Ask HN: Databases have indexes. Then why have partitions?
Because you are keeping 13 months of phone call data and want to easily drop the oldest month every 1st.
Because your queries need to be fast, and you can chop up your data into 1000 pieces (by some column or combination) without adding complexity.
You can easily drop the oldest month every 1st using a delete statement.
Queries can be fast with indexes without adding complexity.
Uh, what?
I think this is a legitimate question. I always wondered the same thing.