Page 2 of 7 FirstFirst 1234 ... LastLast
Results 11 to 20 of 62

Thread: local mail getting marked as spam?

  1. #11
    Join Date
    Nov 2005
    Posts
    175
    Rep Power
    10

    Default

    Ok I did a little homework, and I get what the SMTP Sumbission port is, what it's for, I've got some links on how to enable it, etc.

    Does anyone actually know if using port 587 or maybe port 465 gets around the problem of Zimbra still trying to process the spam rules?

  2. #12
    Join Date
    Jul 2007
    Posts
    45
    Rep Power
    8

    Default

    I have not tested this with 5.0 yet, but with 4.5.6 authenticated SSL smtp on 465 did not skip the spam checking. Maybe someone else has more info on 587, but I do not think it will be any different.

    Mark

  3. #13
    Join Date
    Nov 2005
    Posts
    175
    Rep Power
    10

    Default

    I'm sure you're right. Since it doesn't appear to me that port 587 is even officially supported by Zimbra (I don't think there's any Zimbra command line or administrator option to turn it on, I believe it's a Postfix setting only), I doubt they would have specifically addressed an issue by having us change a setting they didn't even expect us to use.

    Seems to me this is a fairly serious issue. Anyone who travels and doesn't use the Zimbra web client or a VPN would have this problem. Seems like it would be easy enough for a spammer to use my domain or even valid email addresses on my domain as sender addresses, I can't quite justify whitelisting the whole domain. Surely there's some other way to fix this?

  4. #14
    Join Date
    Nov 2005
    Posts
    175
    Rep Power
    10

    Default

    This seems like a problem that would affect a lot of people. Should I file a bug for this or something?

  5. #15
    Join Date
    Jul 2007
    Location
    San Jose, CA
    Posts
    1,027
    Rep Power
    10

    Default

    I think you guys are partly on the wrong track. If you look at the headers of the message that the OP submitted, he must be correctly authenticated already as the only "received" headers are from the internal IP and localhost. But he's getting dinged in Pyzor and SORBS, which both check on IP address. I'm betting that either his real domain (whatever myserver.com obfuscates) or else the internal 192.168.1.x range have landed in the blacklisters, so that other people "out there" who use the same blacklists would also have his email flagged. That's the first thing I would check based on the scores I am seeing.
    Cheers,

    Dan

  6. #16
    Join Date
    Jul 2007
    Location
    San Jose, CA
    Posts
    1,027
    Rep Power
    10

    Default

    Quote Originally Posted by jholder View Post
    Marking them as not spam should train SA to ignore that.
    Not gonna accomplish much since he's already getting a BAYES_05 score which is fairly strongly "not spam." Unless he increases the negative score for BAYES_05 and BAYES_00, it'll not be strong enough to counteract the high positive scores from SORBS and PYZOR
    Cheers,

    Dan

  7. #17
    Join Date
    Jul 2007
    Posts
    45
    Rep Power
    8

    Default

    What I would like is for mail delivered via an authenticated connection to skip the spam check completely.

    Quote Originally Posted by dwmtractor View Post
    I think you guys are partly on the wrong track. If you look at the headers of the message that the OP submitted, he must be correctly authenticated already as the only "received" headers are from the internal IP and localhost. But he's getting dinged in Pyzor and SORBS, which both check on IP address. I'm betting that either his real domain (whatever myserver.com obfuscates) or else the internal 192.168.1.x range have landed in the blacklisters, so that other people "out there" who use the same blacklists would also have his email flagged. That's the first thing I would check based on the scores I am seeing.

  8. #18
    Join Date
    Jul 2007
    Location
    San Jose, CA
    Posts
    1,027
    Rep Power
    10

    Default

    Quote Originally Posted by mdeneen View Post
    What I would like is for mail delivered via an authenticated connection to skip the spam check completely.
    I understand why you would say this (and it's worth filing an RFE) but I'm not sure I agree. If there is something broken with my IPs (maybe my ISP) or my config that gets my messages tagged as spam (for other people) I'd want to know it. . .

    Seems to me that in this case, whitelisting only sidesteps whatever other problem is causing the messages to be flagged
    Cheers,

    Dan

  9. #19
    Join Date
    Jul 2007
    Posts
    45
    Rep Power
    8

    Default

    Quote Originally Posted by dwmtractor View Post
    I understand why you would say this (and it's worth filing an RFE) but I'm not sure I agree. If there is something broken with my IPs (maybe my ISP) or my config that gets my messages tagged as spam (for other people) I'd want to know it. . .

    Seems to me that in this case, whitelisting only sidesteps whatever other problem is causing the messages to be flagged
    Ok, I'll give you this and revise my original statement.

    I would like mail sent via an authenticated SMTP connection TO a local account to skip the spam filters. Doesn't it feel wrong when the results from logwatch sent from servers on the local network end up in the junk mail folder?

  10. #20
    Join Date
    Jul 2007
    Location
    San Jose, CA
    Posts
    1,027
    Rep Power
    10

    Default

    Quote Originally Posted by mdeneen View Post
    Ok, I'll give you this and revise my original statement.

    I would like mail sent via an authenticated SMTP connection TO a local account to skip the spam filters. Doesn't it feel wrong when the results from logwatch sent from servers on the local network end up in the junk mail folder?
    Perhaps, and that is why I said it would be perfectly reasonable to file an RFE (if you don't want to do so yourself, I'll be happy to try to translate it appropriately to a bugzilla request), however, I would definitely want it to be a switchable option. Here's why:

    Let's say you're sending messages to customers but they don't get through, because your IP (or IP range due to a sloppy ISP) has landed in somebody's public blocklist. How are you going to know that something is wrong? Your customers may not know you sent something that didn't come thru, unless you call them to verify, which sort of makes email redundant. On the other hand, if you find that email from yourself to yourself is getting caught in your own spam filters, it gives you a "heads up" that you need to research and determine why you're sitting on somebody's blacklist. You can take the necessary steps to remedy the situation, and your own system will monitor whether you've succeeded or failed.

    Failing that, you're left to wait for some customer to be upset because you never sent him/her that promised email, in order to discover that you have an issue at all. To my way of thinking that's not the ideal way to do the job. . .
    Cheers,

    Dan

Similar Threads

  1. Replies: 7
    Last Post: 02-03-2011, 07:01 AM
  2. Problem with Postfix and MTA
    By ZMilton in forum Administrators
    Replies: 16
    Last Post: 04-16-2008, 07:47 AM
  3. [SOLVED] Mailserver down when send file attach of 50Mb
    By ZMilton in forum Administrators
    Replies: 20
    Last Post: 04-10-2008, 12:44 PM
  4. fresh install down may be due to tomcat
    By gon in forum Installation
    Replies: 10
    Last Post: 07-25-2007, 09:09 AM
  5. receiveing mail
    By maybethistime in forum Administrators
    Replies: 15
    Last Post: 12-09-2005, 04:55 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •