Bart Vetters grabbed a keyboard and wrote:
>
> > I want to use tcp_wrapper to provide an extra layer of control to those
> > services which I allow, but I'm not sure how to go about doing this.
>
> There's two ways. Well, not really two ways, you'll see what I mean.
>
> - It is no longer necessary to use TCP wrappers with the tcpd daemon. All
> of the network daemons installed with Mandrake 8 come with tcp wrappers
> support compiled in (using the tender services of libwrap I believe).
> This means that these daemons will check hosts.allow and hosts.deny
> without intervention of tcpd.
> A caveat: I am not a Mandrake employee and currently do not have access
> to a Mandrake 8 machine to check (a simple ldd of in.telnetd should
> confirm or deny), so I'm not 100% sure the above is correct, and in a
> security related matter less than 100% is not always enough :).
I did a "ldd /usr/bin/in.ftpd" but didn't see anything in the output which
suggested that tcpd was part of it. I did not see anything called libwrap,
either. Is there something else I should be looking for?
> - For those with fond memories of tcpd, you can still call it from xinetd
> as such:
>
> server = /usr/sbin/tcpd
> server_args = in.telnetd
>
> Another caveat: I'm not sure of the syntax of the above, and have as
> mentioned no access to a LM8 machine, but I seem to remember the above
> being the correct options. Anyway, man xinetd should provide more help.
I'll try digging through the man page again and see if I can find what I
missed when trying to look this up. I tried chaging as above to have it
read:
server = /usr/sbin/tcpd
server_args = /usr/sbin/in.ftpd -l -a
(the "-l -a" was in that field originally, so I left 'em in). Restarted
xinetd, did a telnet to localhost 21 to see if it would let me in, and
while I did connect, I never got the greeting message from wu-ftpd. Also,
I noticed that I kept getting short bursts of LOTS of disk activity after
that, even after I closed the connection. Even after putting
/etc/xinetd.d/wu-ftpd back to normal and restarting xinetd again. Started
top up and found that there was now a running tcpd process that was sucking
up lots of CPU time. :-) Killed it and the disk activity went away. Beats
me what all the disk activity was about; I can't see anything in /var/log,
and I had a tail running on syslog, so it wasn't doing stuff there.
Thanks for the reply. Hopefully, you've got me pointed in the right
direction. :-) With a little luck, I might find whatever it was that I
missed when looking in the xinetd man page last time....
--Dave
--
David Guntner GEnie: Just say NO!
http://www.akaMail.com/pgpkey/davidg or key server
for PGP Public key
Want to buy your Pack or Services from MandrakeSoft?
Go to http://www.mandrakestore.com
