Hello, I am using FC15 and following the tutorial for setting up Postfix. I have run into a bit of a dilemma and I have been searching all over the place to try to find an answer. Sorry for the lengthy post, but I figured I'd tell my tale so that everyone can see where I've been; this way I won't be told to go there again. I installed amavisd-new and clamav as per Page 4 of the tutorial. I changed the settings in amavisd.conf as instructed (entering my correct server information, of course). When I entered the command to start amavisd, I was presented with the following error: Code: Starting amavisd (via systemctl): Job failed. See system logs and 'systemctl status' for details. [FAILED] I noticed it did not give me a log to check. So I went to /var/log and looked at the messages log, where I got this Code: Oct 2 00:35:14 server systemd[1]: amavisd.service: control process exited, code=exited status=255 Oct 2 00:35:14 server systemd[1]: Unit amavisd.service entered failed state. I searched for the above error string, only to be sent to articles about what status 255 means and not answering my question. The Force guided me to run Code: # cd /etc/init.d # amavisd (note the absence of 'start') At first it told me that the file /etc/amavisd.conf did not exist and that the script was failing at line 1795 in /usr/bin/amavisd. So, I went to edit said script file, only to find that the variable was set by another file (or at least it appeared to be. Searching for $config_file only sent me on an endless circle with nowhere to set the file path to the conf file). So, I created a symbolic link to point to /etc/amavisd/amavisd.conf; I then ran the amavisd command above again. Apparently $myhostname needed to be a FQDN and did not like my machine's name, which was simply 'server.' Funny, I don't remember that being in the tutorial. I should probably add that I've followed similar tutorials for the previous versions of Fedora and never had amavisd give me this problem. Anyway, I did as requested by amavis: I uncommented the line and entered a FQDN in $myhostname and commanded Fedora to start the flippin' program: Code: # /etc/rc.d/init.d/amavisd start I about threw the screen out the window when the error message about it failing came up again, but this time the code was 102. However, I remembered my Jedi training and did not give in to the Dark Side. So, I went and looked in the messages log... No new error entry there... I looked in maillog, and behold! Amavis created the BDB in /var/spool/amavisd, and all other kinds of activities (PERL modules loading)... But wait, Fedora told me the process failed, so how is this happening? Code: # service amavisd status amavisd.service - SYSV: amavisd is an interface between MTA and content checkers Loaded: loaded (/etc/rc.d/init.d/amavisd) Active: failed since Sun, 02 Oct 2011 00:44:00 -0400; 8min ago Process: 32438 ExecStart=/etc/rc.d/init.d/amavisd start (code=exited, status=102) CGroup: name=systemd:/system/amavisd.service # cd /etc/init.d # amavisd The amavisd daemon is already running, PID: [32355] So now I am really confused! How can I have a PID (in essence, it's running) yet the process failed? To make sure SELinux wasn't giving me any issues, I set it to permissive mode, stopped amavis and started it again. I was given the same failed messages in the beginning and amavis wouldn't start... At this point, if I again process Code: # cd /etc/init.d # amavisd I get: Code: The amavisd daemon is already running, PID: [32355] So, here is where I am now... can anyone please shed some light on this. Did I miss a setting somewhere?
To which tutorial (URL) are you referring? Have you tried to delete the PID file manually and restart amavisd?
http://www.howtoforge.com/virtual-u...urier-mysql-and-squirrelmail-fedora-15-x86_64 I wasn't expecting that to work, but it did... Thanks, falko! You da man!
The same issue I have no pid file anywhere on my system and amavisd still will not start. I even nulled the amavisd.conf file and re-copied the complete contents from the howto. Still no go. It's not your fault Falko, but I really hate error messages that give you no clue where to look. Here they are as such: =============================================== [root@www ~]# service amavisd status amavisd.service - SYSV: amavisd is an interface between MTA and content checkers Loaded: loaded (/etc/rc.d/init.d/amavisd) Active: failed since Tue, 21 Feb 2012 18:44:06 -0600; 11min ago Process: 19117 ExecStart=/etc/rc.d/init.d/amavisd start (code=exited, status=255) Main PID: 2309 (code=exited, status=0/SUCCESS) CGroup: name=systemd:/system/amavisd.service ------------------------------------------------------------- Feb 21 18:35:34 www systemd[1]: amavisd.service: control process exited, code=exited status=255 Feb 21 18:35:34 www systemd[1]: Unit amavisd.service entered failed state. Feb 21 18:44:06 www systemd[1]: amavisd.service: control process exited, code=exited status=255 Feb 21 18:44:06 www systemd[1]: Unit amavisd.service entered failed state. ------------------------------------------------------------- Feb 21 18:44:06 www amavis[19121]: starting. /usr/sbin/amavisd at www.example.com amavisd-new-2.6.6 (20110518), Unicode aware, LANG="en_US.UTF-8" Feb 21 18:44:06 www amavis[19121]: Perl version 5.012004 =============================================== Thank in advance, Will
Further trial and error testing has indicated the following line in amavisd.conf is causing it not to start: @lookup_sql_dsn = ( ['DBI:mysql:database=mail;host=127.0.0.1;port=3306', 'mail_admin, 'Mail_Admin_Password'] ); When I remark it out Amavisd starts. I tried reinstalling the Perl DBI stuff but it didn't help. I've tested the MySQL database connection and it works. All I know now is tomorrow I'll have a lot of users mad at me. <Grin>
Is MySQL using TCP and listening on port 3306? You can check with Code: netstat -tap Also make sure you use the correct MySQL login in your amavisd configuration.
Yes. It is using 3306. Could it be an issue with systemd? I also re-entered the login user and password after verifying they work using mysql login.
I had to do a force rebuild of Perl modules Bundle::Mysql to get it to work. It was some version problem with two versions of DBD::mysql on my system. It wouldn't allow mysql.so to start. Thanks anyway.