David Richards & Jagane Sandar - Hadoop Summit 2013 - theCUBE - #HadoopSummit
summary
Claims production-grade active-active replication of Hadoop 2.0 across data centers thousands of miles apart, demonstrating teragen in US East and terasort in Oregon with failover and zero data loss. Implements a time-independent Paxos for distributed coordination rather than relaxed ZooKeeper-style compromises. Announces commercial support for Spark and Shark with UC Berkeley AmpLab and releases S3 HDFS open sourced for Amazon integration. Talks OEM deals and a sales pipeline aiming to move trials into mission‑critical production.
AI-written summary · about
chapters (10)
- 0:00Introduction to WANdisco
- 1:04Four major announcements
- 3:11Active active replication explained
- 4:36Solving the zero data loss problem
- 7:45Overcoming enterprise skepticism
- 9:28Developer collaboration origins
- 11:14Why others failed at active active
- 12:41Customer applications and OEMs
- 13:45Partnerships and ecosystem strategy
- 15:08Future milestones and execution
