Hello .. just checking if Meltano ELT supports the following: • Right now, we have an etl pipeline that uses tap-mysql as an extractor and target-snowflake as a loader. • One of the tables, named

table_a

, schema was changed as following ◦ removed

user_id

primary key (and by doing so, its unique constraint) ◦ added new ID column (

bigint

, autoincrement), set as new primary key ◦ renamed column

last_checkin

to

time

My question: how would this impact the newly extracted data? also, would this require some changes on the snowflake side? or is it going to keep the old data and add (new column) to the existing table we have in snowflake? because we don't want to contaminate the old data - and we only want to add to it, like the newly renamed column would appear as a separate one? or should it be mapped to keep the old data and build on top of it?

The beauty of Meltano is you can test this with a test environment in Snowflake very quickly. Yes the new

id

column would appear. The renamed column would look like a new column to the extraction so

time

would just start getting populated for new rows. It will keep the old data (assuming you're running incremental and not full table syncs with activate version). highly recommend you try it first and make sure you understand before just running it so you feel confident

Perfect! Appreciate the response ..

The renamed column would look like a new column to the extraction so

time

would just start getting populated for new rows.

Would this still apply if the

last_checkin

was used as a replication key and renamed to

time

? Can't test against prod data - and I only have prod copy atm ..