We have, for a long time now, successfully used MailScanner on our
gateway servers and then routing clean mail to our pop toasters.
The pop toasters are running netqmail, vpopmail 5-4-17, and chkuser 2.0.8.
I am having an issue where users are going overquota and not doing
anything about it until mail begins to bounce. What limits cause chkuser
to switch from "rcpt mailbox is overquota" to "not existing recipient"?
I am unable to find the answer, though it might be staring me in the face.
Right now if a users don't fix their quota issue in time I have to
manually restart milter-ahead on my MailScanner servers to clear the
cache of non-existant users.
Trying to understand the problem before I decide what to do about it.
DOH! chkuser is behaving correctly. Seems milter-ahead caches only "yes
accept the message" or "no refuse the message". Nothing else. Once an
overquota response is cached, milter-ahead responds with "no user".
I reduced my milter cache to one minute.
I'll go get my coffee now.
"Posterity, you will know how much it cost the present generation to
preserve your freedom. I hope you will make good use of it. If you
do not, I shall repent in heaven that ever I took half the pains to
preserve it." John Quincy Adams