Postgres insert optimization

Question

I have a script that generates tens of thousands of inserts into a postgres db through a custom ORM. As you can imagine, it\'s quite slow. This is used for development purpose

Answer 1

Try to do as much as possible in one request!

insert into my_table (col1, col2) 
values (
  unnest(array[row_1_col_value_1, row_2_col_value_1, row3_col_value_1]), 
  unnest(array[row_1_col_value_2, row_2_col_value_2, row_3_col_value_2));

This resembles the suggestion of @a_horse_with_no_name. The advantage of using unnest is: You can use query parameters that contain arrays!

insert into my_table (col1, col2) 
values (unnest(:col_values_1), unnest(:col_values_2));

By collapsing three insert statements into one, you save more than 50% of execution time. And by using query parameters with 2000 values in a single Insert, I get a speed factor of 150 in my application.