In generale un processo (demone) una volta lanciato deve funzionare, punto e basta. Può capitare però che per colpa di qualche bug non ancora risolto, il processo muoia e sia neccessario riavviarlo manualmente.

E' il caso di Spamassassin 3.3.2 con Perl 5.14.2 su Debian Wheezy, ogni tanto il processo spamd muore senza motivo. Nei file di log si trova:

spamd[4563]: prefork: adjust: 10 idle children more than 9 maximum idle children. Decreasing spamd children: 27150 killed.
spamd[4563]: spamd: handled cleanup of child pid [27150] due to SIGCHLD: interrupted, signal 2 (0002)
spamd[4563]: spamd: handled cleanup of child pid [27151] due to SIGCHLD: interrupted, signal 2 (0002)

e poi niente altro (il processo spamd padre è morto). Potrebbe trattarsi di un bug, secondo i seguenti post:

• https://bugzilla.redhat.com/show_bug.cgi?id=812359
• https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6745

Come workaround al problema si installa monit, che semplicemente monitora la presenza del demone spamd e se necessario lo riavvia.

Questo lo snippet di configurazione da salvare in /etc/monit/conf.d/spamd:

 check process spamd with pidfile /var/run/spamd.pid
   group spamd
   start program = "/etc/init.d/spamassassin start"
   stop  program = "/etc/init.d/spamassassin stop"
   if 5 restarts within 5 cycles then timeout
   if cpu usage > 99% for 5 cycles then alert
   if mem usage > 99% for 5 cycles then alert
   depends on spamd_bin
   depends on spamd_rc

 check file spamd_bin with path /usr/sbin/spamd
   group spamd
   if failed checksum then unmonitor
   if failed permission 755 then unmonitor
   if failed uid root then unmonitor
   if failed gid root then unmonitor

 check file spamd_rc with path /etc/init.d/spamassassin
   group spamd
   if failed checksum then unmonitor
   if failed permission 755 then unmonitor
   if failed uid root then unmonitor
   if failed gid root then unmonitor

Poi alcuni parametri sono stati aggiustati in /etc/monit/monitrc:

  set daemon 120            # check services at 2-minute intervals
  with start delay 240      # optional: delay the first check by 4-minutes (by 

  set mailserver localhost

  set alert support@texnet.it                    # receive all alerts

  set httpd port 2812 and
    use address 62.48.51.60  # only accept connection from localhost
    allow 62.48.51.60        # allow localhost to connect to the server and
    allow 62.48.51.8         # allow a remote host to connect to the server and
    allow admin:MySecret     # require user 'admin' with password 'monit'
    allow @monit             # allow users of group 'monit' to connect (rw)
    allow @users readonly    # allow users of group 'users' to connect readonly