I have a very old sendmail configuration on a VM on a dying server. I need to move it to a stable hw platform, so I took the sendmail setup and migrated it to a new system with a current Sendmail installation.

The previous Sendmail configuration worked fine and was trusted by remote senders. It did not have SPF, DKIM, or DMARC set up. The new Sendmail server worked fine (worked as expected), but was not trusted by some remote MTAs sending TO it - including those used by Google/gmail. This seems odd because the only real difference between the two configurations is the Sendmail version -- something seems to have changed in the software. I reverted to the original system after these experiments last summer, but haven't circled back to it until now, when the failure of the underlying hw is forcing my hand.

I haven't been following the evolution of Sendmail. I need to set up a better long-term solution (possibly with another MTA, and definitely with SPF+DKIM+DMARC), but in the short term I urgently need to migrate this configuration over to a box that isn't failing. Is anyone aware of Sendmail changes (either in behavior or defaults) that would explain how remote servers regard the MX?