Torsten Förtsch

Parallel query without parallel query

One of the key improvements in Postgres 9.6 has been the ability to distribute a query to multiple worker processes. Yet, with a few tricks this has been feasible almost forever at least for the really heavy stuff. I prefer to keep the database specific stuff in the database

Lock contention in Postgres

At binary.com, our clients can trade in binary options. We keep all the related information in a PostgreSQL database. The life cycle of such an option begins when it is bought by a client. In our language this is called an open position. Later the client sells the position

PostgreSQL: How to use EXPLAIN output in SQL?

Sometimes you want to process the EXPLAIN output in SQL. Unfortunately, EXPLAIN itself does not offer that option. However, there are tricks... Wrapping EXPLAIN in a function This is probably the simpler version. It requires to create a function: CREATE OR REPLACE FUNCTION explain( p_q TEXT, p_mod TEXT[