Status
Current state: Under DiscussionReleased
Discussion thread: here (<- link to https://mail-archiveslists.apache.org/mod_mbox/flink-dev/)
JIRA: here (<- link to https://issues.apache.org/jira/browse/FLINK-XXXX)
thread/hod6bg421bzwhbfv60lwsck7r81dvo59
JIRA:
| Jira | ||||||
|---|---|---|---|---|---|---|
|
Released: 1.20Released: <Flink Version>
Please keep the discussion on the mailing list rather than commenting on the wiki (wiki discussions get unwieldy fast).
...
This FLIP proposes to promote the Unified Sink API V2 from PublicEvolving to Public and to mark the SinkFunction as Deprecated. Since its introduction in Flink 1.12, the Unified Sink API has undergone extensive development and testing, evidenced by its evolution across multiple FLIPs and its adoption in major connectors like Kafka, CassandraFileSystem, and Elasticsearch AWS since Flink 1.14 and beyond. Over more then four release cycles, the API has demonstrated stability and robustness, aligning with the criteria set forth in FLIP-197 for API stability graduation. This promotion is expected to encourage wider adoption by signaling the API’s maturity and reliability to the user base. This step is essential for standardizing Flink’s API landscape, much like the transition from SourceFunction to Source API, thereby enhancing the framework's overall functionality and maintainability.
The following table shows the relevant FLIPs that lead to the SinkV2 API as it has been released with Flink 1.19
| FLIP | Note | API Annotation | Released with |
|---|---|---|---|
| FLIP-143: Unified Sink API | Introduction of new Unified Sink API | Experimental | Flink 1.12 |
| FLIP-177: Extend Sink API | Extends Unified Sink API | Experimental | Flink 1.14 |
| FLIP-171: Async Sink | Introduces generic Async Sink API, based on the Unified Sink API | PublicEvolving | Flink 1.15 |
| FLIP-191: Extend unified Sink interface to support small file compaction | Extends Unified Sink API with introduction of SinkV2 | Sink: SinkV2: | Flink 1.15 |
| FLIP-371: Provide initialization context for Committer creation in TwoPhaseCommittingSink | Added ability to emit metrics from the committer | SinkV2: | Flink 1.19 |
| FLIP-372: Enhance and synchronize Sink API to match the Source API | Changed the Sink V2 API so that it uses mixin interfaces to enhance the extendibility of the API, similar to the Source API. | SinkV2: | Flink 1.19 |
The following table shows the state of migration to SinkV2 for ASF supported and released connectors:
| Status | Connector | Used Sink/SinkV2 since | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
 | FileSystem | Flink 1.12
| ||||||||||||||||
| Kafka / Upsert-Kafka | Flink 1.14
| ||||||||||||||||
 | Cassandra |
| ||||||||||||||||
| DynamoDB | AWS Connector v3.0 (via ASync API)
| ||||||||||||||||
| Elasticsearch | Flink 1.15
| ||||||||||||||||
| Firehose | Flink 1.15 (via ASync API)
| ||||||||||||||||
| Kinesis | Flink 1.15 (via ASync API)
| ||||||||||||||||
| MongoDB | MongoDB v1.0 (via ASync API)
| ||||||||||||||||
| Opensearch | Opensearch v1.0 (via ASync API)
| ||||||||||||||||
| RabbitMQ |
| ||||||||||||||||
| Google Cloud PubSub |
| ||||||||||||||||
| Pulsar | Flink 1.15
| ||||||||||||||||
| JDBC | JDBC v3.2.0
| ||||||||||||||||
| HBase |
|
This topic has also been previously discussed on the Dev mailing list in https://lists.apache.org/thread/q62nj89rrz0t5xtggy5n65on95f2rmmx
Public Interfaces
...
- The following interfaces will be
...
- marked as
@Deprecated- org.apache.flink.streaming.api.functions.sink#SinkFunction
- org.apache.flink.streaming.api.functions.sink#SocketClientSink
- org.apache.flink.streaming.api.functions.sink#TwoPhaseCommitSinkFunction
- The following interfaces will be marked as
@Public- org.apache.flink.api.connector.sink2#Committer
- org.apache.flink.api.connector.sink2#CommitterInitContext
- org.apache.flink.api.connector.sink2#CommittingSinkWriter
- org.apache.flink.api.connector.sink2#Sink
- org.apache.flink.api.connector.sink2#SinkWriter
- org.apache.flink.api.connector.sink2#StatefulSinkWriter
- org.apache.flink.api.connector.sink2#SupportsCommitter
- org.apache.flink.api.connector.sink2#SupportsWriterState
- org.apache.flink.api.connector.sink2#WriterInitContext
Proposed Changes
Properly annotate the above interfaces with the designated annotation.
...
A public interface is any change to the following:
- DataStream and DataSet API, including classes related to that, such as StreamExecutionEnvironment
- Classes marked with the @Public annotation
- On-disk binary formats, such as checkpoints/savepoints
- User-facing scripts/command-line tools, i.e. bin/flink, Yarn scripts, Mesos scripts
- Configuration settings
- Exposed monitoring information
Proposed Changes
Describe the new thing you want to do in appropriate detail. This may be fairly extensive and have large subsections of its own. Or it may be a few sentences. Use judgement based on the scope of the change.
Compatibility, Deprecation, and Migration Plan
- What impact (if any) will there be on existing users?
- If we are changing behavior how will we phase out the older behavior?
- If we need special migration tools, describe them here.
- When will we remove the existing behavior?
Test Plan
Describe in few sentences how the FLIP will be tested. We are mostly interested in system tests (since unit-tests are specific to implementation details). How will we know that the implementation works as expected? How will we know nothing broke?
Rejected Alternatives
- We need to include in the next release notes that these interfaces are deprecated / designated to by annotated as Public
- Change the existing org.apache.flink.streaming.api.functions.sink#DiscardingSink, org.apache.flink.streaming.api.functions.sink#PrintSink and org.apache.flink.streaming.api.functions.sink#PrintSinkFunction to use the new interfaces
Test Plan
Existing tests
Rejected Alternatives
N/AIf there are alternative ways of accomplishing the same thing, what were they? The purpose of this section is to motivate why the design is the way it is and not some other way.