Simple, unified programming

for the big data audience

Simple, unified programming for the big data audience

SnappyData’s core programming abstraction is the same as Spark. It exposes all its data as a Spark DataSet or DataFrame (Spark’s Resilient Distributed DataSet).Using this simple extension, Spark can capture a wide range of processing workloads that previously needed separate engines, including SQL, streaming, machine learning, and graph processing.

So, for instance, you can write a stream processing pipeline that consumes a stream, transactionally ingests it into a Snappydata table, draws insight by running a query combining a stream with historical data and even updates a Machine learning model all using a single programming paradigm.

Spark's generality has several important benefits. First, applications are easier to develop because they use a unified API. Second, it is more efficient to combine processing tasks; whereas prior systems required writing the data to storage to pass it to another engine, Spark can run diverse functions over the same data, often in memory.

SnappyData Programming Features

  • 100% compatible with Spark APIs

  • Support high concurrency across users and applications

  • Transactions and mutability through SQL and Spark API extensions

  • High availability (not just Fault tolerance)

  • Enterprise grade storage and security

  • SQL extensions: Insert, delete, update, constraints, indexes and more

  • Declarative SQL support for Stream processing

Explore SnappyData's Vision

SnappyData's ACM SIGMOD 2016 paper is available