Support for bypassing RBL checks

[netlink-ie]

I think it would be a great idea to implement a way for an admin, and possibly resellers to be able to disable Exim RBL checks that reject incoming mail.

This would allow us to skip RBL checks on a per user or per domain basis. I've come across this problem when one of my business customers complained that they are not receiving some legitimate mails from their customers. After looking into this, I noticed that emails were being rejected because the [EDIT]sender's IP address, not recipient's[/EDIT] (the IP assigned to them by their ISP - in this case Orange.fr) was blacklisted on spamcop.net.

So without having to completely disable RBL lookups, with this feature I could simply disable the lookups only for one user.

[cPanelJamyn]

That sounds like an interesting suggestion, but it would be difficult to implement without more complexity. This would require moving the checks from the beginning of the SMTP transaction to a point after the RCPT TO stage, negating some of the performance benefits associated with early checks. You'd lose the ability to reject mail outright with a score of 'x' if there are multiple recipients, and some of the recipients choose to skip/whitelist the RBL, as well.

[netlink-ie]

Hi!

To get this working for me, I added the following lines to the Exim configuration:

domainlist skip_rbl_domains = example.com

#[% ACL_RBL_BLOCK %]
deny message= Your e-mail was rejected because $sender_host_address has been blacklisted at $dnslist_domain\n$dnslist_text
dnslists = zen.spamhaus.org:bl.spamcop.net
domains = ! +skip_rbl_domains

I did hear this method might give some problems with multiple recipients.

But I can see how the RBL checks could be a problem for some people too though if legitimate senders are unfortunate enough to have their ISPs assign them IPs that have been blacklisted. I guess this happens when the sender is using their ISP's SMTP servers, even if they have their mail hosted elsewhere. And you can't really ask your customers to tell their customers that the problem lies on their end.

What would be the alternative to my suggestion? Or is there even an alternative?