Hello all-

I have inherited a hybrid environment with many websites and an onsite centralized (linux) mail server. we have moved our organizational mail to Office365 and I have added all my relevant domains into the Admin console of O365. My goal is to decommission the onsite mail server. I have some onsite LOB devices that have created inbound connectors for, and they work fine.

So then the question is:

If I have an onsite webserver (linux) that is the same domain name as a registered domain in O365 running sendmail or postfix with applications using that for mail, can I just point that at the EOP FQDN and be done with it, or do I have to additionally create an inbound connector to tell EOP it's ok to receive and send email from that server?

Example:

Onsite server hosts www.my-domain.com and runs postfix for mail. I have registered my-domain.com in the Admin console of O365 and made the necessary DNS changes, but there are no email addresses at my-domain.com, it only has applications that send OUTBOUND mail. (Course registrations, payment confirmations) Currently those application use localhost and postfix to send mail. I know how to point postfix at the FQDN of my-domain-com.mail.protection.outlook.com, but is that all that is necesaary, or do I have to also create an inbound connector with the public IP of my-domain.com and add my-domain.com to the allowed domains filter in that connector?

I hope this makes sense and that I am not in the wrong area. I did some research and read up on connectors but they refer to devices and apps, not specifically servers, so I am a little confused.

Thanks for any advice/assistance in advance!

-Sam

Hi Sam,

If the APP still use the use localhost and postfix to send mail, then additional connector is not necessary.

Thanks,