Growing Facebook on Mobile, a Realtime Analytics Story
summary
Covers Facebook’s realtime mobile analytics pipeline: how mobile logging, realtime processing, and analysis systems were built to provide faster feedback for product iterations. Describes challenges in mobile data collection and processing and gives concrete examples of using those systems to improve app performance, stability, and engagement.
AI-written summary · about
chapters (23)
- 0:00Start
- 4:57Agenda
- 5:35Story load TTI
- 6:00Data Driven Move fast and build things
- 6:48Traffic is shifting ...
- 7:13Growing Facebook on mobile Move fast and build things
- 7:54Mobile Challenges Data Collection
- 9:24Mobile Challenges Upgrade cycle
- 9:39Example: Story load performance
- 13:43Flexible Sampling Configurable from server side
- 13:54Pipeline Flexible Sampling
- 15:21Event Quality Issues
- 16:27Processing Challenges
- 17:22Example Processing
- 21:47Processing Fault-tolerance
- 23:22Analysis
- 25:27Processing Scale: Mobile Devs
- 25:47Example Recap
- 28:46Life Cycle of a Bug Fix Before
- 30:39Life Cycle of a Bug Fix With realtime data
- 31:10Stability & Performance Real time data changes the development process
- 33:06Faster iteration cycles 2 week release cycle on Android
- 34:44Feature Uses Unlocked Framework flexible enough for production feature use
related talks
10mInterview
27mBig Data in Mobile Games
24mWrangling the CPU iOS Performance in the Facebook App
25mThe importance of data for successful mobile marketing and strategy
53mResearch in mobile games: finding insights from idea to release
45mData as culture as software
55mMobile Application Performance Analysis
28mRealtime Content Delivery: Powering dynamic instant experiences on your mobile apps