xinetd: Is it a good idea to use?

[Francisco Romero]

Seems like when I get the hang on RedHat's Linux, something new comes
up. For Example, xinetd. Do you guys think that xinetd is worth working
with?

[Hal Burgiss]

On Sun, 29 Apr 2001 07:51:07 GMT, Jan Johansson <spamf...@mupp.net>
wrote:

>>Seems like when I get the hang on RedHat's Linux, something new comes
>>up. For Example, xinetd. Do you guys think that xinetd is worth
>>working with?
>

>Lightyears better then inetd.

Agreed. This is why new distro versions come out. Improvements! Anyone
who prefers inetd, when xinetd is already installed, is just being too
lazy to consider the differnces.

--
Hal B
h...@foobox.net
h...@burgiss.net
hbur...@bellsouth.net
Spamtrap: u...@ftc.gov and rep...@fraud.org
--

[Brian Aust]

Would you mind explaining in your own words what those great differences
are?

I'm switching from 6.2 to 7.1, and can see that with inetd, each service is
allowed one single line, which you can comment out or not. With xinetd.d,
each service now gets a whole text file in that directory, so you can
include a bunch more lines, probably allowing for parameters or scripts or
whatnot.

Is this fairly close? What are the primary benefits one can implement by
having a whole file in inetd.d dedicated to a service instead of a single
line in inetd.conf?

Also, I've heard that to disable a service, you just remove it from the
xinetd.d directory. But where do you put it? Killing it would make it a
PITA to recreate if you ever wanted to enable the service again. Do people
simply create another directory to store unused xinetd.d service files? Or
does one simply modify the line in each service file called "disable =
(yes/no)" ?

Thanks much!!
Brian

"Hal Burgiss" <h...@burgiss.net> wrote in message
news:slrn9eoju...@localhost.localdomain...

[Hal Burgiss]

On Sun, 29 Apr 2001 22:45:59 GMT, Brian Aust <ba...@healthdec.com>
wrote:

>Would you mind explaining in your own words what those great
>differences are?

Greater flexibilty and configuration. Especially for limiting
connections, including such features as time of day limitations. More
flexibility in logging. Built in libwrap support. Chkconfig support.
These are what I notice and appreciate ...

Most of us probably would not notice the difference of one to the other
for just basic services.

>I'm switching from 6.2 to 7.1, and can see that with inetd, each service is
>allowed one single line, which you can comment out or not. With xinetd.d,

You can turn off/on with chkconfig:

chkconfig --del rsh

If 'easy' is important, this is easier than editing a text file IMO.

>each service now gets a whole text file in that directory, so you can
>include a bunch more lines, probably allowing for parameters or scripts or
>whatnot.
>
>Is this fairly close? What are the primary benefits one can implement by
>having a whole file in inetd.d dedicated to a service instead of a single
>line in inetd.conf?

Greater flexibility for one. Also, I find it much easier to read what
each option is when it is 'option = something'.

>Also, I've heard that to disable a service, you just remove it from the
>xinetd.d directory. But where do you put it? Killing it would make it a

See above. Nonsense. You can also set the disable option to 'yes' (which
is what chkconfig really does).

>PITA to recreate if you ever wanted to enable the service again. Do people
>simply create another directory to store unused xinetd.d service files? Or
>does one simply modify the line in each service file called "disable =
>(yes/no)" ?