Thanks Sam. You might be the only one who knows enough about the internals for these...

If I do the above, my problem would probably be a fiar bit of inefficiency - the servers have many IP addresses, and if I have to list them all in the generic config instead of a listen_all, that results in a file handle per connection and some overhead in other areas right? That wouldn't be the best - if I have to do this, I guess I could use a different port, and configure webmails to you that port though - that way I can keep both daemons on listen_all, and avoid the difficulties that would otherwise result from IP adding and deleting.

BUT...

Is the limit in effect for localhost? (Just checking if that's a special case... I found an OLD note I made that at one point I thought it was - not sure if I was right then or if it still applies though...)

Is there an error message generated when this condition occurs - something I could find in the logs? At this point I can't even be sure that's my problem. I only get told about the problem about once a week - used to be once a month though...

Same question I guess applies to other limit conditions... like MAXDAEMONS, MAXPERIP etc. Is there a way to detect if the limits are being tested? I don't want to set them unrealistically high, but I also don't want a limit of 40 instead of 50 or whatever to cause 1 or 2 client complaints a day ;-)

Thanks again.

m/