There's always an exception. For example, legacy systems with strange
(but conformant) IP stacks - some VMS systems for example. I would
love to be able to do set per-host concurrency for such systems where
I know *exactly* what the upper limit on concurrent connections it
can handle is, cause it's defined in it's IP / MTA implementation.
There's no point in going above this figure cause it just ties up my
remote connections due to the way it handles connections above this
number - i.e. it becomes a black hole.
-tony
On 20-Oct-99 Russell Nelson wrote:
> And in neither of those cases is it an improvement to keep banging
> on
> the destination host. If it refuses messages, then by inference
> it's
> overloaded. Much better to wait until it isn't.
>
---
E-Mail: Tony Gale <[EMAIL PROTECTED]>
All new:
Parts not interchangeable with previous model.
The views expressed above are entirely those of the writer
and do not represent the views, policy or understanding of
any other person or official body.
