Re: database in firewall

I have to admit I now believe that FM is correct. Oracle Doc implies a redirection occurs, but what probably actually occurs is a Unix fork (clone child and then orphan to create the dedicated server).
Other evidence including netsat and listener logs verifies FM to be correct.

The fork C code is not available for windows which is probably why all NT type instances are MTS, which no one disputes to involve redirection.

Also check out the listener.ora parameter called queues, it would not be needed if redirection always occured.

"FM" <fabrizio.magni_at_mycontinent.com> wrote in message news:41527BBB.1050408_at_mycontinent.com...
> Gerry Sinkiewicz wrote:
> > See Oracle metalink note: 125021.1
> >
> > There are differences among OS's so perhaps everyone is correct?
> >
> > It also provides a solution to the original problem (CMAN), which can be
> > used like an oracle only firewall (cf. oracle classes I took a few years
> > ago).
> >
> >
>
> Quoting the document you pointed out:
>
> "Oracle Multi-Threaded Server (MTS) on Unix platforms, (without
> specifying the address with the ports in the INIT<SID>.ORA file), Oracle
> Secure Sockets Layer (SSL) and Windows NT/2000 platforms will cause port
> redirection. "
>
> Ergo: no port redirection is made on unix system in dedicated mode.
> I tested on my solaris and neither there the system is doing port
> redirection.
> All TCP datagrams are sent on the listener port (you stated by yourself
> that the firewall guys had to open only the 1521).
>
> However I'm ready to retract if consistent and replicable experiment
> show different results.
>
> Could you please post the step you performed to cause port redirection
> on you solaris?
>
> (I'm even interested in the documents you mentioned in your previous
> post. Make you send me references? Thank you).
>
> --
> Fabrizio Magni
>
> fabrizio.magni_at_mycontinent.com
>
> replace mycontinent with europe
Received on Thu Sep 23 2004 - 16:33:06 CDT