Ververica Platform Documentation¶
What is Ververica Platform?¶
Companies running the largest stream processing deployments in the world have adopted Apache Flink® because of its powerful model for stateful stream processing. Stateful stream processing enables companies to derive insight and to take action on data at the moment it’s generated ― when it’s the most valuable.
Ververica Platform is purpose-built for stateful stream processing architectures and makes operating these powerful systems easier than ever before by offering an entirely new experience for developing, deploying, and managing stream processing applications. It’s our mission at Ververica to ensure that developers invest their time on their core business objectives, not on maintenance and infrastructure.
There are three editions of Ververica Platform. The underlying components are briefly described below and covered in detail in the documentation of the corresponding edition.
|Edition||Spring Edition||Stream Edition||River Edition|
|Apache Flink® Support||Yes||Yes||Yes|
|Off Cycle Bug Fixes||Yes||Yes||Yes|
Application Manager is a turn key solution for running stream processing applications in production and is the core orchestration component in the Ververica Platform. It’s what allows developers to easily manage, monitor, and configure streaming jobs without worrying about the underlying infrastructure. Application Manager is stateful-streaming-aware, thus simplifying common stream processing operations tasks such as upgrading applications in a consistent manner.
Ververica Streaming Ledger, available in the River Edition of the Ververica Platform, works as a library for processing event streams across multiple shared states / tables with serializable ACID semantics.
Ververica Platform provides a highly efficient, scalable runtime for Streaming Ledger, to execute millions of transactions per second on top of Apache Flink®. It relies on a patent-pending technology for efficient, conflict-free scheduling of state accesses.
This allows the implementation of use cases going beyond today’s exactly-once processing, accessing multiple states in an atomic, consistent way. Examples include workloads currently running on traditional database systems, financial transaction processing, and multiway joins.