Sample Data (Faker) Migration Guide
All streams (users, products, and purchases) now properly declare id as their respective primary keys. Existing sync jobs should still work as expected but you may need to reset your sync and/or update write mode after upgrading to the latest connector version.
Some columns are narrowing from number to integer. You may need to force normalization to rebuild your destination tables by manually dropping the SCD and final tables, refreshing the connection schema (skipping the reset), and running a sync. Alternatively, you can just run a reset.
Nothing to do here - this was a test breaking change.