1000 FAQs, 500 tutorials and explanatory videos. Here, there are only solutions!
SMTPban or quarantining of outgoing messages in the event of mass mailing being sent out
Sometimes, a large number of messages are sent from a given e-mail address.
This can be INTENTIONAL or UNINTENTIONAL.
The risks incurred by e-mail server administrators such as us are enormous: if somebody lodges a complaint after receiving an e-mail that they never asked to receive (see below), then we are responsible for spamming and our messaging servers are immediately blocked by other providers. None of our customers can then send any more e-mails: a disastrous situation that we absolutely wish to avoid.
To avoid this, bulk e-mails are automatically intercepted and placed under quarantine.
They are sent once we receive assurance that there are no grounds for any potential complaint.
They are destroyed if they were unintentional or if the sender cannot prove that their address list complies with the rules.
Messages placed under quarantine cannot be "viewed" but the headers of one message among others are generally provided during exchanges
As set out in our general terms and conditions, any bulk mail must be sent only to persons who have explicitly requested it, via a sign-up process that must be logged (date, e-mail, IP address) by the sender.
People are becoming increasingly inconvenienced by spam and hence are taking a harder line (instead of simply binning such messages, they actually complain). At our end, as e-mail providers, we are responsible for any messages sent over our messaging system, or— and most crucially of all— for messages sent from outside for customers who have web sites hosted with us.
We have a newsletter service in our admin console which allows a sign-up log to be kept.
Any unintentional mailings are also your responsibility.
For a few years, spammers have unfortunately been working on looking for flaws in web sites that they find on Google, in order to allow them to send spam unbeknown to the owners of the sites in question, or via legitimate e-mail addresses whose passwords they have been able to guess or steal (trojans).
As a result, it is no longer enough to filter just incoming messages; messages leaving our network must also be filtered. In our specific case, we have developed our own tools to monitor the outgoing message stream, which run messages through a spam filter. When a particular format of message reaches a certain score, if all such messages are coming from the same sender and the rate of messages per second reaches a particular threshold (which is as generous as possible) that we have defined, then as a preventive measure the server blocks the e-mail address. (It does not destroy any messages that are already pending but simply suspends them until the customer can be contacted and we can ascertain whether it is actually an instance of spamming or simply a false alarm).
Preventing spam at the early stages enables us to curb bulk e-mailing until the security flaw in question is tracked down, thus preventing us from sending out hundreds of thousands of messages and generally allowing us to halt the flow within the first hundred messages!