Re: Access 2.0 as front-end

From: Jon Machtynger <jonm_at_nessie.be.ingres.com>
Date: 20 Oct 1994 13:35:43 GMT
Message-ID: <385rnf$268_at_zebedee.ingres.co.uk>


Peter Herdman-Grant (100435.153_at_CompuServe.COM) wrote:
: What you are intimating is the key consideration when developing
: ODBC/Access applications. In the absence of stored procedures to
: achieve host side filtering, Access will do its best to pull the
: minimum host data to the PC to achieve a heterogeneous join.
: There are many ways to overcome this and achieve great Apps,
: design your database to limit heterogeneous queries. If the
: major queries in your App involve heterogeneous joins,
: you've designed the data model wrong! Understanding that data
: resides in different places, be sure to re-locate the data before
: trying to run a performant application against it.
 

: Is Oracle any better than Ingres?

Pete long time no speak... Unfortunately, I think Oracle is in the same boat although I suppose there may be a different order in which the heterogenous joins would be performed. Would Oracle 7's rule based optimiser (cf cost based) change the performance based on a different storage place for joins? If you're using the Jet engine, then I don't think so.

I'd be interested to see any documented performance document using Oracle and work arounds required to make Jet work more efficiently. Don't let the ingres.com extension worry you, some of our customers use both and want them *both* to run fast...

Jon

--
#include <disclaimer.h>
// All the above is my OPINION...
+--------------------------------------------------------------------+
| Please 'REPLY' to jonm_at_ingres.com     Otherwise you will 'Bounce'  |
| Jon Machtynger(jonm_at_ingres.com)        +--------------+            |
| Bvd de la Woluwe 34 Bte. 13            | Aussie in    |            |
| Brussels. Belgium.                     |   Overijse.. |            |
| Ph: 02-774 49 23 Fax: 02-773 29 08     +--------------+            |
+--------------------------------------------------------------------+
Received on Thu Oct 20 1994 - 14:35:43 CET

Original text of this message