Re: [Q]: Pre-spawned SQL*Net tasks prevent Win31/Novel Client connection

Dave Mausner wrote:
>
> In article <32507DC1.126D_at_cadvision.com>,
> Wayne Linton <lintonw_at_cadvision.com> wrote:
> >I run a Windows 3.1 client on a Novel LAN. I connect to an ORACLE7 V7.1
> >database via SQL*Net V2.1 via TCP/IP to an AlphaServer, OpenVMS.
> >
> >With a Listener that does NOT pre-spawn tasks, I can connect. As soon
> >as I specify any pre-spawn tasks in the LISTENER.ORA for that database,
> >I can not connect from my client. I get a timeout error on the client.
>
> sure because it takes N times longer for VMS to create all those pre-spawned
> processes, if you happen to be the first to open that connection. the answer
> may be to increase the timeout delay, or forget about pre-spawned processes.
>
> you might prefer to use MTS instead of pre-spawned servers.
>
> Dave Mausner, Managing Consultant, Braun Technology Group, Chicago.

When the listener starts with pre-spawn tasks = 1, a task is started up and sitting there waiting for someone to connect. So my problem is not that I timeout waiting for a task to be spawned on VMS - its already there. However the Port address of this pre-spawned task (as listed by the LSNRCTL SERVICES command) is something that puzzles me - like how does it ever get connected to anything? Is this the key to the problem?

Without the prespawned task, the listener connects me okay but it takes at least 30 seconds to connect. I even coded the IP address in the TNSNAMES.ORA and it makes no difference.

Other suggestions?

Wayne Linton Received on Tue Oct 01 1996 - 00:00:00 CEST