Paimon
Paimon source connector
Support Those Engines
Spark
Flink
SeaTunnel Zeta
Description
Read data from Apache Paimon.
Comparison between SeaTunnel and Paimon version
| Seatunnel Version | Paimon Version |
|---|---|
| 2.3.2 - 2.3.3 | 0.4-SNAPSHOT |
| 2.3.4 | 0.6-SNAPSHOT |
| 2.3.5 - 2.3.11 | 0.7.0-incubating |
| 2.3.12 - 2.3.13 | 1.1.1 |
Key Considerations for Upgrading Paimon from 0.7.0-incubating to 1.1.1
- Backup Recommendations Although compatibility is ensured, it is strongly recommended to backup critical data, especially the metadata directory, before initiating the upgrade.
- Gradual Upgrade Process
- Test Environment Validation: First validate the upgrade process in a staging environment.
- Update JAR Files: Replace Paimon JAR files with version 1.1.1.
- Automatic Format Upgrade: The system will automatically detect and upgrade older file formats.
- Configuration Check Review your configurations to ensure no deprecated options are in use. While most configurations remain backward-compatible, deprecated settings may require updates.
- Post-Upgrade Validation
Verify the following after upgrading:
- Read/Write Operations: Ensure data ingestion and retrieval workflows function normally.
- Query Performance: Confirm that query response times meet expectations.
- New Feature Verification: Test all newly introduced features (e.g., time travel, enhanced compaction) to ensure proper functionality.
Note: These steps help minimize risks and ensure a smooth transition to the stable version 1.1.1.
Key features
Options
| name | type | required | default value | description |
|---|---|---|---|---|
| warehouse | String | Yes | - | Paimon warehouse path. For rest, use the managed warehouse name instead of a path. |
| catalog_name | String | No | paimon | The name of the Paimon catalog. |
| catalog_type | String | No | filesystem | Catalog type of Paimon, supports filesystem, hive and rest. |
| catalog_uri | String | Yes when catalog_type is hive or rest | - | Catalog URI of Paimon: the Hive metastore URI, or the REST catalog endpoint. Required when catalog_type is hive or rest. |
| token.provider | String | Yes when catalog_type is rest | - | Authentication provider for the REST catalog: exactly bear or dlf. |
| token | String | Yes when token.provider is bear | - | Bearer token for the REST catalog. |
| dlf.access-key-id | String | Yes when token.provider is dlf | - | DLF access key ID. |
| dlf.access-key-secret | String | Yes when token.provider is dlf | - | DLF access key secret. |
| dlf.security-token | String | No | - | Optional temporary DLF security token. |
| dlf.region | String | No | - | DLF region. Inferred from catalog_uri when omitted. |
| database | String | Yes | - | The database you want to access. |
| table | String | Yes when table_list is absent | - | The table you want to access. Configure exactly one of table and table_list. |
| table_list | array | Yes when table is absent | - | The list of tables to read. Each item must contain database and table, and may contain its own query. |
| user | String | No | - | The Paimon user used to access the table (for example, with a hive catalog that enables authentication). |
| password | String | No | - | The Paimon user password. Required when user is configured. |
| hdfs_site_path | String | No | - | The file path of hdfs-site.xml. Deprecated; prefer paimon.hadoop.conf or paimon.hadoop.conf-path for new jobs. |
| query | String | No | - | The filter condition applied to the table read. If not specified, all rows are read. |
| paimon.hadoop.conf | Map | No | - | Properties applied to the Hadoop configuration. |
| paimon.hadoop.conf-path | String | No | - | The loading path for core-site.xml, hdfs-site.xml, and hive-site.xml files. |
warehouse [string]
Paimon warehouse path, or the managed warehouse name for REST catalogs. Must not be blank.
catalog_name [string]
The name of the Paimon catalog. Default value is paimon.
catalog_type [string]
Catalog type of Paimon. Supports filesystem, hive and rest. Default value is filesystem.
catalog_uri [string]
Catalog URI of Paimon. Required when catalog_type is hive (for example, thrift://hadoop04:9083) or rest (the REST catalog endpoint, for example https://dlf.cn-hangzhou.aliyuncs.com).
database [string]
The database you want to access.
table [string]
The table you want to access. Configure exactly one of table and table_list.
table_list [array]
The list of tables to read. Configure exactly one of table and table_list. Each item must contain database and table, and may contain its own query filter.
table_list = [
{
database = "default"
table = "table1"
query = "select * from table1 where id > 100"
},
{
database = "default"
table = "table2"
}
]
user [string]
The Paimon user used to access the table. Use it together with password. This is mainly required when the underlying catalog enforces authentication.
password [string]
The Paimon user password. Required when user is configured.
hdfs_site_path [string]
The file path of hdfs-site.xml. This option is deprecated; prefer paimon.hadoop.conf or paimon.hadoop.conf-path for new jobs.
query [string]
The filter condition of the table read. For example: select * from st_test where id > 100. If not specified, all rows are read.
Currently, where conditions only support <, <=, >, >=, =, !=, or, and,is null, is not null, between...and, in, not in, like, and others are not supported.
The Having, Group By, Order By clauses are currently unsupported, because these clauses are not supported by Paimon.
you can also project specific columns, for example: select id, name from st_test where id > 100.
Supports dynamic options settings:
SELECT * FROM table /*+ OPTIONS('incremental-between' = 'test-tag1,test-tag2') */;
Note: When the field after the where condition is a string or boolean value, its value must be enclosed in single quotes, otherwise an error will be reported. For example: name='abc' or tag='true'
The field data types currently supported by where conditions are as follows:
- string
- boolean
- tinyint
- smallint
- int
- bigint
- float
- double
- date
- timestamp
- time
paimon.hadoop.conf [string]
Properties in hadoop conf
paimon.hadoop.conf-path [string]
The specified loading path for the 'core-site.xml', 'hdfs-site.xml', 'hive-site.xml' files
Filesystems
The Paimon connector supports reading data from multiple file systems. Currently, the supported file systems are hdfs and s3.
If you use the s3 filesystem. You can configure the fs.s3a.access-key、fs.s3a.secret-key、fs.s3a.endpoint、fs.s3a.path.style.access、fs.s3a.aws.credentials.provider properties in the paimon.hadoop.conf option.
Besides, the warehouse should start with s3a://.
Examples
REST catalog authentication
catalog_type = "rest" uses the REST implementation included in Paimon 1.1.1. Set catalog_uri to the HTTPS catalog endpoint and warehouse to the managed warehouse name, not an HDFS/S3 path. The REST server supplies table locations; workers still need access and credentials for the underlying storage. Existing filesystem and Hive configurations are unchanged.
For REST, catalog_uri must be an HTTP or HTTPS URL with a host and no user information. Use HTTPS outside loopback testing: HTTP to a non-loopback host produces a warning because authentication credentials travel in clear text. token.provider is required and accepts exactly bear or dlf. bear requires a non-blank token; dlf requires both dlf.access-key-id and dlf.access-key-secret.
Bearer tokens and DLF credentials are masked in parsed-configuration logs. For configuration encryption, token is included in the default shade.identifier handling; add the DLF secret keys explicitly to shade.options to encrypt and decrypt them. Log masking alone does not encrypt configuration. Do not put credentials in endpoint URLs. Supply secrets through environment variables or your deployment's secret mechanism.
Only explicitly configured Hadoop storage properties are forwarded to REST. Catalog and authentication keys (uri, warehouse, metastore, token, token.*, dlf.*, including hadoop.-prefixed forms) are rejected in paimon.hadoop.conf and the configured Hadoop XML files. Configure REST authentication with the top-level options shown below. dlf.token-path, dlf.token-loader, dlf.token-ecs-*, and token.refresh-time are not supported in this version and cannot be enabled through paimon.hadoop.conf.
Parsed-configuration masking does not redact dependency logs. Paimon 1.1.1 logs Hadoop storage-property values, which may include credentials, when org.apache.paimon.utils.HadoopUtils DEBUG logging is enabled. Keep this logger at INFO or higher when using credential-bearing storage configuration.
Paimon 1.1.1's shared OkHttp client follows cross-host redirects and forwards x-dlf-security-token, although it strips bearer Authorization. Use trusted catalog endpoints that do not redirect to another host; do not rely on redirects to protect DLF temporary credentials. This dependency limitation is not changed by the connector.
Bearer example:
source {
Paimon {
catalog_type = "rest"
catalog_uri = "https://catalog.example.com"
warehouse = "managed-warehouse"
database = "example_db"
table = "example_table"
"token.provider" = "bear"
token = ${PAIMON_REST_TOKEN}
}
}
DLF example (replace catalog_uri with your DLF endpoint):
source {
Paimon {
catalog_type = "rest"
catalog_uri = "https://catalog.example.com"
warehouse = "managed-warehouse"
database = "example_db"
table = "example_table"
"token.provider" = "dlf"
"dlf.access-key-id" = ${DLF_ACCESS_KEY_ID}
"dlf.access-key-secret" = ${DLF_ACCESS_KEY_SECRET}
"dlf.security-token" = ${?DLF_SECURITY_TOKEN}
"dlf.region" = "cn-hangzhou"
}
}
dlf.security-token is optional and only needed for temporary credentials. dlf.region may be omitted if Paimon can infer it from the DLF endpoint. The connector forwards static credentials; it does not implement automatic refresh of temporary credentials. Live DLF authentication and cloud-storage permissions must be verified in the target deployment.
Simple example
source {
Paimon {
warehouse = "/tmp/paimon"
database = "default"
table = "st_test"
}
}
Multiple tables
source {
Paimon {
warehouse = "/tmp/paimon"
database = "default"
table_list = [
{
database = "default"
table = "table1"
query = "select * from table1 where id > 100"
},
{
database = "default"
table = "table2"
query = "select * from table2 where id > 100"
}
]
}
}
Filter example
source {
Paimon {
warehouse = "/tmp/paimon"
database = "full_type"
table = "st_test"
query = "select c_boolean, c_tinyint from st_test where c_boolean= 'true' and c_tinyint > 116 and c_smallint = 15987 or c_decimal='2924137191386439303744.39292213'"
}
}
S3 example
env {
execution.parallelism = 1
job.mode = "BATCH"
}
source {
Paimon {
warehouse = "s3a://test/"
database = "seatunnel_namespace11"
table = "st_test"
paimon.hadoop.conf = {
fs.s3a.access-key=G52pnxg67819khOZ9ezX
fs.s3a.secret-key=SHJuAQqHsLrgZWikvMa3lJf5T0NfM5LMFliJh9HF
fs.s3a.endpoint="http://minio4:9000"
fs.s3a.path.style.access=true
fs.s3a.aws.credentials.provider=org.apache.hadoop.fs.s3a.SimpleAWSCredentialsProvider
}
}
}
sink {
Console{}
}
Hadoop conf example
source {
Paimon {
catalog_name="seatunnel_test"
warehouse="hdfs:///tmp/paimon"
database="seatunnel_namespace1"
table="st_test"
query = "select * from st_test where pk_id is not null and pk_id < 3"
paimon.hadoop.conf = {
hadoop_user_name = "hdfs"
fs.defaultFS = "hdfs://nameservice1"
dfs.nameservices = "nameservice1"
dfs.ha.namenodes.nameservice1 = "nn1,nn2"
dfs.namenode.rpc-address.nameservice1.nn1 = "hadoop03:8020"
dfs.namenode.rpc-address.nameservice1.nn2 = "hadoop04:8020"
dfs.client.failover.proxy.provider.nameservice1 = "org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider"
dfs.client.use.datanode.hostname = "true"
}
}
}
Hive catalog example
source {
Paimon {
catalog_name="seatunnel_test"
catalog_type="hive"
catalog_uri="thrift://hadoop04:9083"
warehouse="hdfs:///tmp/seatunnel"
database="seatunnel_test"
table="st_test3"
paimon.hadoop.conf = {
fs.defaultFS = "hdfs://nameservice1"
dfs.nameservices = "nameservice1"
dfs.ha.namenodes.nameservice1 = "nn1,nn2"
dfs.namenode.rpc-address.nameservice1.nn1 = "hadoop03:8020"
dfs.namenode.rpc-address.nameservice1.nn2 = "hadoop04:8020"
dfs.client.failover.proxy.provider.nameservice1 = "org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider"
dfs.client.use.datanode.hostname = "true"
}
}
}
paimon enable privilege example
source {
Paimon {
warehouse = "/tmp/paimon"
database = "default"
table = "st_test"
user = "paimon"
password = "******"
}
}
Read Paimon Table Changelog
If you want to read the changelog of the Paimon table, first set the changelog-producer for the Paimon source table and then use the SeaTunnel stream task to read it.
Note
Currently, batch reads are always the latest snapshot read, so to read full changelog data, you need to use stream reads and start stream reads before writing data to the Paimon table. To ensure ordering, the parallelism of the stream read task should be set to 1.
Streaming read example
env {
parallelism = 1
job.mode = "STREAMING"
}
source {
Paimon {
warehouse = "/tmp/paimon"
database = "full_type"
table = "st_test"
}
}
sink {
Paimon {
warehouse = "/tmp/paimon"
database = "full_type"
table = "st_test_sink"
paimon.table.primary-keys = "c_tinyint"
}
}
FAQ
Which read modes does the Paimon source support?
The SeaTunnel Paimon source supports both batch snapshot queries and streaming changelog consumption. In batch mode, it scans the target table's latest snapshot or a specified snapshot range. In streaming mode, it continuously consumes newly committed changes and splits.
How do I configure storage and catalog backends for Paimon?
Configure warehouse pointing to the storage root (e.g. hdfs:///paimon/warehouse, s3a://bucket/warehouse, or local path) and set catalog_type (such as filesystem or hive). Required storage or authentication properties can be provided via paimon.hadoop.conf.
For catalog_type = "rest", use the managed warehouse name instead of a storage path and set catalog_uri and the catalog authentication options shown above. Storage properties still belong in paimon.hadoop.conf; they are separate from REST catalog credentials.
Does Paimon source support column projection?
Yes. The Paimon source reads only the projection fields defined in the SeaTunnel job schema, avoiding unnecessary column deserialization and optimizing read performance on columnar formats like ORC and Parquet.
Changelog
Change Log
| Change | Commit | Version |
|---|---|---|
| [Feature][Connectors-v2] Clean up temporary files for paimon sink (#9819) | https://github.com/apache/seatunnel/commit/c43d57de31 | dev |
| [Feature][Connector-v2] Support multi paimon source (#9759) | https://github.com/apache/seatunnel/commit/0d52102241 | 2.3.12 |
| [Chore] fix typos filed -> field (#9757) | https://github.com/apache/seatunnel/commit/e3e1c67d29 | 2.3.12 |
| [Feature][connector-paimon] Paimon connector supports paimon privilege (#9722) | https://github.com/apache/seatunnel/commit/b2bb2f8d78 | 2.3.12 |
| [Improve][Core] Update apache common to apache common lang3 (#9694) | https://github.com/apache/seatunnel/commit/6e5737c1ec | 2.3.12 |
| [Improve][API] Optimize the enumerator API semantics and reduce lock calls at the connector level (#9671) | https://github.com/apache/seatunnel/commit/9212a77140 | 2.3.12 |
| [fix][connectors-v2] repeated commit cause task exceptions (#9665) | https://github.com/apache/seatunnel/commit/085023ad0d | 2.3.12 |
| [Improve][Connector-V2] Support like predicate pushdown in paimon (#9653) | https://github.com/apache/seatunnel/commit/9e01c84e76 | 2.3.12 |
| [Feature][Connectors-v2]Paimon version upgrade to 1.1.1 (#8074) | https://github.com/apache/seatunnel/commit/96b26a68dc | 2.3.12 |
| [Fix][Connectors-v2] fix dynamic bucket for paimon sink (#9595) | https://github.com/apache/seatunnel/commit/d29a531a48 | 2.3.12 |
| [Feature][Connector-V2] Support like predicate pushdown in paimon (#9484) | https://github.com/apache/seatunnel/commit/a19720ccf6 | 2.3.12 |
| [Fix][Connector-V2] Update waitCompaction value for batch mode and writeonly (#9479) | https://github.com/apache/seatunnel/commit/63993a6197 | 2.3.12 |
| [Future][Connector-V2]Support the automatic creation of non-primary key table (#9219) | https://github.com/apache/seatunnel/commit/93e539cc9f | 2.3.12 |
| [Fix][Connector-V2] Optimize Paimon DECIMAL type check to prevent precision loss (#9480) | https://github.com/apache/seatunnel/commit/c114682a6b | 2.3.12 |
| [Bug][Connector-V2] fix NPE when decimal type precision is incompatible for Paimon (#9452) | https://github.com/apache/seatunnel/commit/37762c93f0 | 2.3.12 |
| [feature][connectors-v2] Support in predicate pushdown in paimon (#9379) | https://github.com/apache/seatunnel/commit/1ec43755d5 | 2.3.12 |
| [Improve][Connector-V2] Fix the word misspellings for paimon connector (#9332) | https://github.com/apache/seatunnel/commit/ba7f5c9e30 | 2.3.11 |
| [Feature][Transform] Support define sink column type (#9114) | https://github.com/apache/seatunnel/commit/ab7119e507 | 2.3.11 |
| [improve] paimon options (#9167) | https://github.com/apache/seatunnel/commit/b0889305c2 | 2.3.11 |
| [Fix][Paimon] nullable and comment attribute was lost during automatic table creation (#9020) | https://github.com/apache/seatunnel/commit/eb54fdd52c | 2.3.11 |
| [Feature][Connector-V2] Support between predicate pushdown in paimon (#8962) | https://github.com/apache/seatunnel/commit/3b141cf621 | 2.3.10 |
| [Feature][Connector-V2] Suppor Time type in paimon connector (#8880) | https://github.com/apache/seatunnel/commit/9f1e590091 | 2.3.10 |
| [Feature][Paimon] Customize the hadoop user (#8888) | https://github.com/apache/seatunnel/commit/2657626f93 | 2.3.10 |
| [Improve][Connector-v2][Paimon]PaimonCatalog close error message update (#8640) | https://github.com/apache/seatunnel/commit/48253da8d6 | 2.3.10 |
| [Improve] restruct connector common options (#8634) | https://github.com/apache/seatunnel/commit/f3499a6eeb | 2.3.10 |
| [Improve][Connector-v2] Support checkpoint in batch mode for paimon sink (#8333) | https://github.com/apache/seatunnel/commit/f22d4ebd4d | 2.3.9 |
| [Feature][Connector-v2] Support schema evolution for paimon sink (#8211) | https://github.com/apache/seatunnel/commit/57190e2a3b | 2.3.9 |
| [Improve][dist]add shade check rule (#8136) | https://github.com/apache/seatunnel/commit/51ef800016 | 2.3.9 |
| [Feature][Connector-v2] Support S3 filesystem of paimon connector (#8036) | https://github.com/apache/seatunnel/commit/e2a4772933 | 2.3.9 |
| [Feature][transform] transform support explode (#7928) | https://github.com/apache/seatunnel/commit/132278c06a | 2.3.9 |
| [Feature][Connector-V2] Piamon Sink supports changelog-procuder is lookup and full-compaction mode (#7834) | https://github.com/apache/seatunnel/commit/c0f27c2f76 | 2.3.9 |
| [Fix][connector-v2]Fix Paimon table connector Error log information. (#7873) | https://github.com/apache/seatunnel/commit/a3b49e6354 | 2.3.9 |
| [Improve][Connector-v2] Use checkpointId as the commit's identifier instead of the hash for streaming write of paimon sink (#7835) | https://github.com/apache/seatunnel/commit/c7a384af2b | 2.3.9 |
| [Feature][Restapi] Allow metrics information to be associated to logical plan nodes (#7786) | https://github.com/apache/seatunnel/commit/6b7c53d03c | 2.3.9 |
| [Fix][Connecotr-V2] Fix paimon dynamic bucket tale in primary key is not first (#7728) | https://github.com/apache/seatunnel/commit/dc7f695537 | 2.3.8 |
| [Improve][Connector-v2] Remove useless code and add changelog doc for paimon sink (#7748) | https://github.com/apache/seatunnel/commit/846d876dc2 | 2.3.8 |
| [Hotfix][Connector-V2] Release resources even the task is crashed for paimon sink (#7726) | https://github.com/apache/seatunnel/commit/5ddf8d461e | 2.3.8 |
| [Fix][Connector-V2] Fix paimon e2e error (#7721) | https://github.com/apache/seatunnel/commit/61d1964361 | 2.3.8 |
| [Feature][Connector-Paimon] Support dynamic bucket splitting improves Paimon writing efficiency (#7335) | https://github.com/apache/seatunnel/commit/bc0326cba8 | 2.3.8 |
| [Feature][Connector-v2] Support streaming read for paimon (#7681) | https://github.com/apache/seatunnel/commit/4a2e27291c | 2.3.8 |
| [Hotfix][Seatunnel-common] Fix the CommonError msg for paimon sink (#7591) | https://github.com/apache/seatunnel/commit/d1f5db9257 | 2.3.8 |
| [Feature][CONNECTORS-V2-Paimon] Paimon Sink supported truncate table (#7560) | https://github.com/apache/seatunnel/commit/4f3df22124 | 2.3.8 |
| [Improve][Connector-v2] Improve the exception msg in case-sensitive case for paimon sink (#7549) | https://github.com/apache/seatunnel/commit/7d31e5668c | 2.3.8 |
| [Hotfix][Connector-V2] Fixed lost data precision for decimal data types (#7527) | https://github.com/apache/seatunnel/commit/df210ea73d | 2.3.8 |
| [Improve][API] Move catalog open to SaveModeHandler (#7439) | https://github.com/apache/seatunnel/commit/8c2c5c79a1 | 2.3.8 |
| [Improve][Connector] Add multi-table sink option check (#7360) | https://github.com/apache/seatunnel/commit/2489f6446b | 2.3.7 |
| The isNullable attribute is true when the primary key field in the Paimon table converts the Column object. #7231 (#7242) | https://github.com/apache/seatunnel/commit/b0fe432e99 | 2.3.6 |
| [Feature][Core] Support using upstream table placeholders in sink options and auto replacement (#7131) | https://github.com/apache/seatunnel/commit/c4ca74122c | 2.3.6 |
| [Paimon]support projection for paimon source (#6343) | https://github.com/apache/seatunnel/commit/6c1577267f | 2.3.6 |
| [Improve][Paimon] Add check for the base type between source and sink before write. (#6953) | https://github.com/apache/seatunnel/commit/d56d64fc04 | 2.3.6 |
| [Improve][Connector-V2] Improve the paimon source (#6887) | https://github.com/apache/seatunnel/commit/658643ae53 | 2.3.6 |
| [Hotfix][Connector-V2] Close the tableWrite when task is close (#6897) | https://github.com/apache/seatunnel/commit/23a744b9b2 | 2.3.6 |
| [Fix][Connector-V2] Field information lost during Paimon DataType and SeaTunnel Column conversion (#6767) | https://github.com/apache/seatunnel/commit/6cf6e41da7 | 2.3.6 |
| [Improve][Connector-V2] Support hive catalog for paimon sink (#6833) | https://github.com/apache/seatunnel/commit/4969c91dc4 | 2.3.6 |
| [Hotfix][Connector-V2] Fix the batch write with paimon (#6865) | https://github.com/apache/seatunnel/commit/9ec971d942 | 2.3.6 |
| [Feature][Doris] Add Doris type converter (#6354) | https://github.com/apache/seatunnel/commit/5189991843 | 2.3.6 |
| [Improve][Connector-V2] Support hadoop ha and kerberos for paimon sink (#6585) | https://github.com/apache/seatunnel/commit/20b62f3bf3 | 2.3.5 |
| [Feature][Paimon] Support specify paimon table write properties, partition keys and primary keys (#6535) | https://github.com/apache/seatunnel/commit/2b1234c7ae | 2.3.5 |
| [Feature][Connector-V2] Support multi-table sink feature for paimon #5652 (#6449) | https://github.com/apache/seatunnel/commit/b0abbd2d89 | 2.3.5 |
| [Feature][Connectors-v2-Paimon] Adaptation Paimon 0.6 Version (#6061) | https://github.com/apache/seatunnel/commit/b32df930e9 | 2.3.4 |
| [Fix][Connectors-v2-Paimon] Flink table store failed to prepare commit (#6057) | https://github.com/apache/seatunnel/commit/c8dcefc3be | 2.3.4 |
| [Improve][Common] Introduce new error define rule (#5793) | https://github.com/apache/seatunnel/commit/9d1b2582b2 | 2.3.4 |
[Improve] Remove use SeaTunnelSink::getConsumedType method and mark it as deprecated (#5755) | https://github.com/apache/seatunnel/commit/8de7408100 | 2.3.4 |
| [Hotfix][Connector-V2][Paimon] Bump paimon-bundle version to 0.4.0-incubating (#5219) | https://github.com/apache/seatunnel/commit/2917542bfa | 2.3.3 |
| [Improve] Documentation and partial word optimization. (#4936) | https://github.com/apache/seatunnel/commit/6e8de0e2a6 | 2.3.3 |
| [Connector-V2][Paimon] Introduce paimon connector (#4178) | https://github.com/apache/seatunnel/commit/da507bbe0e | 2.3.2 |