Here is the scenario:
We are running one Exchange 2003 server in a small network; server has been running without any issues for years. Server is not blacklisted. Internal email and incoming email are working fine.
The MX record is not the same as the PTR record because the server is collocated and there are 2 outside addressing schemes, one for the LAN, and one for the ISP that controls the pipe to the colo.
The PTR record resolves correctly; but suddenly outgoing mail stopped working. The PTR record was established to make SMTP work. I have an explanation for this that I could post.
I noticed after the outgoing emails stopped, that the machine that holds to PTR record is not reachable by tracert or ping.
The ISP, controller of the PTR record server, claims that the machine does not need to be reachable, as long as the PTR record is resolving on the internet. (really??) MXtoolbox shows the PTR resolution is good and the MX record is good, but the SMTP test there fails.
What I did:
I added 184.108.40.206 to the Exchange servers sender ID and filtering section as external DNS, and most mail now goes out. (this tells me that the path usually used for SMTP was dead i.e.. (no tracert to the server whose address I was using before this started).
I looked for answers all over forums but found nothing definitive.
However, some emails are still failing to go out unless I add an SMTP connector for each failed domain recipient. So far I have added a dozen. This is getting old. Many of the required additional SMTP resolve to outlook.com.
Question: what is the real cause of the failure message #5.5.0 smtp;550 Requested action not taken: mailbox unavailable> Why does adding an SMTP record work, is there a better way, a final fix?