Dear community,
happy to share a belated update for the past week. This time with the release of Flink 1.11.2, a couple of discussions and FLIPs on improving Flink's APIs and dropping some baggage, most notably Scala 2.11, a new unified sink API and a bit more. Flink Development ============== * [releases] The vote for Flink 1.11.2 has passed today. Release announcement will follow shortly. [1] * [releases] Feature Freeze for Stateful Function 2.2.0 happened as planned five days ago. Release candidate is expected soon. [2] * [apis] Seth has started a discussion to drop Scala 2.11. Many in favor so far. [3] * [apis] Seth proposes to separate the concepts of "Statebackend" and "Checkpointing" in Flink's APIs. So far, the statebackend (RocksDB vs Heap-Based) and checkpointing storage (usually a distributed file system) are handled by the same classes and configured together. He believes this is the reason for some of the confusion of users around these concepts. [4,5] * [apis] Aljoscha started a discussion to deprecate and later remove UnionList OperatorState. Some current users of UnionList OperatorState have voiced concerns. No conclusion so far. [6] * [connectors] Guowei has published FLIP-143 "Unified Sink API". This is another follow up FLIP in order for the DataStream to supersede the DataSet API (FLIP-131). The basic idea is to provide an interface that allows the development of sinks that provide exactly-once guarantees for both bounded and unbounded workloads, but don't require the developer of the sink to make this distinction. [6,7] [3] http://apache-flink-mailing-list-archive.1008284.n3.nabble.com/DISCUSS-Drop-Scala-2-11-tp44607.html ============== * Jark has published an implementation of the Nexmark benchmark for Apache Flink. [8] Notable Bugs ========== Nothing came to my attention. Events, Blog Posts, Misc =================== * Arvid Heise & Niels Basjes are Apache Flink Committers now! Congratulations to both. [9,10] * The third video of my colleague Alexander's "Introduction to Flink" series has been published. This time about building event-driven applications with Apache Flink. [11] * On the Ververica Blog, Jark and Qingshen give an outlook to their upcoming talk at Flink Forward Global [12] on Change Data Capture with Flink SQL. [13] Cheers, Konstantin |
Free forum by Nabble | Edit this page |