Re: Theoretical Basis for SELECT FOR UPDATE
From: Tony Andrews <andrewst_at_onetel.com>
Date: 6 Oct 2005 07:58:18 -0700
Message-ID: <1128610698.627761.73780_at_g49g2000cwa.googlegroups.com>
Date: 6 Oct 2005 07:58:18 -0700
Message-ID: <1128610698.627761.73780_at_g49g2000cwa.googlegroups.com>
Tony Andrews wrote:
> Whether we lock the rows and block the second user, or don't block them
> and instead have our transaction fail, either we are ARE "forcing all
> the potentially concurrent transactions modifying a common set of rows
> execute serially", are we not?
That didn't make sense. What I meant to type was:
Whether we lock the rows and block the second user, or don't block them and instead have our transaction fail, either way we ARE "forcing all the potentially concurrent transactions modifying a common set of rows execute serially", are we not? Received on Thu Oct 06 2005 - 16:58:18 CEST