Re: pg advisory locks
От | Alban Hertroys |
---|---|
Тема | Re: pg advisory locks |
Дата | |
Msg-id | 8A4E4371-7B0C-4627-8EBE-D71241915F79@gmail.com обсуждение исходный текст |
Ответ на | pg advisory locks (Arun Gokule <arun.gokule@gmail.com>) |
Список | pgsql-general |
On 12 Oct 2014, at 18:41, Arun Gokule <arun.gokule@gmail.com> wrote: > Hi, > > I am executing pg_advisory_locks using the following set of statements: > > SELECT pg_advisory_lock(317,2); > UPDATE posts SET dislikers = array_remove(dislikers, 7) WHERE id = 317; > update posts set num_dislikes = icount(dislikers), updated_at = now() where id = 317; > WITH update_likers AS (SELECT pg_advisory_unlock(317,2)) select num_likes, num_dislikes, (7 IN (select(unnest(likers))))as liked, (7 IN (select(unnest(dislikers)))) as disliked from posts where id = 317 LIMIT 1; > > These are issued from a multithreaded app. One in 1000 queries, I get a deadlock after the execution of the above set ofstatements. i.e. SELECT try_pg_advisory_lock(317,2) returns false. Is there something obvious that I am doing wrong? Your last query doesn’t call the unlock statement, that WITH section is going to be ignored as the query isn’t using it. Alban Hertroys -- If you can't see the forest for the trees, cut the trees and you'll find there is no forest.
В списке pgsql-general по дате отправления: