GP Database Distribution key Query

The tables in the GP database generally set the distribution key, which, if not set, is usually the first field by default, so it is generally recommended that a single record be used as the distribution key, so that the records in the table are evenly distributed across each segment.

GP Database Distribution key Query method:

select c.schema_name,       c.table_name,        c.table_comment as  "Chinese show",        c.column_name  as  "Distribution Key",  from  ( select d.nspname as schema_name,                 c.relname as  Table_name,                obj _description (c.oid)  as table_comment,                 a.attname as column_name            from pg_class c            left join pg_attribute a              on a.Attrelid = c.oid           join pg_ Namespace d             on d.oid  = c.relnamespace           left join  pg_inherits f             on  C.oid = f.inhrelid          where a.attnum  > 0            and f.inhrelid  is null) c  join  (select a.attrnums[i.i] as a.attrnums,                b.attname,                a.localoid::regclass as  localoid          from gp_distribution_policy a,                 (Select generate_series (1,10)  as generate _series)  i (i),                Pg_attribute b        where a.attrnums[i.i] is not  null          and a.localoid =  b.attrelid          and a.attrnums[i.i] =  B.attnum)  t    on c.column_name = t.attname   and  ((c.schema_name::text | |   '. ':: text)  | |  a.table_name::text):: Regclass::oid = t.localoid::oid order by c.schema_name,           c.table_name,          c.column_name; 

GP Database Distribution key Query