Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: Avoiding any locks in SQL Servers - read and understand....itsmagic.

Re: Avoiding any locks in SQL Servers - read and understand....itsmagic.

From: Guido Stepken <g.stepken_at_t-online.de>
Date: Mon, 01 Sep 2003 18:37:38 +0200
Message-ID: <3F5375D2.3060200@t-online.de>


Hi, Joel !

Tnx, was my fault .. i really meant same data set, not database. I still   had this in mind, when writing that nonsense. But - it is still true, that, according to oracle's documentation (simply register for free at oracle), that oracle has no complete MVCC implementation, but one, that comes very near, MVRC (Multi Versioning Read Concurrency). No two transactions can write into the same row at the same time, but one can read and one can write. MVCC in PostgreSQL can have several instances writing or reading ...a complete implementation.
To avoid deadlocks, transactions are simply dropped. So, any writing transaction into postgresql should be read again.

MS SQL Server 2000 is not able to get that managed and causes very much programming overhead....costy.

regards, Guido Stepken

Joel Garry wrote:

> Well, were they all updating the same row? Mea culpa if I gave Guido
> too much credit for possibly mistyping database when he meant row. I
> was under the impression he was using MVCC in the sense of "instead of
> locking at that time" as opposed to the Oracle norm of optimistic
> commits with the ability to construct a version if needed for rollback
> or consistency.
Received on Mon Sep 01 2003 - 11:37:38 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US