Subject: | |
From: | |
Reply To: | |
Date: | Wed, 12 Apr 1995 16:42:53 EDT |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
On Wed, 12 Apr 1995 07:47:01 CDT R J Kennedy said:
>--We have been having a problem with Oracle's SQLNET program reading
> the RESLVCNF.NET.SYS file. When the file is there the SQLNET program
> is unable to connect from our Hp/9000 to the Hp/3000 and access Image.
Hmmm... really strange unless they're trying to reverse-map you, and that
should fail in either case if it fails with it present. The primary goal
of RESLVCNF is to enable DNS resolution through an outside server.
> domain mmm.com
> nameserver (IP address of our name server)
It might help if the 9000 is defined in HOSTS.NET.SYS. It sounds very
much like a security-based thing rather than an access problem since the
presence of RESLVCNF increases your connectivity rather than the converse.
> My question is: Is this file related in any way to the entries in
> the net directory file in the NMMGR program.
Well, no, other than perhaps changing it's order of preference in the
process of resolving a name. In 4.0, having RESLVCNF caused DNS name
resolution to be performed FIRST, then HOSTS/NSDIR (not sure about the
order there). I don't know if this was changed in 5.0.
> My thoughts are that it's the information in the file that is causing
> SQLNET the problems and not the existance of the file.
The data should be valid; but it might be upset if your Probe name is not
the same as your DNS name (wild guess there).
[\] Jeff Kell, [log in to unmask]
|
|
|