On 1 Dec 2005 at 15:29, Nelson R. Pardee <[log in to unmask]> wrote:

> On Thu, 1 Dec 2005, Francoise Becker wrote:
> 
> >On 1 Dec 2005 at 12:13, Nelson R. Pardee <[log in to unmask]> wrote:
> >
> >> But I'm still puzzled because, for example, I see failures
> >> for one address every day all month, but it isn't deleted. So the Delay(5)
> >> should catch the person regardless of the MAX(100).
> >
> >I'd have to see more details to diagnose that specific problem.
>
> Be glad to do it. What is needed?

Well, to start with, a recent Error Monitoring report that shows it's 
been recording bounces for over 5 days for that address and the 
results of a QUERY on the subscription for another. If you have 
changelogs enabled, a check of the changelog entries for that address 
over the last month may or may not be instructive.
 
> >However, if specifically the *probe* processing is causing
> >performance problems, you can change the Probe rate on the Auto-
> >Delete= keyword so that fewer addresses get probed each day.
> >
> >If your site isn't running HPO, you should consider upgrading to HPO
> >as well.
> 
> Our cpu utilization on the box averages less than 5%. This is by far our
> most active large list (those that are larger are once a day distribution
> lists). I wouldn't mind going to HPO, but I don't have a compelling case
> to take to management, at least yet.

You should start by reducing the probing rate anyway. 

Is the list set to Mail-merge=Yes, by any chance? That would increase 
the number of probes, because then every message would be a probe 
instead of just max one probe per day per person.

-- 
Francoise Becker <[log in to unmask]>

Knowledge is just a click away: http://www.lsoft.com/optin.html