When you create a mapping in SQL ELT mode to load data to Amazon Redshift, incorrect messages are logged in the session log.

When you create a mapping in SQL ELT mode to load data to Amazon Redshift, the mapping fails in the following cases:

If you use upsert target operation.
If you read a complex file from an Amazon S3 source.
If you use the Log() function with optional arguments.

When you run multiple Amazon Redshift V2 mappings concurrently, the connector creates duplicate S3 staging locations and the mappings fail at runtime with the following error:

When you run a mapping task that is based on a mapping in SQL ELT mode, and the mapping contains a parameterized unconnected Lookup transformation, the task fails with the following error:

When you use custom query as source and enable the Select Distinct option in the advanced source properties in a mapping in SQL ELT mode, the Select Distinct query overrides the custom query and eliminates the duplicate rows.

When the lookup condition in a connected or unconnected lookup transformation contains operators other than equal to (=) and the Multiple Match Policy option is set to Return Any in a mapping in SQL ELT mode, the mapping fails.

When you enable the staging optimization property for lookups configured to use persistent cache, the mapping fails at runtime for consecutive runs. (July 2023)

When you read from an Amazon Redshift source table that has NULL data in char data type and then configure the staging optimization property, the mapping is successful but there is a mismatch in the data that is written to the target. (April 2023)

When you enable the staging optimization property with the unload command ESCAPE=OFF and run a mapping to read from an Amazon Redshift source that have data delimiters, the data written to the target is truncated until the delimiter value. (April 2023)

When you enable the staging optimization property for the source and run a parameterized mapping, the mapping task runs without optimizing the performance. (April 2023)