In-Hadoop Analytics: Bringing analytics to big data
summary
Claims that analytics run where Hadoop stores it, turning clusters of tens to thousands of nodes holding terabytes to petabytes into live analytics platforms. Presents a reference architecture with a data lake/landing zone and co-located engines — Hadoop, Spark for interactive workloads, SQL engines, text analytics and machine learning — and a concrete flow: information extraction → SQL transformations → ML training → scoring back into SQL. Architecture is under development and aims to enable iterative discovery and mixed workloads.
AI-written summary · about
related talks
10mData Rivers — The New Analytics Architecture
58mUnderstanding Big Data for Software Engineers
07mComplementary Solutions with Hadoop
11mBreaking Data Silos: Enterprise Data Management
47mWhat's Ahead for Big Data?
49mBuilding a 100‑Petabyte Analytics Warehouse
31mData Warehouses Meet Data Lakes
17mDemocratising data analytics