INSERT deadlocks (bug or feature?) - Mailing list pgsql-bugs

From evgeny gridasov
Subject INSERT deadlocks (bug or feature?)
Date
Msg-id 20050527140607.443a55af.eugrid@fpm.kubsu.ru
Whole thread Raw
Responses Re: INSERT deadlocks (bug or feature?)  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-bugs
Hi everybody!

Recently, I've discovered an interesting feature (or a bug?) of PostgreSql (checked 7.4.x and 8.0.x):

One may define such tables:

create table ref(
id serial primary key,
name text);

create table dat(
id serial,
ref_id int references ref(id),
comment text);

Let us fill the ref table:

insert into ref(name) values('feature');
insert into ref(name) values('bug');

The test case:
For example we have 2 concurrent transactions (tr1 and tr2):

tr1: begin;
tr2: begin;
tr1: insert into dat(ref_id, comment) values (1, 'all ok');
tr2: insert into dat(ref_id, comment) values (2, 'all ok');
tr1: insert into dat(ref_id, comment) values (2, 'lockup');
tr2: insert into dat(ref_id, comment) values (1, 'deadlock');


... and we recieve a deadlock!
Easy to understand why: each insert statement generates query like:
SELECT 1 FROM ONLY "public"."ref" x WHERE "id" = $1 FOR UPDATE OF x


So, is this behaviour of postgresql is a bug or feature?

Thanks,
Eugene.

pgsql-bugs by date:

Previous
From: Michael Fuhr
Date:
Subject: Re: BUG #1681: Selection problem
Next
From: Tomasz Brzezina
Date:
Subject: UNION makes strange duplicates