Changelog
Keep Snowflake and data-warehouse loads correct for more source data shapes
Preserve PostgreSQL array NULLs, handle oversized Redshift arrays, and write SQL Server binary values safely.
Some common source values could interrupt replication or be represented incorrectly: PostgreSQL NULL array elements could arrive as the literal string null, large Redshift array values could halt a load, and SQL Server binary primary keys could collide when stored as text.
Artie now preserves SQL NULL elements in PostgreSQL destination arrays, handles oversized array values for Redshift using the existing exceeded-value behavior, and can write SQL Server binary values as VARBINARY when raw binary values are enabled. Multi-dimensional PostgreSQL integer, boolean, and text arrays can also continue through Reader; unsupported multi-dimensional element types return a clear error.
This is for teams replicating tables with arrays or binary identifiers into PostgreSQL, Redshift, or SQL Server destinations.
Why this matters
- Preserve NULL semantics in PostgreSQL arrays
- Prevent a large new Redshift array value from stopping a pipeline
- Use binary SQL Server keys without text-collation collisions