Too Many Hops

[gnulab]

Hi,

I am using Proxmox MG and Zimbra as the webmail portion.

Everything had been fine until one day all users were unable to send emails to gmail and any custom domains hosted on gmail infrastructure.

Till now all users can still send email to other domains, except for gmail.

Sending email to this temp email address (https://temp-mail.org/en/) is received finely.

Am I correct to say the issue is at the proxmox portion? And how do I correct this issue?


Thank you

Here is the Mail Header

Received: from webmail.example.com (webmail.example.com [111.222.333.444])
    by mg.example.com (Proxmox) with ESMTPS id A9615460A1B
    for <htw@gmail.com>; Tue,  8 Aug 2023 10:59:39 +0700 (WIB)
Received: from localhost (localhost [127.0.0.1])
    by webmail.example.com (Postfix) with ESMTP id 69A904007C5
    for <htw@gmail.com>; Tue,  8 Aug 2023 10:59:39 +0700 (WIB)
Received: from webmail.example.com ([127.0.0.1])
    by localhost (webmail.example.com [127.0.0.1]) (amavisd-new, port 10032)
    with ESMTP id GALT-4ZJFlIM for <htw@gmail.com>;
    Tue,  8 Aug 2023 10:59:39 +0700 (WIB)
Received: from localhost (localhost [127.0.0.1])
    by webmail.example.com (Postfix) with ESMTP id 421C24007C7
    for <htw@gmail.com>; Tue,  8 Aug 2023 10:59:39 +0700 (WIB)
X-Virus-Scanned: amavisd-new at example.com
Received: from webmail.example.com ([127.0.0.1])
    by localhost (webmail.example.com [127.0.0.1]) (amavisd-new, port 10026)
    with ESMTP id ca7z242B3dQH for <htw@gmail.com>;
    Tue,  8 Aug 2023 10:59:39 +0700 (WIB)
Received: from mg.example.com (mg.example.com [149.28.147.58])
    by webmail.example.com (Postfix) with ESMTPS id 1DE6D400674
    for <htw@gmail.com>; Tue,  8 Aug 2023 10:59:39 +0700 (WIB)
Received: from mg.example.com (localhost.localdomain [127.0.0.1])
    by mg.example.com (Proxmox) with ESMTP id 0097E46125A
    for <htw@gmail.com>; Tue,  8 Aug 2023 10:59:39 +0700 (WIB)

 

[Stoiko Ivanov]

Seems you have created a mail-loop:
* check what zimbra does with those mails

* do you have listed gmail.com in your relay-domains? - do you have configured a transport for gmail.com?!

 

[tempomails]

Just adding to Stoiko’s point, since the header does look like a loop: mg → webmail → amavis → webmail → back to mg. A few things worth checking:

1. In PMG, make sure gmail.com isn’t in your relay domains and has no transport entry pointing to Zimbra.
2. In Zimbra, check that outbound mail for external domains goes to the PMG as a relay host, but that PMG doesn’t hand it straight back to Zimbra for non-local domains.
3. Compare the Received lines between a Gmail test and a working domain. The loop usually shows up as the same two hostnames repeating.

One handy way to isolate it is to test with several different recipient domains. A free temporary email works well as a throwaway recipient to confirm that only Gmail-routed mail is looping. (Disclosure: I work on that tool.)