Jagane Sundar, WANdisco | BigDataNYC 2016
summary
Claims a Paxos-based algorithm delivers strongly consistent replication across WANs (HDFS to S3/Azure/Wasabi and NetApp filers) so replicas never return different answers. Demonstrates live use cases: disaster recovery, active/active, Cloud bursting, Spark Streaming and IoT ingestion. Shows migrations that run without taking apps offline and without pounding source appliances, moving massive datasets in days rather than weeks. Announces an SDK for custom plugins. Reports the system resists wide-area failures and succeeds in customer migrations with low impact and seamless cutover.
AI-written summary · about
chapters (10)
- 0:00Introduction and welcome
- 1:11Expanding beyond Hadoop
- 2:39Strong vs eventual consistency
- 4:24Google Spanner-like capabilities
- 5:30Data as a strategic asset
- 7:17Hybrid Cloud redesign
- 9:04Migration beyond big data
- 10:52Low-downtime migration tech
- 12:40IOT data consistency needs
- 14:29Integration challenges and roadmap
