Skip to main content

Senderbase Poor Reputation Help

Comments

7 comments

  • cPanelMichael
    Hello, Have you tried adding a new IP address to your server and using it for sending email? You can find details on how to configure this at:
    0
  • ag1266
    Thanks for the reply, but I really need to get to the underlying cause of the problem or any new IP address is just likely to end up with a "poor reputation" again, putting me back at square one. Funny thing is, I just tried to follow-up with senderbase support as this problem has only occured since I migrated to new servers and new IP addresses in December. Everything before was identical so I'm wondering if these IPs have historical spamming reputations, but lo and behold my email to them was bounced because of "poor reputation". You couldn't make this stuff up...
    0
  • cPanelMichael
    Everything before was identical so I'm wondering if these IPs have historical spamming reputations

    This is what I suspected. It might be worth adding a new IP address and using it for sending just to rule that out as the cause of the problem. Thank you.
    0
  • ag1266
    Senderbase have come back saying "Using generic ISP strings can be suspicious. The cure would not to add additional PTRs (that will not help at all) but to use a more specific HELO like mailserver.domain.tld (well, and then have the PTR match that obviously)." How can you have a HELO of mailserver.domain.tld and have the PTR record match when you have multiple domains on the server without having multiple PTRs? There must be thousands of people out there hosting multiple sites/domains on a single server who also use those servers for email so I can't understand why I'm having these problems. It's taken god knows how long so far and I'm getting absolutely nowhere. :(
    0
  • cPanelMichael
    Right, the standard behavior is to setup RDNS for the hostname of the server. My advice is to try adding a new IP address and using it for sending just to see if it makes a difference. Thank you.
    0
  • ag1266
    Just want to (hopefully!) close this off in case it's of help to anyone else in the future. I eventually found out it's not down to IP address, it's down to server hostname and, in particular reverse DNS. Senderbase now flag any server that has a non-unique generic ISP name as "poor" by default. They require a unique domain name per server for reverse DNS so, for example if you have server1.isp-provider.com and someone else has server2.isp-provider.com that is bad in their eyes as reverse DNS for both servers point to the same domain name. A point to note is that even if reverse DNS is correctly set up using the default hostname/server name, they will flag on their reputation page that forward/reverse DNS do NOT match. It's very confusing as every other test utility out there will indicate all is fine. To get round it, pick the domain name of one of the sites hosted on the server (doesn't matter which as long as it's unique) and set reverse DNS to that domain name. An hour after making the change Senderbase showed reputation on both servers as "neutral" from "poor". I also had to change /etc/mailhelo to reflect the change. It was set to use default ISP provided hostname and has been changed to mail..com - bounce EXIM after making this change. One other thing cropped up though - after a reboot (for reasons not related) I could not send or receive mail. Transpired that the rules for POP3, IMAP and SMTP within iptables were only configured for the default ISP provided hostname. I had to add new iptables rules for mail..com (even though the original hostname is still the overall hostname for the server). I have no idea why this is - whether iptables uses reverse DNS in some way or not, I just know doing that fixed it. Rules need to be added to /etc/sysconfig/iptables to be permanent and re-applied after reboot. Now, four days after making the change, reputation on one server is showing as "good", the other is showing as "neutral" and the delayed/queued mail on both servers has cleared. It was a LOT of investigation/work and Senderbase support are useless, but all appears to now be working as it should. I thought I might have to change SPF records for all hosted domains to say that mail..com was a valid sender, but so far that doesn't seem to be neccessary. Am continuing to monitor.
    0
  • cPanelMichael
    I am happy to see you were able to address the issue. Thank you for updating us with the outcome.
    0

Please sign in to leave a comment.