pg.ddx.io  pgsql-sql@postgresql.org mailing list archive  
help / color / mirror / Atom feed
From: wieck@debis.com
To: genov@digsys.bg
Cc: pgsql-sql@postgreSQL.org
Cc: pgsql-novice@postgreSQL.org
Subject: Re: [SQL] Optimizing a query through its syntax and indices
Date: Wed, 3 Nov 1999 18:16:33 +0100 (MET)
Message-ID: <m11j41R-0003kLC@orion.SAPserv.Hamburg.dsh.de> (raw)
In-Reply-To: <199911031545.RAA13904@lorna.digsys.bg>

>
>
> Hello,
>
> I talk PostgreSQL 6.5.2 ...
>
> Suppose we have some tables we join using some of their fields.
> Suppose there are "restrictions" for the values of some (or all) of the tables'
> fields of this kind:
>    tableM.fieldN [ NOT ] IN (constA,constB,constC),
>    tableP.fieldQ [ NOT ] IN (constD,constE,constF),
>    etc...
> These restrictions can occur for the fields by which we join the tables, as well
> as for fields of tables we do not use for joins.
>
> Are there (can there be) general ( or not so general :-) ) rules for optimizing
> the query ( and the performance and the speed ) through the particular order we
> put the join statements in the WHERE clause, trough mixing/ordering the join
> parts and the restrictions ( in the means defined above ) in the values of the
> fields in the WHERE clause, through changing the syntax ( for example, using
> UNION or EXCEPT clauses ), or through using one- or more-than-one-field
> indices, or through doubling some of the restrictions if they refer to a
> field we join by ( for example, "... WHERE tableM.fieldN = tableP.fieldQ and
> tableM.fieldN in (constA,constB)" complemented by " and tableP.fieldQ in (constA
> ,constB)" ).
>

    The  last part, complementing the qualifications, is known to
    speedup  the  join  significantly.  At   least   for   simple
    expressions  that  can be put down into the scan itself. This
    is because it reduces the amount  of  data  before  the  join
    already.  Don't  know if this is true for IN expressions too,
    so you might give it a try (and report the result back to us,
    please).

    There  had  been  the idea to do this automatically in a step
    between rewriting and planning. Unfortunately noone seems  to
    have the time to tackle it.


Jan

--

#======================================================================#
# It's easier to get forgiveness for being wrong than for being right. #
# Let's break this rule - forgive me.                                  #
#========================================= wieck@debis.com (Jan Wieck) #





view thread (3+ messages)  latest in thread

Message-ID: <m11j41R-0003kLC@orion.SAPserv.Hamburg.dsh.de>
Permalink:  ../m11j41R-0003kLC@orion.SAPserv.Hamburg.dsh.de/
Also on:    postgresql.org/message-id/m11j41R-0003kLC@orion.SAPserv.Hamburg.dsh.de

 · 

reply

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Reply to all the recipients from your mail client; the link fills in
  the headers below: reply via email
  Some clients, webmail especially, drop In-Reply-To and References from
  a link, and the reply then starts a new thread: use the mbox below.

  To: pgsql-sql@postgresql.org
  Cc: wieck@debis.com, genov@digsys.bg, pgsql-novice@postgreSQL.org
  Subject: Re: [SQL] Optimizing a query through its syntax and indices
  In-Reply-To: <m11j41R-0003kLC@orion.SAPserv.Hamburg.dsh.de>
  References: <199911031545.RAA13904@lorna.digsys.bg> <m11j41R-0003kLC@orion.SAPserv.Hamburg.dsh.de>

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox