You cannot apply migration for mappings configured with the multi-object source type. When you migrate to Amazon Redshift V2, the mapping fails to read from multiple Amazon Redshift source objects and displays the following error:

Operation failed: error [String index out of range: -1]

When you run an Amazon Redshift V1 mapping with an SQL query to read from a date/time column, the mapping runs successfully. The same mapping fails after the migration to Amazon Redshift V2 with the following error:

If you specify a value for the

Number of Sorted Ports

field in the advanced properties in an Amazon Redshift V1 source object, this value does not apply for Redshift V2 after the migration. To configure a sort operation, you can use the

Sort

property from the

Query

tab in the Amazon Redshift V2 Source transformation after the migration.

When you enable the

Turn on client side encryption

property to define the source attributes to read from an Amazon Redshift V1 source, these configured values are not retained after migrating to Amazon Redshift V2.

When a mapping contains a parameterized Lookup transformation, the configured encryption in the connection used in the lookup is not retained after the migration.

When you migrate and the mapping is configured with partitioning for the source, the number of rows processed by each partition varies in Amazon Redshift V2 when compared to Amazon Redshift V1.

When you configure any operation other than insert, and then migrate to Amazon Redshift V2, you must select the applicable target attribute

Treat Source Row As

in the Amazon Redshift V2 target properties before you run the mapping.

If a mapping is configured with an override to the Amazon Redshift table where the overridden columns are of greater precision, the entire column with higher precision is skipped and an exception is logged. The difference in precision between the columns at design and runtime does not affect the same mapping that uses the Amazon Redshift V1 connection.

The attribute

Number of bytes needed to support multibytes for varchar

is available in the Amazon Redshift V1 connection but not in Amazon Redshift V2. When you migrate to Amazon Redshift V2, the data appears truncated in the target columns. As a workaround, you can edit the metadata of the target before you run the mapping.

If you create a completely parameterized mapping, do not specify the connection type in the connection parameterization. If you specify the connection type in the connection parameterization and then migrate the connection, the mapping fails at runtime.