Re: Informix vs. Sybase vs. Oracle vs. (gasp) MS SQL Server

From: J. Duhl <jduhl_at_highway1.com>
Date: 1997/04/18
Message-ID: <3357F0AF.2A94_at_highway1.com>#1/1


Scott Gray wrote:
>
> In article <33538a45.29051694_at_news.jax.mindspring.com>,
> Reid Fleming <RFleming_at_aol.com> wrote:
> >Dave Raddatz <raddatz_at_austin.ibm.com> wrote:
> >
> >>Steve Phelan wrote:
> >>
> >>I think Larry's referring to the SP machines (mutltiple RS/6000
> >>node environment) vs. what you describe as SMP machines (multiple
> >>chips in a single box).
> >>
> >>I believe several DB vendors have written "parallel" versions
> >>of their products to take advantage of the SP type machines.
> >>
> >
> >If they have, I would be interested in seeing how they turned out. We
> >had one of the first SP2's that we tried using with Sybase System 10.
> >Performance was poor; Sybase recommended that we wait until their
> >'Navigation Server' became available. We waited for a while, but the
> >product, which they were co-developing with NCR (I believe) was
> >dropped.
>
> Nope, it wasn't dropped, and is available right now. Only
> the name was changed, it is currently called Sybase MPP 11.1.
> I managed to play with it a while ago, and it is quite cool
> (although, as with *any* clustering system, is pretty hard
> to maintain). Also, Sybase's future direction with the new
> Adaptive server architecture is to roll MPP into the Server
> itself.
>
> >I'm not an internals expert, but several DB vendors in the past have
> >spoken of how much more difficult it is to 'parallelize' between
> >machines, versus between processors (like Informix, and to a lesser
> >extent, Oracle, are able to do). I like the idea of separate boxes,
> >and hope that they can make (or have made) this work.
>
> It works, but is a really hard problem (just think of the variables
> involved with optimization). Now with the hybrid NUMA approach
> I bet you'll start to see a lot of development in the direction
> of tuning databases towards a more distriubuted approeach.
>
> Cool stuff.
Interesting to see where the relational database architectures are now relative to some of the object databases architectures.

There are several object databases that already scale for large distributed networks (e.g. Objectivity, Versant, Object Design). That is one of their advantages over an RDBMS. They also handle complex processing much better with higher performance and on less resources. Oh and they also support SQL (usually via an ODBC interface).

$0.02  

  • Joshua Duhl

Stillpoint Consulting
74443.2610_at_compuserve.com

"From here, I can see the movement of everything else" Received on Fri Apr 18 1997 - 00:00:00 CEST

Original text of this message