Re: Correction: VoltDB says it can replace Spark Streaming, not the entire Spark framework
Thank you Doug.
VoltDB competes with Spark streaming - especially in cases where spark streaming is used to sessionize, transform incoming streams or implement real time analytics (in the style of a Lambda speed layer).
However, VoltDB integrates with a breadth of OLAP and Hadoop tooling - including Spark. In these cases, developers use Spark to calculate OLAP style analytics against large data sets and use those outputs in VoltDB to build real time applications for personalization, fraud alerting, and streaming analytics.
For example, one VoltDB user deploys VoltDB in combination with Spark to detect online advertising bot fraud. VoltDB is performing a sessionization function - processing 100,000's of events per second into sessions - while Spark is used to run machine learning algorithms that detect fraud.