I'm using Mercury/32, v4.90.298 with SMTP Client (end-to-end version).

In the past three days, I've seen six emails being sent to known valid addresses that fail with:
Permanent error 250 (no data or no address - name server conf) resolving 'validdomain.com'

Checking the MX records for the domains finds all of the failures are pointing to something like validdomain-com.mail.protection.outlook.com

In each case, after the failure, checking the DNS records shows no DNS problems, and resending the message succeeds.

I searched logs back a few months. It happened twice in December. None in January. Four times in February. Three times yesterday, and once so far today. (Edit: twice today now.)

They are apparently transient errors, but the error message indicates permanent failure.

Anybody else seeing anything like this?

Any thoughts?

Regards,
Tony

Could be some problem with the name servers that your system is using. It's usually best to use the name servers in Windows for address resolution rather than specifying one in Mercury.

If a name server needs to be replaced there are several open alternatives around that usually are reliable, like Google (8.8.8.8 and 8.8.4.4) and Cloudflare (1.1.1.1 and 1.0.0.1).

Rolf,

Thanks for your input. Name servers was my first thought too. But since it has only been observed on email addresses hosted by Microsoft (outlook.com), I feel like there's something more to it.

I use SpamAssassin for spam filtering and it doesn't work with public DNS servers. So, I have been running a DNS server here for years. Up until last week, I was using Simple DNS Plus. Thinking that the errors might be name server issues, I changed to Technitium DNS Server (after a lot of research and testing). But, still seeing the same thing.

I've upped the logging level on Mercury and Technitium since the most recent failure. Waiting for the next failure to see if it reveals any hints.

Regards,
Tony

In case anybody is interested...

After getting more Permanent error 250 failures with the verbose logging on the DNS server, and with lots of googling for similar situations, I think I understand what is happening. It is a name server problem, but the real problem is on the remote domain's name server, Microsoft's in this case.

Microsoft's name servers are not responding within the timeout period of the local DNS server, which is the root cause.

MercuryE requests the MX record from the local DNS server. It succeeds. Then the A record query times out. MercuryE requests the A record again right away. If it still times out, the message is failed with "Permanent error 250 (no data or no address - name server conf) resolving 'validdomain.com'"

It's really a transient error, but it is being handled as a fatal one. The MercuryE DNS retries setting does not seem to apply to this as it only tries twice regardless of the retries setting.

I'm guessing it's mostly happening with Microsoft's mail servers because so many people are now hosting email on outlook.com (whether they know it or not.) I found a couple of examples that were not Microsoft, so it's not only...