I don't know why postfix suddenly cannot send out email. I found that clamd.amavisd cannot start, I tried reinstall it and no luck. I know I must re-activate it but now more urgent is let my client send emails. I searched the whole night and found that Falko post comment out belows in main.cf #content_filter = amavis:[127.0.0.1]:10024 #receive_override_options = no_address_mappings After restart postfix, all emails still send through port 10024 and state, delivery temporarily suspended: connect to 127.0.0.1[127.0.0.1]:10024: Connection refused all emails hold in queue. Then I also found another post to comment out master.cf #amavis unix - - - - 2 smtp # -o smtp_data_done_timeout=1200 # -o smtp_send_xforward_command=yes #127.0.0.1:10025 inet n - - - - smtpd # -o content_filter= # -o local_recipient_maps= # -o relay_recipient_maps= # -o smtpd_restriction_classes= # -o smtpd_client_restrictions= # -o smtpd_helo_restrictions= # -o smtpd_sender_restrictions= # -o smtpd_recipient_restrictions=permit_mynetworks,reject # -o mynetworks=127.0.0.0/8 # -o strict_rfc821_envelopes=yes # -o receive_override_options=no_unknown_recipient_checks,no_header_body_checks # -o smtpd_bind_address=127.0.0.1 After that then restart postfix again, now it shows "mail transport unavailable". All emails still hold in queue. I think that ISP3 changed some settings in master.cf for amavis, and wiped out original master.cf settings. Can anyone help me what I have to add back in master.cf for normal transportation? Really very urgent!! Thanks so much!!
I am using Fedota 15, updated all rpms and update to ISPConfig 3.1 since the clamd.amavisd cannot start. This is the only log in /var/log/messages Oct 23 11:59:23 ws1 systemd[1]: clamd.amavisd.service operation timed out. Terminating. Oct 23 11:59:23 ws1 systemd[1]: Unit clamd.amavisd.service entered failed state. The emails status in postfix are delivery temporarily suspended: connect to 127.0.0.1[127.0.0.1]:10024: Connection refused
Check your log files to try to determine why amavis won't start up. You did answer yest to "reconfigure services" during the ispconfig upgrade?
Thanks Jesse, I temporarily made postfix work now without amavis, clamav and spamassassin. But 100% sure want those back to postfix. Just before, config in master.cf are ports 10025 and 10027. I just thought postfix will pass emails to amavis through 10024 to check and amavis will pass emails back to postfix through 10025. So, if I stopped amavis, postfix doesn't know where to pass, that's why I got 10024 suspended. And, if I comment out config 10025 and 10027 in master.cf, postfix will show "mail transport unavailable", that's correct that no ports to pass through. Then I uncomment those ports and change them to 10024 and 10026. Everything goes fine, just no anti-everything. With postfix works, I tried erase and reinstall all RPMs, clamav-server, amavisd-new, spamassassin with all dependents. amavisd.service can start, clamd.amavisd fail. There's just a 2 logs in /var/log/messages Oct 23 12:52:55 ws1 systemd[1]: clamd.amavisd.service operation timed out. Terminating. Oct 23 12:52:55 ws1 systemd[1]: Unit clamd.amavisd.service entered failed state. What can I do? Can I install RPM of amavisd-new and self compile clamav 0.99.2? What I have to change if I make this choice? Really need all of you help. Thanks.
Dear Till, Thanks for your reply. There's no error log when start amavis. Amavis start successful, just clamd.amavisd cannot start, the error just got 2 lines, Oct 23 12:52:55 ws1 systemd[1]: clamd.amavisd.service operation timed out. Terminating. Oct 23 12:52:55 ws1 systemd[1]: Unit clamd.amavisd.service entered failed state.
There should be some logs for why clamd doesn't start, check clamd.conf for LogSyslog or LogFiles settings. It's also curious that amavis won't work without clamd .. at least on debian, if clamd doesn't run, it falls back to clamscan command (which is not efficient and bogs, but better than absolutely nothing).
Dear Jesse, thx for your reply but while the server is running, I can only try on weekend. One question, the server is now running well without amavisd and clamd installed, can I yum install it without change postfix settings? I just afraid it cause the postfix hold the queue again. I will die... thx man!
I imagine installing amavis and clamd won't affect your postfix configuration, though I don't run Fedora, so I'm just assuming it acts sanely. Starting clamd won't affect postfix either, and you should be able to work on getting that to run (even though no mail is being processed by it). I'd guess starting amavis would result in an error that port 10024 is in use (by postfix); and potentially a problem, if you rebooted, amavis might get started first, causing postfix to not start correctly (because port 10024 would then be in use by amavis) - so if you install it, make sure amavis is set to not start for now.
By any chance was port 10026 used anywhere else besides amavis? I had that problem when I upgraded to ispconfig 3.1. amavisd needed 10026. Beforehand I had configured dspam to use 10026 then after the upgrade I had to move dspam to another port. For me, because of the conflict of the port numbers made amavisd fail. Hope this helps.
Dear Jesse and Anthony, Thanks so much for your reply. I also just think that may cause by any port conflict but actually that we haven't touch the server those days. So, there won't any update or upgrade or newly installed services. Anyway, I will try to figure it out. By the way, how can I check if those ports get used? I am not familiar with "ps" command or any other I can check that out? Thanks.
Try this http://superuser.com/questions/529830/get-a-list-of-open-ports-in-linux or http://www.cyberciti.biz/faq/how-do...-are-listeningopen-on-my-linuxfreebsd-server/ or https://access.redhat.com/documenta...rity-Verifying_Which_Ports_Are_Listening.html