[stunnel-users] Zombie problem with stunnel-4.11

[Ole Husgaard]

Hi,

I am trying to set up stunnel to be invoked from xinetd.d to secure pop3
access, and I'm having problems with zombie processes. It looks like
stunnel does not clean up after the pop3 session:
31679 ?        Ss     0:00 xinetd -stayalive -reuse -pidfile /var/run/xinetd.pid
32049 ?        Ss     0:00  \_ stunnel /etc/pop3s/stunnel.cnf
32050 ?        Z      0:00      \_ [ipop3d] <defunct>

The contents of my stunnel.cnf file is:
   CAfile = /var/www/html/ca-cert.crt
   CRLfile = /var/www/html/ca-crl.crl
   cert = /etc/pop3s/pop3s.cert
   key = /etc/pop3s/pop3s.key
   verify = 2
   exec = /usr/sbin/ipop3d
   execargs = ipop3d

I use a non-standard port 997 for the service, so I have the line
   pop3s-eXerp           997/tcp
in my /etc/services file.

The xinetd configuration file for the service is:
service pop3s-eXerp
{
         disable        = no
         socket_type    = stream
         wait           = no
         user           = root
         server         = /usr/sbin/stunnel
         server_args    = /etc/pop3s/stunnel.cnf
         log_on_success += USERID
         log_on_failure += USERID
}

I have searched the net, but have not seen anybody else who have had this
problem.

Does anybody have an idea of what is causing this problem, or what I can do
to work around it?

Best Regards,

Ole Husgaard.