Forgive me for not quite understanding, but is the Global message limit for all virtual domains combined or is it using whatever I use as a global for a per domain limit?
Is this limit being applied to incoming and outgoing or just outgoing?
It’s been a while since I got snubbed on a question I posted… oh well.
It does appear that the global limit applies to all the virtual accounts combined. I just had an account get abused and the limit was reached, causing all domains to have mail rejected.
Wouldn’t it make more sense to have the global number be applied individually to each virtual account?
Hmm, that’s how it should work - the limit should be applied for each sender individually. Well actually, it gets applied to whatever the milter is instructed to use as its database key. Jamie should be able to answer that since his code configures the milter. 
Jamie, what is being used to identify each remote user being monitored by the milter? Their IP address?
It would appear that it looks at the “FROM”, or the sender email address. The config is in /etc/mail/greylist.conf. The config lines for global are:
ratelimit “virtualmin_limit” rcpt 30 / 1h
racl blacklist from /.*/ ratelimit “virtualmin_limit” msg “Message quota exceeded”
racl whitelist default