I run a couple of email relay hosts at work. A number of customers use ProofPoint as email washers, so their MX record in DNS is something like mxa-002d7b01.gslb.pphosted.com which washes the mail and forwards good ones onward.
Recently, proofpoint blacklisted the IP address of one of my relays. Sendmail logs showed this.
Jul 31 11:15:27 smtp1 sendmail[600]: x6VZBFRHN0:
to=<[email protected]>,
relay=mxa-002d7b01.gslb.pphosted.com. [148.163.155.188],
dsn=5.0.0, stat=Service unavailable
The sender address is a [email protected] and is configured as a DISCARD in /etc/mail/access so we ended up dumping a lot of customer emails.
QUESTION: is it possible to configure sendmail to treat 500 "fatal" errors as 400 "transient" errors, and requeue the emails rather than bouncing them back to the sender ?
Details:
Queues will grow, that's totally acceptable. I don't have a way to regenerate those lost emails.
We're sending machine-generated emails, so they all look similar but not the same. Messages are not spam. Our end users pay for a larger service and these alerts are part of that service.
I don't know why proofpoint blocked that IP address, seems they don't know either. Yes I ended up routing email via another relay, and proofpoint accepted those messages, but that's a workaround, not a solution and could go wrong again.
0 Answers