list of schema with sizes (relative and absolute) in a PostgreSQL database

Question

I'm looking for a query that returns a result of the form for any database (see example below supposing total space used by the database is 40GB)

schema | size | relative size
----------+-------------------
foo    | 15GB |   37.5%      
bar    | 20GB |     50%
baz    |  5GB |   12.5%

I've managed to concoct a list of space using entities in the database sorted by schema, which has been useful, but getting a summary per schema from this doesn't look so easy. See below.

SELECT relkind,
       relname,
       pg_catalog.pg_namespace.nspname,
       pg_size_pretty(pg_relation_size(pg_catalog.pg_class.oid))
FROM   pg_catalog.pg_class
       INNER JOIN pg_catalog.pg_namespace
         ON relnamespace = pg_catalog.pg_namespace.oid
ORDER  BY pg_catalog.pg_namespace.nspname,
          pg_relation_size(pg_catalog.pg_class.oid) DESC;

This gives results like

  relkind |                relname                |      nspname       | pg_size_pretty 
---------+---------------------------------------+--------------------+----------------
  r       | geno                                  | btsnp              | 11 GB
  i       | geno_pkey                             | btsnp              | 5838 MB
  r       | anno                                  | btsnp              | 63 MB
  i       | anno_fid_key                          | btsnp              | 28 MB
  i       | ix_btsnp_anno_rsid                    | btsnp              | 28 MB
  [...]
  r       | anno                                  | btsnp_shard        | 63 MB
  r       | geno4681                              | btsnp_shard        | 38 MB
  r       | geno4595                              | btsnp_shard        | 38 MB
  r       | geno4771                              | btsnp_shard        | 38 MB
  r       | geno4775                              | btsnp_shard        | 38 MB

It looks like using an aggregation operator like SUM may be necessary, no success with that thus far.

Answer 1

Try this:

SELECT schema_name, 
       sum(table_size),
       (sum(table_size) / database_size) * 100
FROM (
  SELECT pg_catalog.pg_namespace.nspname as schema_name,
         pg_relation_size(pg_catalog.pg_class.oid) as table_size,
         sum(pg_relation_size(pg_catalog.pg_class.oid)) over () as database_size
  FROM   pg_catalog.pg_class
     JOIN pg_catalog.pg_namespace ON relnamespace = pg_catalog.pg_namespace.oid
) t
GROUP BY schema_name, database_size

Edit: just noticed the workaround with summing up all tables to get the database size is not necessary:

SELECT schema_name, 
       pg_size_pretty(sum(table_size)::bigint),
       (sum(table_size) / pg_database_size(current_database())) * 100
FROM (
  SELECT pg_catalog.pg_namespace.nspname as schema_name,
         pg_relation_size(pg_catalog.pg_class.oid) as table_size
  FROM   pg_catalog.pg_class
     JOIN pg_catalog.pg_namespace ON relnamespace = pg_catalog.pg_namespace.oid
) t
GROUP BY schema_name
ORDER BY schema_name