sathiya psql wrote:
> EXPLAIN ANALYZE SELECT count(*) from call_log_in_ram ;
> QUERY
> PLAN
> ------------------------------------
> ----------------------------------------------------------------------------------------------
> Aggregate (cost=90760.80..90760.80 rows=1 width=0) (actual
> time=6069.373..6069.374 rows=1 loops=1)
> -> Seq Scan on call_log_in_ram (cost=0.00..89121.24 rows=3279119
> width=0) (actual time=0.012..4322.345 rows=3279119 loops=1)
> Total runtime: 6069.553 ms
> (3 rows)
You will never get good performance automatically with COUNT(*) in
PostgreSQL. You can either create your own infrastructure (triggers,
statistics tables, etc) or use an approximate result like this:
CREATE OR REPLACE FUNCTION fcount(varchar) RETURNS bigint AS $$
SELECT reltuples::bigint FROM pg_class WHERE relname=$1;
$$ LANGUAGE 'sql';
Use the above function as:
SELECT fcount('table_name');
fcount
--------
7412
(1 row)
--
Sent via pgsql-performance mailing list (pgsql-performance@xxxxxxxxxxxxxx)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-performance
