The move to converged platforms is becoming mainstream, as businesses seek to reduce complexity in their IT solutions, and the latest platform to see a move to convergence is big data and provider MapR who have developed two new solutions to allow businesses to ingest and analyse data in real-time.
The new MapR Streams solution is a global event streaming system that connects data producers and data consumers across shared topics of information and integrates directly into the new Converged Data Platform which consists of the new MapR Streams, and the current MapR-DB and MapR-FS products.
MapR claims the integration of MapR Streams into the new Converged Data Platform enables organisations to collect, analyse and act on real-time data from current applications and to prepare for the expected deluge of data from the Internet of Things (IoT) applications. From advertisers providing relevant real-time offers, to healthcare providers improving personalised treatment, to retailers optimising inventory, to telecom carriers dynamically adjusting mobile service areas.
Additionally MapR also says Streams is robust enough to scale to handle massive data flows and long-term persistence and will allow high availability (HA), disaster recovery (DR), security, and full data protection.
The MapR Converged Data Platform brings together three standard data systems
Apache Hadoop, Apache Spark and MapR’s NoSQL database in one package. Allowing users to work with a single supplier system across mutiple silos with data at rest and data in motion. One early adopter is comScore who is using the new system to analyse over 65 billion new events a day. As Michael Brown, CTO, comScore, explains he chose MapR Streams because it’s “built to ingest and process these events in real time, opening the doors to a new level of product offerings for our customers.”
Features of MapR Streams include:
- Ingesting data from thousands of locations with millions of topics and billions of messages
- Kafka API for real-time producers and consumers for easy application migration.
- Out-of-box integration with popular stream processing frameworks like Spark Streaming, Storm, Flink, and Apex.
- Includes auto-failover, order consistency, cross-site replication and unlimited persistence of all messages in a stream
MapR Streams will be generally available in early 2016.