Due to some intricacies of reverse-proxy stuff (which I'm really not going to go into for security reasons), I have added an entry to the /etc/hosts file for every node in a particular Proxmox VE cluster. This is so that the FQDN of the SMTP server I am sending notifications through resolves to a particular IP, instead of the one for the reverse-proxy. This is so that each node, when sending any E-Mail notification, directly connects to said SMTP server.
When I send a Test E-Mail, it works, and connects to the SMTP server. I get the E-Mail, looks good.
When I do a manual single backup, the E-Mail comes through just fine.
My daily backup tasks for VMs I care about, however, fails. And the problem is that when this task happens and the E-Mail is sent, it resolves to the wrong IP (the reverse-proxy) as it ignores the /etc/hosts line. I know this because I receive a message rejected alert (somehow?) and it reports the FQDN resolves to a particular IP, which... is the reverse-proxy IP and not the SMTP server IP.
I'm not sure if this classifies as a bug, but this inconsistent behaviour does not meet expectations, so to say. I see a bunch of possible solutions I could try, but I wanted to hear from the forums/devs before going further.
All nodes in the cluster are on the latest Proxmox VE version as of a few days ago.
When I send a Test E-Mail, it works, and connects to the SMTP server. I get the E-Mail, looks good.
When I do a manual single backup, the E-Mail comes through just fine.
My daily backup tasks for VMs I care about, however, fails. And the problem is that when this task happens and the E-Mail is sent, it resolves to the wrong IP (the reverse-proxy) as it ignores the /etc/hosts line. I know this because I receive a message rejected alert (somehow?) and it reports the FQDN resolves to a particular IP, which... is the reverse-proxy IP and not the SMTP server IP.
I'm not sure if this classifies as a bug, but this inconsistent behaviour does not meet expectations, so to say. I see a bunch of possible solutions I could try, but I wanted to hear from the forums/devs before going further.
All nodes in the cluster are on the latest Proxmox VE version as of a few days ago.