One really interesting problem I figured out while researching this area is the damage that has been done by giving the connection pool on the database client side and then "connection pooler" as the server side "database frontend" the same name. People often confuse these two concepts and misinterpret advice that applies to the other. We should rename the server side concept but that isn't gonna happen, so the next best thing is to be aware that this confusion exists and to communicate clearly. You can use a connection pool with and without a connection pooler and you can use a connection pooler with and without a connection pool. The right thing, however, is to always use both.
I recently went on a journey to deeply understand connecting from serverless functions to SQL databases. There is a lot of FUD around serverless "needing more concurrent connections". This kind of makes no sense but as I was benchmarking I realized that sometimes the connection