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.

Atlas Ad-Serving Server - Advertising at scale at Facebook

Facebook's Atlas is an ad server that also allows ad buyers to measure, target, and optimize digital and mobile ads across digital (i.e., not just on Facebook). Atlas operates separately from Facebook, does not access personal information from the social network or share marketing data with Facebook.

Atlas Ad-Serving Server - Advertising at scale at Facebook

  1. 1. Rebuilding Atlas – Advertising at Scale at Facebook Jason McHugh, Software Engineer | March, 2015
  2. 2. Outline • Ad serving technology • History of Atlas • The acquisition • Challenges • Architecture • Lessons learned
  3. 3. Advertising
  4. 4. Advertising • Advertising a huge industry with massive budgets • Digital advertising is the fastest growing advertising medium • More time spent with digital media than TV • Time spent on mobile exceeded TV Source: eMarketer, 2014; Flurry, 2014
  5. 5. Third Party Ad Serving
  6. 6. How Third Party Ad Serving Works Advertiser
  7. 7. How Third Party Ad Serving Works Advertiser Campaigns Click-Through URLs Atlas Creative Concepts
  8. 8. Serving An Ad Atlas You, me, or other random human
  9. 9. Retargeting Atlas
  10. 10. Retargeted Ad Atlas
  11. 11. History of Atlas • 1997 – Atlas started life as Avenue A • 2003 – Avenue A becomes aQuantive • 2007 – aQuantive was purchased by Microsoft for $6 billion dollars • 2012 – Microsoft took a $6.2 billion dollar writedown • 2013 – Facebook acquired Atlas in April • 2014 – Atlas New publically launched at Ad Week in September
  12. 12. Challenges
  13. 13. • Ad Tech Stack • Architecture • Data Model and databases • Data flows • Deployment Challenge: Understand the System One Single DB Instance
  14. 14. Challenge: Huge Product • Third party ad server • Advertiser and publisher negotiations – RFP • Search management • Search optimization • Email tracking • Custom analysis and reporting • Rich media including video • Franchise Management
  15. 15. Challenge: No Lift and Shift • Lift and Shift • Common approach after acquisition • Take as much as exists at the time of the purchase • Move to your data centers and then evolve • Not possible here • Non-open compute hardware • Usage of Microsoft close-source technologies
  16. 16. Architecture • Holistic view of the logical architecture • Detail a piece of the physical architecture
  17. 17. Logical Architecture
  18. 18. Logical Architecture
  19. 19. Logical Architecture
  20. 20. Logical Architecture - Focused
  21. 21. Physical Architecture – Ad Delivery
  22. 22. Physical Architecture – Data Processing Pipeline
  23. 23. Scribe • Large-scale, high throughput message queue • Not lossless but guarantees are excellent and perfect for us • Decouples producers from consumers • Persistent for n days • Sharded consumption • Checkpoint streams
  24. 24. Physical Architecture – Data Processing Pipeline
  25. 25. Detailed Look – Stream Processing • Mini workflows • Scalability is hard • Message queues can be costly • Repeatable re-execution
  26. 26. Physical Architecture
  27. 27. Physical Architecture
  28. 28. Physical Architecture
  29. 29. Physical Architecture
  30. 30. Physical Architecture
  31. 31. Physical Architecture
  32. 32. Lessons Learned
  33. 33. Mistake – Minimize the www/hack tier
  34. 34. Mistake – Minimize the www/hack tier • Minimize the code in www tier to ensure a higher level of availability • Hack and the www tier have come a long way in 2 years • Huge improvements in availability • API changes hit two separate systems • Couldn’t leverage all the improvements and investments in hack
  35. 35. Mistake – Looking ahead • Lesson • Look ahead to where the industry is going • Or look to where an organization (or group) will invest efforts • Plan to meet them there
  36. 36. Questions?

×