You keep talking about parquet and data lakes and other external data sources. But is pg_duckdb a good fit for the use case where you already have a large PostgreSQL instance with billions of rows and many tables and you want to speed up your existing analytical queries?
I think you need to store the data in Parquet files to really benefit from DuckDB. You can copy the data from your Postgres heap tables to Parquet files that you can access via DuckDB (or pg_duckdb). The speed is incredible. That being said, are Hive partitions supported?
Can we use Postgres functions on DuckDB tables? Thanks for your hard work! 👍️
How does pg_duckdb compare to other columnar extensions in Postgres?
Is Postgres likely to pickup the enhanced network protocols of DuckDB?
You keep talking about parquet and data lakes and other external data sources. But is pg_duckdb a good fit for the use case where you already have a large PostgreSQL instance with billions of rows and many tables and you want to speed up your existing analytical queries?
I think you need to store the data in Parquet files to really benefit from DuckDB. You can copy the data from your Postgres heap tables to Parquet files that you can access via DuckDB (or pg_duckdb). The speed is incredible.
That being said, are Hive partitions supported?