In citusDB, how do I find out how my data is distributed?

Question

In CitusDB, I can create an empty table with:

CREATE TABLE table1 (col1 text, col2 text);

I can tell table1 how to partition the data, which will later be loaded into the table, by running this:

SELECT create_distributed_table('table1', 'col1');

In this moment, I will then know how my table is distributed across CitusDB nodes.

However, if I come across a new table that I didn't create, but I know it is distributed, how do I know what column the table is distributed on?

Answer 1

You want to use the citusDB column_to_column function described in the citus db docs: http://docs.citusdata.com/en/v9.3/develop/api_udf.html

SELECT column_to_column_name(logicalrelid, partkey) AS dist_col_name
  FROM pg_dist_partition
 WHERE logicalrelid='<table>'::regclass;