* I just installed Clement binary RPM and nothing is happening, what need to be done? As it is the first installation, Clement daemon is not started. During install Clement configure itself to wait for an SMTP connexion on eth0. Make sure your current SMTP application (sendmail, postfix, exim,...) is not listening on that IP already, then start Clement the usual daemon way: 'service Clement start' Once started you can check Clement process is up and running by doing 'telnet "your_eth0_ip_number" smtp' you should receive a sign-on like 220 host.dom.name ESMTP Clement-2.0-XXX; Tue, 02 Aug 2005 10:16:07 -0400 type 'quit' to exit. Next time you reboot, Clement daemon will be started automatically as well as any other daemon application needed by your server. * Before starting Clement I want to check the configuration done by the install process, where are the critical file? All configuration file are stored with the conf directory within the application base directory (/usr/local/clement). Critical file is clement.conf, format is plain ASCII and you can change it according your need (field description are embedded within the file itself). We strongly suggest to add ADMIN entries. The install process configured the 'root' account to be the super user while accessing clement via the WEB. Better to add real user name of those in charge to manage Clement. * Well, I do not mind to configure Clement by editing file, but some of my colleague prefer to use the WEB interface. Is Clement WEB interface available? Yes, while installing, RPM process add some definition to your apache configuration. URL 'clement.your_domain_name' give you access to the application. Once authenticated (via PAM or POP), if your are not defined in the configuration admin list you will reach the standard Clement WEB page, the one accessible to all your end-users. User among the admin listed, have access to extra icon, allowing to configure Clement binding interface. In order to have the Clement WEB interface accessible by plain end-user, we suggest to add 'clement.your_domain_name' within you DNS definition. [esearch.png] * Clement seems quite useful but I am a little bit reluctant to put the zillion domains my E-mail server is taking care on its only control right now, is there a way to challenge Clement with one 'test domain' first? No problem. To do so, you need to have an IP interface available on the server (set an ip alias with ifconfig). Access the Clement's web interface (clement.your_domain_name) and once authenticated change the binding IP to the IP alias dedicated to your test domain. Within minute, Clement should bind itself to the new interface. Then change the MX assigned to the 'test domain' to be the IP alias set aside for the test. * Clement seems not using my spam daemon (spamassassin) to trap obvious spam contents, do I need to define something? Cross check if the variable SPMDSRV (in clement.conf) is defined and set on the good host using the good port. By default, SPMDSRV is defined as localhost looking for port 783, if your spamassassin is not configured that way, you need to adjust SPMDSRV setting. * Clement seems not working where can I find debug information? All trouble, alert or Clement 'questioning' are reported via syslog. You must make sure a debug entry is defined within your syslogd configuration file (/etc/syslog.conf). Expected entry is like: *.=debug /var/log/thedebugfilename If missing, add one at the syslog.conf file end, then restart your syslog daemon (Be aware to update the logrotate definition to have new file expired). By default, clement report only real problem to syslog, you can increase the debug log verbosity by adding: options="-dx" in file '/etc/sysconfig/clement". Higher the 'x' value is (0 is production verbosity) more verbose Clement will be. You can start Clement in foreground and in 'verbatim' mode too, please try 'clement -h' to check about the available command line option. * Clement is working and filtering E-mail on my 'test domain', is there any logs I can look at? Yes, within the /usr/local/clement/logs directory you can find 2 file collecting data for the current day. The clement-YYYY-MM-DD.jrl, collect all clement action, updating the virus library,connecting to 'sender' or 'checker' server, dispatching acceptable E-mail, etc... The event-YYYY-MM-DD.jrl, collect all E-mail receiving status and decision making. Event file are very useful to find out why a E-mail was rejected and give feed back to end-user. * Clement put some E-mail in server quarantine area, how can I have that E-mail transmitted to end-users. You have nothing to do, end-user need to connect on URL 'clement.your_domain_name', authenticate himself, then select E-mail to be forwarded and click on the send icon to have E-mail sent within 2 minutes. If you want to make a 'special' to an end-user you can do it yourself provided you are defined as a Clement admin in the configuration file. If your Clement is connected to our Mentor application, Clement will learn about E-mail originator and next time the E-mail will be transmitted directly to recipient. [qua.png] * What happen to old E-mail stored in server quarantine area? A cron shell is executed every day and E-mail older than 20 days in quarantine area are removed. Cron shell is stored within /usr/local/clement/conf (named cron.conf) and can be tailored to fit your needs. * Ok, Clement is working fine, I want to have all my zillion of domains filtered by it, where do I defined them? Using the WEB interface (or editing the clement.conf file), define the special domain name 'All_Other_Domains' to be binded to your MX listening interface, that's all! (be sure to have no other SMTP server using that interface). * It is my understanding Clement can be connected to a remote 'Mentor' process, what is the purpose? Clement is able to request advice to remote Mentor to make an even better analyze of the received E-mail. Mentor is working on the sending context and past recipient exchanges and advice clement about E-mail status, reducing unneeded E-mail to a trickle.