Are subselects treated as atomic single commands?

From: Thomas Chille <thomas(dot)chille(at)gmail(dot)com>
To: pgsql-general(at)postgresql(dot)org
Subject: Are subselects treated as atomic single commands?
Date: 2004-11-24 15:12:42
Message-ID: cad2de1c041124071225c92cf2@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

Hi,

the docu about the Read Committed Transaction Isolation Level pointed
out: "... The point at issue here is whether or not within a single
command we see an absolutely consistent view of the database.".

Because i dont want to use the Serializable Transaction Isolation
Level or table locks if it not necessary i have one question:

Would the insert command with that subselect treated as one single
command and can i so prevent a race condition between multiple
function calls?

CREATE OR REPLACE FUNCTION "public"."count_parameter" (name, integer)
RETURNS "pg_catalog"."void" AS'
BEGIN

INSERT INTO parameter (parameter_name, parameter_value) SELECT $1,
$2 WHERE (
SELECT COUNT(*) = 0 FROM parameter WHERE parameter_name = $1
);
IF NOT FOUND THEN
UPDATE parameter SET parameter_value = parameter_value + $2
WHERE parameter_name = $1;
END IF;

RETURN;

END;
'LANGUAGE 'plpgsql' VOLATILE CALLED ON NULL INPUT SECURITY INVOKER;

best regards,
thomas

Browse pgsql-general by date

  From Date Subject
Next Message Peter Eisentraut 2004-11-24 15:21:08 Re: Moving/Using Postgres Binaries on multiple machines
Previous Message Richard Huxton 2004-11-24 15:08:35 Re: Performance