Based on such TPCH test background:
Table: orders
Data Amount: 150,000,000
SQL to use duckdb mysql extension: copy my.orders to '/tmp/orders.mysql.duckdb.parquet' (format parquet);
Time to consume: 534 seconds.
SQL to use Doris: select * from my.test.orders into outfile "file:///tmp/orders.mysql.parquet" format as parquet;
Time to consume: 122 seconds.
With simple JDBC application to access orders table and write, it will consume about 150 seconds which is closed to Doris.
So in most of cases, duckdb_mysql takes over 4 times cost to achieve same task. It could get improved.
Based on such TPCH test background: Table: orders Data Amount: 150,000,000 SQL to use duckdb mysql extension: copy my.orders to '/tmp/orders.mysql.duckdb.parquet' (format parquet); Time to consume: 534 seconds.
SQL to use Doris: select * from my.test.orders into outfile "file:///tmp/orders.mysql.parquet" format as parquet; Time to consume: 122 seconds.
With simple JDBC application to access orders table and write, it will consume about 150 seconds which is closed to Doris.
So in most of cases, duckdb_mysql takes over 4 times cost to achieve same task. It could get improved.