Re: Multiple indexes, huge table

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Aram Fingal <fingal(at)multifactorial(dot)com>
Cc: Postgres-General General <pgsql-general(at)postgresql(dot)org>
Subject: Re: Multiple indexes, huge table
Date: 2012-09-06 21:54:10
Message-ID: 13407.1346968450@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

Aram Fingal <fingal(at)multifactorial(dot)com> writes:
> I have a table which currently has about 500 million rows. For the most part, the situation is going to be that I will import a few hundred million more rows from text files once every few months but otherwise there won't be any insert, update or delete queries. I have created five indexes, some of them multi-column, which make a tremendous difference in performance for the statistical queries which I need to run frequently (seconds versus hours.) When adding data to the table, however, I have found that it is much faster to drop all the indexes, copy the data to the table and then create the indexes again (hours versus days.) So, my question is whether this is really the best way. Should I write a script which drops all the indexes, copies the data and then recreates the indexes or is there a better way to do this?

Yes, that's actually recommended practice for such cases.

> There are also rare cases where I might want to make a correction. For example, one of the columns is sample name which is a foreign key to a samples table defined with " ON UPDATE CASCADE." I decided to change a sample name in the samples table which should affect about 20 million rows out of the previously mentioned 500 million. That query has now been running for five days and isn't finished yet.

That sounds like you lack an index on the referencing column of the
foreign key constraint. Postgres doesn't require you to keep such
an index, but it's a really good idea if you ever update the referenced
column.

regards, tom lane

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message Andrew Barnham 2012-09-06 22:23:35 PostgreSQL server embedded in NAS firmware?
Previous Message Sergio Basurto 2012-09-06 21:43:48 Re: regexp_matches question SOLVED