. Watch the video with slide synchronization on InfoQ.com! http://www.infoq.com/presentations/ uber-scalability-arch
  • MeganWard46

    Dec. 4, 2021
  • MrsBFarley

    Dec. 3, 2019
  • NbaysItsolusenzMadur

    Nov. 2, 2018
  • Vhjkjh

    Jul. 26, 2018
  • zeeshanamin71216

    Jan. 22, 2018
  • ssuser8e8e40

    Jan. 11, 2018
  • JayavardhanAluri

    Dec. 29, 2017
  • nnttung

    Sep. 13, 2017
  • FranciaFranco3

    Sep. 12, 2017
  • GopalaKR

    Jul. 23, 2017
  • DANDUMEAnasMMande

    Jul. 8, 2017
  • TheRealAkansha

    May. 24, 2017
  • KaTem10

    May. 16, 2017
  • besa533

    May. 9, 2017
  • Jerrybecnel

    Apr. 27, 2017
  • CryptoConsumer

    Apr. 26, 2017
  • omaruriel

    Apr. 2, 2017
  • MwebazeMark

    Mar. 27, 2017
  • AmarnathMahadevSalga

    Mar. 27, 2017
  • 786201616

    Mar. 21, 2017

Video and slides synchronized, mp3 and slide download available at URL http://bit.ly/1ncT8iO. From its simple roots as a PHP program, Uber has grown into a complex distributed system deployed across multiple datacenters using multiple databases and programming languages. Matt Ranney covers the evolution of Uber's architecture and some of the systems they built to handle the current scaling challenges. Filmed at qconsf.com. Matt Ranney is the Chief Systems Architect at Uber. He has a computer science degree which has come in handy over a career of mostly network engineering, operations, and analytics.

Views

Total views

32,586

On Slideshare

0

From embeds

0

Number of embeds

67

Actions

Downloads

0

Shares

0

Comments

0

Likes

166

×