From: | Gregory Stark <stark(at)enterprisedb(dot)com> |
---|---|
To: | "Perry Smith" <pedz(at)easesoftware(dot)com> |
Cc: | <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: Concurrency Question |
Date: | 2007-07-17 05:00:59 |
Message-ID: | 87abtvei9g.fsf@oxford.xeocode.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
"Perry Smith" <pedz(at)easesoftware(dot)com> writes:
> Looking at the documentation above, I would expect the insert into A to get a
> Row exclusive lock for table A. And, I'm guessing it would get an ACCESS
> SHARE lock for table B. But this would not prevent the delete from B from
> happening at the same time (if I am reading this correctly).
The bit you quoted was for tables. The RI trigger does indeed take a share
lock on the referenced record in table B which prevents it from being deleted.
(In older versions it used to take an exclusive lock because there were no
share locks on records.)
--
Gregory Stark
EnterpriseDB http://www.enterprisedb.com
From | Date | Subject | |
---|---|---|---|
Next Message | Jason Nerothin | 2007-07-17 07:03:07 | interaction with postgres defined types in custom c functions |
Previous Message | Gregory Stark | 2007-07-17 04:56:37 | Re: why postgresql over other RDBMS |