Score:0

On Premise antispam and MS 365 Hybrid deployment

in flag

I'm configuring a MS 365 Hybrid deployment, MX will be keeped to Exchange on premise server.

I have a On premise antispam (LibraEsva) that I don't want to remove.

This is the actual configuration:

enter image description here

Following MS guide lines, the connection between Exchange On Premise and Cloud has to be direct without any filter.

As you can see the there is a SMTP 25 from public internet to MX interface filtered by Antispam. Moreover, there is another SMTP 25 connection between Exchange On Premise and Cloude for internal routing between local and cloud mailboxed.

I made 2 NAT

  • From Any with port SMTP-25 forwarded to Antispam
  • From MS365 ips with port SMTP-25 forwarded to Exchange

This works, public email are filtered, and email between Exchange server are not filtered...

BUT

The NAT made from MS365 hostname are getting all public emails from MS 365 NOT ONLY the internal ones, if any user of 365 sends an email to me, it is not filtered anymore.

How can I solve this brain teaser? I think there should be a solution.

Score:2
ng flag

You need two different public IP addresses.

One will forward TCP port 25 to your antispam system and will be used as your public MX record.

The second one will forward TCP port 25 to your Exchange server and will only be used by your Exchange Online hybrid connectors; you can restrict inbound traffic to this IP address to Exchange Online IP address ranges.

Tobia avatar
in flag
Yes, I tought about this solution for the second NAT, but it looked not very safe to me... Maybe it is safe enought because nobody can use Exchange online to send email to an ip different from the public MX... but I'm not sure. However if this is the only solution, I'll take it. Thanks!
I sit in a Tesla and translated this thread with Ai:

mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.