Near Real-Time Joins at Scale10/05/2017 Objective: Guidance Three types of distributed joins at near real-time scale are explored. 1. Real time stream (such as Kafka) ⋈; static table 2. Real time stream ⋈; real time stream 3. Real time stream ⋈; mutable table: This is the hardest scenario; a novel approach is featured. Each of these joins have a different solution explored in the presentation, and is supported by case studies from eBay. Speaker(s)
|