Spamd not working on 1.2.3



  • I'm have not much experience not in pfSence nor in freeBSD, but I'll try to describe the problem.
    After upgrading from 1.2.2 to 1.2.3 beta spamd stopped whitelisting at all.
    It runs in graylisting mode, and in 1.2.2 all was OK. But after upfrading  to 1.2.3 only messages from addresses in /var/db/whitelist.txt passed to my mail server. All incoming messages from other sources, including already whitelisted by spamd, remains in GRAY forever.
    After downgrading back to 1.2.2 spamd works well again. Whitelisted sources starts do deliver immediately.
    I suppose, that spamd can't set whitelisted sources in pf in 1.2.3.
    Is it possible to config something in 1.2.3 to repair spamd behavior, or we need to wait new spamd release for 1.2.3?



  • Can you please add your config file for spam d?



  • /etc/spamd.conf

    all:\
    	:whitelist:blacklist:custom:UofATraplist:Nixspam:
    
    whitelist:\
    	:method=file:\
    	:white:\
    	:file=/var/db/whitelist.txt:
    
    blacklist:\
    	:black:\
    	:msg="Sorry, you spammed us before.":\
    	:method=file:\
    	:file=/var/db/blacklist.txt:
    
    custom:\
    	:white:\
    	:msg="":\
    	:method=file:\
    	:file=/var/db/whitelist.my:
    
    UofATraplist:\
    	:black:\
    	:msg="Sorry, you are on The University of Alberta traplist.http://www.openbsd.org/spamd/traplist.gz":\
    	:method=http:\
    	:file=www.openbsd.org/spamd/traplist.gz:
    
    Nixspam:\
    	:black:\
    	:msg="Sorry, you are on The Nixspam spamlist.http://www.openbsd.org/spamd/nixspam.gz":\
    	:method=http:\
    	:file=www.openbsd.org/spamd/nixspam.gz:
    

    in GUI modified items:

    Grey listing - yes
    Passtime - 2
    Grey Expiration - 4
    White Exp - 2160
    Stutter Secs - 15
    Delay Secs -2
    Window Size - 1
    Enable RRD graphing - yes



  • I never got it working on 1.2.1!



  • It works OK on my box from 1.2.0
    But not on 1.2.3…



  • That package is not being actively maintained, so there's a good chance it won't be functional in the 1.2.* branch.  Consider posting a bounty.



  • Under pfsense 1.2.3 it seems, that the '_spamd' user needs more authorization.
    Set 'user pw user mod _spamd -u 0' and try it again!



  • TNX. Will try.



  • OK so I'm going to show my ignorance of BSD:

    I went to the shell and tried:

    Set 'user pw user mod _spamd -u 0'

    and then tried:

    user pw user mod _spamd -u 0

    but neither worked. What am I doing wrong?

    Thanks,

    Mike



  • you have to type:
    pw user mod _spamd -u 0
    it will work



  • @kongar:

    you have to type:
    pw user mod _spamd -u 0
    it will work

    This does seem to fix SpamD up for 1.2.3, which I've been trying to figure out (should come read the forums lol!)

    Thanks Kongar!



  • Yeh - I have it now working under 1.2.3 - It has MANY limitations though but does work.

    1. It doesn't seem to support multi WAN.
    2. White-listing and black-listing are flaky.
    3. And those jiggly buttons are just plain stupid and annoying!

    Don't create any NAT or firewall rules for port 25 it opens it up for you.

    I did run the user mod tweak above to get it working, not sure if you still need too.



  • @cheesyboofs:

    Yeh - I have it now working under 1.2.3 - It has MANY limitations though but does work.

    1. It doesn't seem to support multi WAN.
    2. White-listing and black-listing are flaky.
    3. And those jiggly buttons are just plain stupid and annoying!

    Don't create any NAT or firewall rules for port 25 it opens it up for you.

    I did run the user mod tweak above to get it working, not sure if you still need too.

    Yes, that about covers it… :) I haven't seen the white list or black list be flaky tho..  are you talking the counts or the actual blocking/allowing from the list?
    The jiggly buttons do work, despite being annoying ;) there could definitely be a better feedback for clicking them, the package as a whole is what'd i'd call 3/4 done.

    So, if I were to overhaul the package besides multi-wan what else needs to be included in it?  I use it in a very un-normal way (I just require every mail server to hit it twice to get through, that's it, no wait time) but it cuts 90% of our spam which was from Dynamic addresses... Do you want the ability to specify which IP to capture traffic on? (I've kind of wanted this, so you could send one through spamd but another mail server/ip doesn't go through for instance..)

    Lets get a hit list together..



  • Just a few niggles,

    1.) Loose the jiggly buttons  - I hate them.
    2.) Cant seem to add a range of addresses to the whitelist i.e. 81.103.221.0/24 I had to add each ISP server manually some 18 servers.
    3.) Why don't things that have been auto white-listed appear in the Spamd White list?
    4.) Would be nice to tick a box so that the Spamd database would just show ALL the GREY items I range from 13 to 35 at any one time but every time I'm over 25 I have to adjust the filer limit.
    5.) Date and time stamp the Spamd database, I still to this day don't know if it runs top to bottom, or bottom to top.

    I was always led to believe spamd listens on 127.0.0.1 so should work on multi wan but if I do a shields up port scan https://www.grc.com/x/ne.dll?bh0bkyd2 port 25 is only open on the wan, maybe its just not opening up port 25 on the other interfaces.

    I does kind of work and is a worthy addition to pfsense.

    Cheers


Log in to reply