Step 2: Create Monitoring User

We recommend creating a separate monitoring user on your PostgreSQL database for pganalyze.

As an RDS superuser, connect to the database you will be monitoring and run the following (we've generated a random password for you, but you can replace it with one of your choosing):

CREATE USER pganalyze WITH PASSWORD 'mypassword' CONNECTION LIMIT 5;
GRANT pg_monitor TO pganalyze;

CREATE SCHEMA pganalyze;
GRANT USAGE ON SCHEMA pganalyze TO pganalyze;
GRANT USAGE ON SCHEMA public TO pganalyze;

If you enable the optional reset mode (usually not required), you will also need this helper method:

CREATE OR REPLACE FUNCTION pganalyze.reset_stat_statements() RETURNS SETOF void AS
$$
  /* pganalyze-collector */ SELECT * FROM public.pg_stat_statements_reset();
$$ LANGUAGE sql VOLATILE SECURITY DEFINER;

Then, connect to each database that you plan to monitor on this server as a superuser (or equivalent) and run the following to enable the collection of additional column statistics:

CREATE SCHEMA IF NOT EXISTS pganalyze;
GRANT USAGE ON SCHEMA pganalyze TO pganalyze;
CREATE OR REPLACE FUNCTION pganalyze.get_column_stats() RETURNS SETOF pg_stats AS
$$
  /* pganalyze-collector */ SELECT schemaname, tablename, attname, inherited, null_frac, avg_width,
    n_distinct, NULL::anyarray, most_common_freqs, NULL::anyarray, correlation, NULL::anyarray,
    most_common_elem_freqs, elem_count_histogram
  FROM pg_catalog.pg_stats;
$$ LANGUAGE sql VOLATILE SECURITY DEFINER;

Note: We never collect actual table data through this method (see the NULL values in the function), but we do collect statistics about the distribution of values in your tables. You can skip creating the get_column_stats helper function if the database contains highly sensitive information and statistics about it should not be collected. This will impact the accuracy of Index Advisor recommendations.

Then, connect to each database that you plan to monitor on this server as a superuser (or equivalent) and run the following to enable the collection of additional extended statistics:

CREATE SCHEMA IF NOT EXISTS pganalyze;
GRANT USAGE ON SCHEMA pganalyze TO pganalyze;
CREATE OR REPLACE FUNCTION pganalyze.get_relation_stats_ext() RETURNS TABLE(
  statistics_schemaname text, statistics_name text,
  inherited boolean, n_distinct pg_ndistinct, dependencies pg_dependencies,
  most_common_val_nulls boolean[], most_common_freqs float8[], most_common_base_freqs float8[]
) AS
$$
  /* pganalyze-collector */ SELECT statistics_schemaname::text, statistics_name::text,
  (row_to_json(se.*)::jsonb ->> 'inherited')::boolean AS inherited, n_distinct, dependencies,
  most_common_val_nulls, most_common_freqs, most_common_base_freqs
  FROM pg_catalog.pg_stats_ext se;
$$ LANGUAGE sql VOLATILE SECURITY DEFINER;

Note: We never collect actual table data through this method (we omit fetching fields like most_common_vals from the pg_stats_ext view, as you can see in the function definition), but we do collect statistics about the distribution of values in your tables. You can skip creating the get_relation_stats_ext helper function if the database contains highly sensitive information and statistics about it should not be collected. This will impact the accuracy of Index Advisor recommendations.

Note that it is important you run these as RDS superuser in order to pass down the full access to statistics tables.

Write down the username and password of the monitoring user, we will use it in the last step of this tutorial.

Next we continue by installing the pganalyze collector software to your environment: