Jagane Sundar | Hadoop Summit 2014
summary
WANdisco claims a working system that makes HBase tolerate region-server failures by storing each region on three servers with a distributed coordination engine so a single region-server outage is a non-event, enabling active-active region servers. It runs a single HDFS across data centers with a single authentication point, supports Tier‑1 financial and healthcare apps with virtually no failover, runs on Cloudera and Hortonworks, and cites a Wikibon figure that ~70% of deployments span multiple data centers.
AI-written summary · about
chapters (11)
- 0:00Introduction at Hadoop Summit
- 1:11Non-stop Hadoop across data centers
- 2:35Enterprise security in Hadoop
- 5:05Multi-data center deployments grow
- 7:11Financial sector's tier one applications
- 8:40Working with Hadoop distributions
- 10:22Future of Apache Hadoop ecosystem
- 12:50Enterprise Hadoop table stakes
- 14:54Spark and YARN integration
- 15:35CloudEra vs Hortonworks differences
- 17:00Market adoption and predictions
