The drop in list traffic yesterday was my fault. The problem was on a system
unrelated to the classiccmp server, but it only affected the classiccmp
server. I believe it would have impacted just over a day of traffic. Problem
resolved, and my apologies.
Details for the tech oriented:
Our hosting service has two front end mail processors (mx1/mx2) that do a
lot of quick/basic checks before passing mail onto the backend mailbox
servers. One of them had an issue and had to be restored from a backup that
was a couple weeks old. We don't normally think anything of that in this
particular case, because the config files on a front end mx basically never
change so any backup is good. Much to my chagrin, I didn't realize that the
backup for mx2 was taken BEFORE switching the mail stream for
classiccmp.org
from the old ccmp server to the new one. Since both
front end mail servers
are equal preference, that means that about half the inbound
mail for
classiccmp.org was being directed to the old server which has pretty much
been sliced up as things were migrated to the new server. Last night I
changed mx2 to deliver to the correct/new classiccmp server (mx1 was already
doing so), and that fixed the problem. Newbie mistake, my apologies.
Best,
J