Apache Flink: Building a Stream Processor for fast analytics, event-driven applications, event time, and tons of state
Apache Flink is known today largely as a stream processor for high-volume streams with low latency, event time aware,
supporting large and consistent state. Following the "batch is a special case of streaming" paradigm, Flink doubles
also as a fast batch processor.
Over the recent past, we see more and more users using Flink to build *stateful event-driven* applications, for example
applications that directly power online services behind websites or social networks. With features like "queryable state",
we see Flink taking broader roles than "just stream processing" in the application stack.
In the talk, we describe some of these new applications and take a look at Flink's core architecture and building
blocks (streams, state, event time, checkpoints) and why this architecture turns out to lend itself so well to building
both stream processing jobs and stateful event-driven applications.
In a second part of the talk, we will talk about how we are pushing Flink further to handle even bigger use cases:
We will in particular take a look at new features like
- broader support of Stream SQL (and its embedded counterpart, the Table API)
- next steps for Queryable Stat
- support much larger application state inside the Flink runtime (incremental and asynchronous checkpoints, finer grained recovery)
- Flink's new resource management model which elastically allocates and releases resources from Yarn / Mesos / etc, supports more fine grained control over process isolation and container sizes.
Stephan Ewen is PMC member of Apache Flink and co-founder and CTO of data Artisans.
Before founding data Artisans, Stephan was leading the development that led to the creation of Apache Flink. He holds a Ph.D. from TU Berlin.
Launchpad Space (4th floor, 301 Howard, SF)
Subscribe here (https://goo.gl/forms/kllPDSpCSzuJOX7I2) to receive biweekly event updates from Launchpad Space, Google's new event Space in SF where developers and startups can receive free technical training, one-on-one mentoring and more!