Hey there! The March 2026 report is here. The Apache SeaTunnel community has been incredibly active. A total of 26 contributors participated, version 2.3.13 was released, five new connectors were added, and major improvements were made across the core engine, file connectors, CDC, and Transform modules. More than 20 bugs were also fixed.
On top of that, infrastructure upgrades were rolled out. Whether you’re an enterprise or individual user, it’s a great time to upgrade, explore new features, and stay in sync with the community momentum.
Reporting period March 1, 2026 to March 30, 2026
1. Release Overview
| Version | Release Date | Notes |
|---|---|---|
| 2.3.13 | March 14, 2026 | Released this month with 50+ new features and 20+ bug fixes |
Download
https://seatunnel.apache.org/download
2. Key Updates in Version 2.3.13
2.1 New Connectors
| Connector | Description | PR |
|---|---|---|
| HugeGraph Sink | Adds support for Apache HugeGraph | #10002 |
| DuckDB | Introduces DuckDB as both Source and Sink | #10285 |
| Lance | Adds support for writing to Lance datasets | #9894 |
| AWS DSQL | Adds AWS DSQL Sink connector | #9739 |
| IoTDB | Adds Source and Sink support for IoTDB 2.x | #9872 |
2.2 Core Engine Enhancements
| Module | Feature | PR |
|---|---|---|
| Zeta Engine | Supports arbitrarily nested arrays and map types | #9881 |
| Zeta Engine | Adds min-pause checkpoint configuration | #9804 |
| Zeta Engine | Introduces REST API to inspect pending queue details | #10078 |
| Flink | Adds support for Flink 1.20.1 | #9576 |
| Flink | Enables schema evolution for CDC sources | #9867 |
| Metrics | Adds sink committed metrics and commit rate calculation | #10233 |
2.3 File Connector Improvements
| Connector | Enhancement | PR |
|---|---|---|
| HdfsFile | Enables parallel reading for large files | #10332 |
| LocalFile | Supports chunked parallel reading for CSV, TEXT, JSON files | #10142 |
| Parquet | Adds logical partitioning support | #10239 |
| HdfsFile and LocalFile | Adds sync_mode=update support | #10437, #10268 |
| HBase | Supports time-range scanning | #10318 |
| Hive | Supports automatic failover across multiple Metastore URIs | #10253 |
2.4 CDC Improvements
| Feature | Description | PR |
|---|---|---|
| Maxwell Canal Debezium | Optimizes JSON format and supports merging update_before and update_after | #9805 |
| Kafka | Adds Protobuf deserialization support via Schema Registry wire format | #10183 |
| Kafka | Injects record timestamp as EventTime metadata | #9994 |
| MySQL CDC | Optimizes wait time for schema evolution | #10040 |
2.5 Transform Enhancements
| Transformation | Feature | PR |
|---|---|---|
| Multimodal Embeddings | Adds support for multimodal embeddings | #9673 |
| RegexExtract | Introduces regex-based extraction transform | #9829 |
| SQL to Paimon | Adds support for MERGE INTO syntax | #10206 |
3. Bug Fixes in Version 2.3.13
| Module | Issue | PR |
|---|---|---|
| CSV Reader | Fixes parsing failure caused by empty first column | #10383 |
| ClickHouse | Improves batch parallel reads by replacing limit offset with last batch sort value | #9801 |
| PostgreSQL | Adds support for TIMESTAMP_TZ type | #10048 |
| Redis | Fixes cluster mode bug and adds end-to-end tests | #9869 |
| MongoDB | Improves writer close logic | #10051 |
| Elasticsearch | Optimizes resource cleanup for Scroll API | #10124 |
| MySQL CDC | Optimizes schema evolution wait time | #10040 |
4. Community Highlights
Contributors in March 2026
| Rank | Contributor | PR Count | Role |
|---|---|---|---|
| 🏅 | @zhangshenghang | 6 | Contributor |
| 🥈 | @yzeng1618 | 4 | Contributor |
| 🥈 | @davidzollo | 4 | Contributor |
| 🥈 | @chl-wxp | 4 | Contributor |
| 🥉 | @liunaijie | 3 | Contributor |
| 🥉 | @dybyte | 3 | Contributor |
| 🥉 | @ricky2129 | 3 | Contributor |
| 🥉 | @corgy-w | 3 | Contributor |
| @zooo-code | 2 | Contributor | |
| @kuleat | 2 | Contributor | |
| @LeonYoah | 2 | Contributor | |
| @OmkarK-7 | 1 | Contributor | |
| @icekimchi | 1 | Contributor | |
| @assokhi | 1 | Contributor | |
| @Sephiroth1024 | 1 | Contributor | |
| @Best2Two | 1 | Contributor | |
| @ic4y | 1 | Contributor | |
| @misi1987107 | 1 | Contributor | |
| @CosmosNi | 1 | Contributor | |
| @chocoboxxf | 1 | Contributor | |
| @xiaochen-zhou | 1 | Contributor | |
| @qingzheguo-flash | 1 | Contributor | |
| @rameshreddy-adutla | 1 | Contributor | |
| @CNF96 | 1 | Contributor | |
| @MuraliMon | 1 | Contributor | |
| @ocean-zhc | 1 | Contributor |
A total of 51 PRs were merged in March. Huge thanks to all 26 contributors.
Full contributor list
https://github.com/apache/seatunnel/graphs/contributors
Infrastructure Updates
End-to-end test Docker images migrated to the seatunnelhub repository
JDK Docker images upgraded
CI timeout optimization with Kafka set to 140 minutes and Kudu to 60 minutes
Added Metalake support for managing data source metadata
5. Recommendations for Enterprises
Upgrade Guidance
Production environments are strongly recommended to upgrade to version 2.3.13
This release includes more than 50 new features and over 20 bug fixes
Features to Watch
New connectors including HugeGraph, DuckDB, IoTDB, AWS DSQL, and Lance
Improved large file processing with parallel chunked reads in HdfsFile and LocalFile
Enhanced CDC capabilities including schema evolution and multi-format Kafka support
Improved observability with new sink committed metrics
Support for Flink 1.20.1
Notes
Some connector APIs have changed, so reviewing the upgrade documentation is recommended
Using the seatunnelhub image repository is strongly encouraged
6. Key Metrics
| Metric | March Data |
|---|---|
| Releases | 1 release (2.3.13) |
| New Connectors | 5+ |
| Feature Enhancements | 50+ |
| Bug Fixes | 20+ |
| Contributors | 50+ |
7. What’s Coming Next
Further optimization of CDC performance
More cloud-native data source integrations
Improved metrics and monitoring capabilities
Compiled and edited by the SeaTunnel Community
Top comments (0)