postgresql - how and why indexes are bigger than their tables

Question

I'm using postgresql 9.3 and trying to understand how and why indexes are bigger than their tables.

Sample output:

 database_name | database_size |                          table_name                          | table_size | indexes_size | total_size
---------------+---------------+--------------------------------------------------------------+------------+--------------+------------
 foo_12345 | 412 MB        | "foobar_dev_12345"."fact_mobile_sends"                       | 57 MB      | 131 MB       | 189 MB
 foo_12345 | 412 MB        | "foobar_dev_12345"."fact_mobile_started"                      | 17 MB      | 39 MB        | 56 MB
 foo_12345 | 412 MB        | "foobar_dev_12345"."fact_mobile_stopped"                      | 16 MB      | 35 MB        | 51 MB

I'm running the following query to get the table and index sizes.

SELECT
    table_catalog AS database_name,
    pg_size_pretty(pg_database_size(current_database())) As database_size,
    table_name,
    pg_size_pretty(table_size) AS table_size,
    pg_size_pretty(indexes_size) AS indexes_size,
    pg_size_pretty(total_size) AS total_size
FROM (
    SELECT
        table_catalog,
        pg_database_size(current_database()) AS database_size,
        table_name,
        pg_table_size(table_name) AS table_size,
        pg_indexes_size(table_name) AS indexes_size,
        pg_total_relation_size(table_name) AS total_size
    FROM (
        SELECT ('"' || table_schema || '"."' || table_name || '"') AS table_name, table_catalog
        FROM information_schema.tables
    ) AS all_tables
    ORDER BY total_size DESC
) AS pretty_sizes;

Is my query correct? What would cause indexes to be bigger?

Answer 1

Possible reasons:

• Numerous and probably overlapping indexes on the table; have a look with \d

• Bloat due to high update churn can sometimes affect indexes more than tables, depending on update patterns. Examine the size of each individual index to see if it makes sense.

• GiST indexes, if used, can be quite large

Unlike what I originally thought this is not an issue with TOAST out-of-line storage not being counted, since pg_table_size includes TOAST tables.

Note that if you're concerned about index bloat and decide to REINDEX some of all of the involved indexes, consider setting a non-default FILLFACTOR first if the table is subject to lots of updates (or inserts+deletes). Otherwise you'll take a write performance hit because the index doesn't have any space to insert new values so it'll force lots of page splits and be less efficiently structured.