I migrated data from Postgres to Redshift and doing data validation checks, so taking MD5 for complete row. Visual checks are ok between Postgres to Redshift, but MD5 value differs for Float8 data types.
In Postgres:
balance|md5(balance::VARCHAR)
-140.2|d59f47f21a88e8b73d4ca309d75cc64b
In Redshift:
balance|md5(balance::VARCHAR)
-140.2|539dc58d834b1cf24252705b4f40b7f1
Anyone got this issue? any thought on why is this difference? Is there any other way to do data validation for all rows across database after migration?