Recent content by steven99

Seems to be intentional considering there are separate firewall macros for POP3 and IMAP (and their SSLs versions) and I can see the benefit in having the different kinds of mail traffic being in different macros. Not all mail servers want to be MTAs and not all MTAs want to be fetch points.

Yup, that would be correct.

Not via the API, but there is nothing stopping you from writing a simple PHP script that collects a POST and then uses a SMTP via a SMTP lib to then send the mail out via PMG.

Is the state for the node2 still syncing? If so, that is likely where the issue is and node2 didn't get the blacklist sync'd . If you check node2's blacklist settings, does it have the blacklist setup with that domain? If not, you need to find out what is causing the syncing issue.

Domains don't have owners and so what you're looking for isn't possible. What you could do is build a wrapper system that uses the API and keep track of what domain should go to what user in that wrapper system. For example, in a WHMCS or Blesta server module what you're looking for can be...

You can create a new cluster on existing setup but you can't join a cluster on existing IIRC. Though if you plan on using Ceph, you will need other drives to use as OSDs.

No disadvantage. HELO is part of the SMTP RFC I am sure and any MTA not following RFCs are likely spammers anyhow.

Just to confirm, you did a nslookup, dig, or other DNS lookup tool on the PMG server it self to confirm it could find test.com's MX record? What about test.local and rDNS for the relaying server? Not on PMG it self, but I have seen that error in postfix logs when it could not lookup a rDNS on...

In postfix, you just set the smarthost to be: pmg-server.example.com:26 where pmg-server.example.com is your PMG server address and :26 is the Internal SMTP port set under Mail proxying -> Ports . With that said, any mail going out from your postfix server will go through PMG and then PMG...

Looks like you're are not passing the cookie and header. $currentDomain, $PMGcookie, and $CSRFtoken need to be set before the curl happens . If you know PHP, it might be easier to take the VE API php examples and adapt them here.

You replaced "hostname" with "test" within the tildes and that is likely where the issue is as test is not a valid command. Putting something between tildes ("`") instructs bash to run that as as command and to use that output. You could remove that part if you want and just have the comment...

You need to first get the access ticket via curl, wget, whatever and get the PMGAuthCookie and CSFR token into bash script variables. How you get those there is up to you. It returns a json, so you'll need to parse it out. Using the jq tool seems to be a commonly recommended method for json...

If you just want to run the API URLs in your browser, just login to the node and then run the API url in another browser tab. The API will use your current session cookie for the API calls -- very handy for testing out calls during dev. Otherwise, if connecting to API from a script, like...

Would second a built-in rate control.

Configuration -> Spam detector -> Quarantine -> Quarantine host should be the setting you want.