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

Home -> Community -> Usenet -> c.d.o.tools -> From Unix client to Unix Oracle server?

From Unix client to Unix Oracle server?

From: Ramon F Herrera <ramon_at_conexus.net>
Date: 3 Aug 2001 05:26:14 -0700
Message-ID: <c9bc36ff.0108030426.40ea56aa@posting.google.com>

Hello:

I am working on an application which will run on several Unix hosts
(as clients). The central server is another Unix machine running Oracle.

The server is the easy part: it runs Oracle 8.1.7 and RedHat Linux 7.0 The client part is giving me a lot of grief, because the host is an SCO OpenServer 5.0.5. Oracle has dropped support for SCO and when I called them, they suggested that I use ODBC from a 3rd. party vendor. Following Oracle's advise, I have been looking into EasySoft, OpenLink and InterSolv/Merant.

What bothers me about these companies' ODBC products (BTW: Merant doesn't support SCO anymore, so I am left with only 2) is the many layers that they have. They talk about "agents", they talk about "brokers", they talk about "ODBC to ODBC bridges", etc. All that stuff sound like a lot of overhead and my application needs to be as fast as possible. Being a Unix guy, I can't stop thinking: "why don't they have a single ODBC layer, which I install on the SCO client, giving me the ability to write ODBC apps, and connecting to the server (which should remain _untouched_; no extra software, after all it already has Oracle's Net8, and a listener on port 1521) using Oracle protocols?" I don't need all those extra "brokers" and "agents", just a good ODBC library (located in the client only, I repeat).

EasySoft has 3 pieces of software, one goes in the client (which I already tested, it comes with a small demo that connects to their MS Access NT in England, I just hope that the performance over the USA Internet is better that what I experienced). In addition, I am supposed to install 2 more layers in my Linux Oracle server, making all this even slower. It seems that what they do is run sqlplus internally, bypassing the obvious faster connection (port 1521).

So my first question is: does anyone out there know of a company that does ODBC in a simpler way, running in the client only and that has support for the SCO OS5?

The other question that comes to mind is: how does Oracle do the above
(Unix-to-Unix) connection? -assuming that both client and server run
some supported Unix like Solaris or Linux. I have checked and used OCI, but that won't help me now because OCI does not go over a network: the app must be in the same host as the Oracle server. Is this correct?

What tools does Oracle Corp. offer that will allow me to write a C application which will query a remote Oracle server?

Thanks for your insight and suggestions.

-Ramon F. Herrera Received on Fri Aug 03 2001 - 07:26:14 CDT

Original text of this message

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