Date: October 24, 2018 Time: 6:00pm EST Duration: 1 Hour
Apache Flink is an Apache Software Foundation open source framework, more mature and superior to Spark, for distributed stream and batch processing.
Flink features include “Streaming-first” processing. This is not micro-batching but true event-time streaming that was built into Flink from the start. Streams are treated as inifinite, unbounded data sets in Flink. Batch processing is little more than a subset of Flinks capabilities as Batch data is treated as “finite streams” or bounded data sets. Flink provides a flexible API to handle late and out-of-order data. It provides stateful/fault tolerant streaming with exactly-once processing semantics. Along with its programming API, Flink has powerful SQL and R interfaces for event time data analysis.
Flink performs uniformly at large scale, running on thousands of nodes. It can run in single JVM, standalone cluster, YARN, or cloud.
This is why companies like Alibaba, Bouygues Telecom, Capital One, Comcast, Ebay, Ericsson, Huawei,King.com,Lyft,Netflix, Portugal Telecom, ResearchGate, Styx, Telefonica, Uber and Zalando all trust Flink in mission critical production.
SLI will be sponsoring Chris Ruegger, a Flink SME, in its first DUG talk (www.dugtalks.com) October 24. Please be sure to register for this first of many DUG Talks at SLI!