From: Tom Lane <tgl@sss.pgh.pa.us>
To: Ioannis Anagnostopoulos <ioannis@anatec.com>
cc: pgsql-performance@postgresql.org
Subject: Re: Sequential scan instead of index scan
In-reply-to: <501FE1C9.6010209@anatec.com>
References: <501FDDD5.2000106@anatec.com> <501FDFCD.2030806@oneit.com.au>
	<501FE1C9.6010209@anatec.com>
Comments: In-reply-to Ioannis Anagnostopoulos <ioannis@anatec.com>
	message dated "Mon, 06 Aug 2012 16:24:57 +0100"
Date: Mon, 06 Aug 2012 11:34:08 -0400
Message-ID: <10966.1344267248@sss.pgh.pa.us>

Ioannis Anagnostopoulos <ioannis@anatec.com> writes:
>         I think this is a pretty good plan and quite quick given the
>         size of the table (88Million rows at present). However in real
>         life the parameter where I search for msg_id is not an array of
>         3 ids but of 300.000 or more. It is then that the query forgets
>         the plan and goes to sequential scan. Is there any way around?

If you've got that many, any(array[....]) is a bad choice.  I'd try
putting the IDs into a VALUES(...) list, or even a temporary table, and
then writing the query as a join.  It is a serious mistake to think that
a seqscan is evil when you're dealing with joining that many rows, btw.
What you should probably be looking for is a hash join plan.

			regards, tom lane