Re: tnsnames vs nameserver

It has to do with the way in which the SQL*Net software resolves a service name. Say that you connect to a database named "TEST". Well, what is TEST? The traditional method is for SQL*Net to read a file on your PC called TNSNAMES.ORA. This file will contain an entry for TEST that provides a node name and an instance name. When people refer to the tnsnames method, this is what they are talking about.

One problem with using tnsnames files is that each and every PC needs to have their own file. I've seen sites where they place this file on a LAN to be shared, but that can be difficult to set up, especially if you have a large user base.

Oracle's solution to this problem is Oracle Names. It's a nameserver that sits on the LAN and provides translations from database names like TEST.

You get a couple of advantages by using Oracle Names. One is that you can move instances transparently. If you move TEST from machine A to machine B, you only have to update the information in the names server database. With TNSNAMES files, you would have to update all the files, possibly disrupting some of your users. Another advantage of Oracle Names is that as you add databases, you aren't forced to go around to each user's machine and update their TNSNAMES.ORA file.

regards,

Jonathan

On Thu, 07 Jan 1999 15:43:16 -0500, Rob Williamson <robw_at_physics.umd.edu> wrote:

>We are about to have an Oracle 8 install done and we were told we needed
>to decide
>on tnsnames ( flat file ) or nameserver ( database ) for the Net8
>install.
>What does this mean? And how should we decide?
>Is it based on how many users we will have?
Received on Thu Jan 07 1999 - 19:13:11 CST