You probably don't need to configure these files. They have to be correct, otherwise things won't work, but the standard configurations (from recent distributions) are all pre-configured.
/etc/services lists the TCP and UDP port
numbers used by different services. Sendmail for
example, will be listening on port 25 because it
smtp port by name, which when
looked-up by the software becomes
... many other lines ... smtp 25/tcp mail finger 79/tcp www 80/tcp http # WorldWideWeb HTTP www 80/udp # HyperText Transfer Protocol
Since sendmail does use /etc/services, when I want to
reboot my machine without starting sendmail, I
simply edit /etc/services, and set the
to 2525. Then it listens on the wrong port, and doesn't
receive (or bounce) any email. The sender simply gets
a service not available, and retries later.
You might have to edit /etc/inetd to switch between daemons that are always running, and daemons that get started (by inetd) on the first request over the network. RTFM inetd(8) tcpd(8)
finger stream tcp nowait daemon /usr/sbin/tcpd /usr/sbin/in.fingerd systat stream tcp nowait guest /usr/sbin/tcpd /usr/bin/ps -auwwx netstat stream tcp nowait guest /usr/sbin/tcpd /bin/netstat -f inet
For example, the finger line, means that on my system, ther isn't
a finger daemon running all the time, but if you call me on that
port (as listed in /etc/services),
inetd will start up
in.fingerd for that connection. That saves memory and
gives a slightly faster boot-up time.
/etc/inetd also contains "wrappers", that can be used to log every connect to your box, and authenticate who the remote caller is.
After inetd, individual services have their own permission control and configuration files, such as:
There are several of these, but you probably don't have to do anything just yet, as the defaults are usually reasonable. At least not until you notice that that particular service isn't running as you expected.
Though of course you have to go through the list, to see what your system is and isn't doing.