On Mon, Oct 06, 2003 at 12:27:09PM +0100, Tony Finch wrote: > >Yes, I've seen evidence of limits between the 10,000 and 20,000 messages
> >per hour range, in a large ISP environment. Apologies for being vague,
> >I'll get more data.
>
> That's only 5 messages per second. You must be doing something wrong.
> That might be about the limit of a single disk system without message
> logs turned off or other IO optimizations, but that's hardly a large
> ISP environment.
You'll see a problem, when you're using exim on a fallback in this
situation. The server is trying to connect to machines which have
recently been known to be unavailible. So the changes of writing to the
retry database are *high*. I really is pretty bad.
At the same time you've accepting new mail from your relays and tryign
to deliver that, which also causes rights.
Finally when you manage to flush the mail for a domain, that also causes
writes. Lots and lots of writes. Only one queue runner can write at a
time. Bad.
