Ingest Java SDK release notes for 2024¶

With this release, Snowpipe Streaming can ingest data into Snowflake-managed Apache Iceberg tables.

Fixed dependency issues and error messages in the SDK.

Beginning with release 2.3.0, numeric values preserve their format. The numeric values will not be converted to and from scientific notation.

Made updates to support a new table format.

Fixed vulnerable dependencies.

Upgraded Hadoop to fix vulnerability issues.

Removed unnecessary dependencies to reduce JAR size.

Fixed a critical issue by updating the location for the file name in metadata.

Added ExternalVolumeManager to support multiple stages for a new table format.

Upgraded dependency versions.

Updated parameters to support a new table format.

Improved code logic to support different storage volumes.

Fixed a critical issue that could potentially cause conflicts when change_tracking is enabled for streams and dynamic tables.

Improved InsertRows performance.

Added or improved various logs for better observability.

Fine-tuned channel and chunk sizes.

Fixed an issue with failover across deployments.

Returned more detailed error messages for the INVALID_CHANNEL error.

Added support for external OAuth 2.0.

Upgraded several dependencies, including vulnerability fixes.

Fixed an issue where HTTP connections are leaked due to error responses.

Relaxed the file size constraints to deal with issues where longer client flush lags produce larger files.

Set Zstandard as the default compression algorithm.

Allowed clients to drop channels.

Upgraded JDBC to 3.14.5.

Implemented a change for sending the start and end offset tokens for a channel.

Implemented a change for sending the column ordinal data to the server side for cross-checking table schema changes.

Added support to pass verification logic for a user-defined offset token as part of channel creation.

Fixed an overflow issue that caused silent data issue.

Added an optional offset token parameter for openChannel.

Added support for specifying compression algorithm to be used for BDEC Parquet files.

Updated to support customized URL and added Snowflake account name in request header.

Implemented a change to send spansMixedTables flag in blob registration requests.

Deprecated BUFFER_FLUSH_INTERVAL_IN_MILLIS parameter, instead use the MAX_CLIENT_LAG parameter.

Implemented the refresh of downscoped GCS tokens.

Reverted one change that updated public API for internal use case.

Fixed the end-to-end JAR test so it can run on all cloud platforms.