From: Bruce Momjian <pgman@candle.pha.pa.us>
Message-Id: <200507220513.j6M5DM209271@candle.pha.pa.us>
Subject: Re: ROW SHARE/SELECT ... FOR UPDATE + foreign keys
In-Reply-To: <7B27AD9D-67FC-47BE-9956-B4066E6B9C53@sitening.com>
To: "Thomas F. O'Connell" <tfo@sitening.com>
Date: Fri, 22 Jul 2005 01:13:22 -0400 (EDT)
Cc: pgsql-docs@postgresql.org
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset=US-ASCII

Thomas F. O'Connell wrote:
> What would people think of adding a note to the ROW SHARE section of  
> Table-Level Locks remarking on the fact that referential integrity  
> checks expressed through foreign keys do SELECT ... FOR UPDATES.
> 
> http://www.postgresql.org/docs/8.0/static/explicit- 
> locking.html#LOCKING-TABLES
> 
> It was somewhat challenging today to try to track down a pervasive  
> locking issue related to a table not directly referenced in a stored  
> procedure but that was referenced by a key in a table that was being  
> updated.
> 
> I'm not sure how this changes with the new shared row locking  
> implementation in 8.1...
> 

8.1 will use _shared_ row locks for such cases.

-- 
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 359-1001
  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073