Re: Two instances with same SID in a same machine

Mark Bole wrote:

> Ok, consider yourself proven wrong.

Lest there be any doubt, I do. And I've posted as such elsewhere. It happens.

> Fabrizio beat me to it, but I
> tested and received the same result. No RAC, no VMware, no tricks up my
> sleeve -- just a plain vanilla Solaris 8 box.
>
> I read the OP as an academic exercise, not a "is this a good idea in the
> real world" question.

So did I, actually.

> But then some people in this newsgroup don't even
> like the idea of running two DIFFERENTLY named instances on the same
> box, so it's not surprising the level of resistance to even debating the
> possibility.

I wasn't resistant, actually. I was 100% convinced it couldn't be done. It's what coming from a Windows background does for you (where it can't be done. Famous last words, I realise!).

> It's due to good design decisions made a long time ago in the area of
> Unix shared memory (System V IPC).

I believe you. I don't know what it means, but I believe you!

[snip much good technical stuff]

> I'm quite willing to believe that it won't work under Windows, but then
> fork() won't work under Windows either, last time I checked.
>
> Setting up a separate listeners (or separate service names in the same
> listener) for each instance is left as an exercise for the reader. ;-)
>
> --Mark Bole

Regards
HJR Received on Fri Aug 27 2004 - 18:48:26 CDT