Preamble: This is not an authoritative roadmap in the sense of a strict plan with a specific timeline. Rather, we, the community, share our vision for the future and give an overview of the bigger initiatives that are going on and are receiving attention. This roadmap shall give users and contributors an understanding where the project is going and what they can expect to come.
The roadmap is continuously updated. New features and efforts should be added to the roadmap once there is consensus that they will happen and what they will roughly look like for the user.
Last Update: 2019-05-08
Flink views stream processing as a unifying paradigm for data processing (batch and real-time) and event-driven applications. The APIs are evolving to reflect that view:
The Table API / SQL is becoming the primary API for analytical use cases, in a unified way across batch and streaming. To support analytical use cases in a more streamlined fashion, the API is extended with additional functions (FLIP-29).
Like SQL, the Table API is declarative, operates on a logical schema, and applies automatic optimization. Because of these properties, that API does not give direct access to time and state.
The DataStream API is the primary API for data-driven applications and data pipelines. It uses physical data types (Java/Scala classes) and there is no automatic rewriting. The applications have explicit control over time and state (state, triggers, proc. fun.).
In the long run, the DataStream API should fully subsume the DataSet API through bounded streams.
Flink’s approach is to cover batch and streaming by the same APIs, on a streaming runtime. This blog post gives an introduction to the unification effort.
The biggest user-facing parts currently ongoing are:
Table API restructuring FLIP-32 that decouples the Table API from batch/streaming specific environments and dependencies.
The new source interfaces FLIP-27 generalize across batch and streaming, making every connector usable as a batch and streaming data source.
The introduction of upsert- or changelog- sources FLINK-8545 will support more powerful streaming inputs to the Table API.
On the runtime level, the streaming operators are extended to also support the data consumption patterns required for some batch operations (discussion thread). This is also groundwork for features like efficient side inputs.
The community’s goal is to make Flink’s performance on bounded streams (batch use cases) competitive with that of dedicated batch processors. While Flink has been shown to handle some batch processing use cases faster than widely-used batch processors, there are some ongoing efforts to make sure this the case for broader use cases:
Faster and more complete SQL/Table API: The community is merging the Blink query processor which improves on the current query processor by adding a much richer set of runtime operators, optimizer rules, and code generation. The new query processor will have full TPC-DS support and up to 10x performance improvement over the current query processor (FLINK-11439).
Exploiting bounded streams to reduce the scope of fault tolerance: When input data is bounded, it is possible to completely buffer data during shuffles (memory or disk) and replay that data after a failure. This makes recovery more fine grained and thus much more efficient (FLINK-10288).
An application on bounded data can schedule operations after another, depending on how the operators consume data (e.g., first build hash table, then probe hash table). We are separating the scheduling strategy from the ExecutionGraph to support different strategies on bounded data (FLINK-10429).
Caching of intermediate results on bounded data, to support use cases like interactive data exploration. The caching generally helps with applications where the client submits a series of jobs that build on top of one another and reuse each others’ results. FLINK-11199
External Shuffle Services (mainly bounded streams) to support decoupling from computation and intermediate results for better resource efficiency on systems like Yarn. FLIP-31.
Various of these enhancements can be taken from the contributed code from the Blink fork.
To exploit the above optimizations for bounded streams in the DataStream API, we need break parts of the API and explicitly model bounded streams.
Flink will get the new modes to stop a running application while ensuring that output and side-effects are consistent and committed prior to shutdown. SUSPEND commit output/side-effects, but keep state, while TERMINATE drains state and commits the outputs and side effects. FLIP-34 has the details.
The new source interface effort (FLIP-27) aims to give simpler out-of-the box support for event time and watermark generation for sources. Sources will have the option to align their consumption speed in event time, to reduce the size of in-flight state when re-processing large data volumes in streaming (FLINK-10887).
There is a big effort to design a new way for Flink to interact with dynamic resource pools and automatically adjust to resource availability and load. Part of this is becoming a reactive way of adjusting to changing resources (like containers/pods being started or removed) FLINK-10407, while other parts are resulting in active scaling policies where Flink decides to add or remove TaskManagers, based on internal metrics.
To support the active resource management also in Kubernetes, we are adding a Kubernetes Resource Manager FLINK-9953.
The Flink Web UI is being ported to a newer framework and getting additional features for better introspection of running jobs FLINK-10705.
The community is working on extending the interoperability with authentication and authorization services. Under discussion are general extensions to the security module abstraction as well as specific enhancements to the Kerberos support.
The community is working on extending the support for catalogs, schema registries, and metadata stores, including support in the APIs and the SQL client (FLINK-11275). We are adding DDL (Data Definition Language) support to make it easy to add tables and streams to the catalogs (FLINK-10232).
There is a broad effort to integrate Flink with the Hive Ecosystem, including metastore and Hive UDF support FLINK-10556.
There is also a big effort to support Python for Table API FLIP-38. We will divide the work into following stages:
Support for additional connectors and formats is a continuous process.
To reduce compatibility issues with different Scala versions, we are working using Scala only in the Scala APIs, but not in the runtime. That removes any Scala dependency for all Java-only users, and makes it easier for Flink to support different Scala versions FLINK-11063.