RE: a stupid idea, a stupid test
I can't comment on the tunneling but I have a similar problem with dhcp.
The linux dhcpcd will give back the ip-addr on shutdown. At my site (with
something like 1500 pc's, 500 printers, etc.) wins is used for the m$ desktops.
My linux box did not behave like the nt systems. What I eventually did
was to replace the "stop" case in /etc/init.d/dhcpc with
"killall -9 dhcpcd". Unless I leave the box turned off for a long time I
always get the same ip-addr back. I just can't leave it off when the lease
expires on the ip-address. You can put the ip-addr in the host file and
get by for a while...
jim
----------
From: Stephen Carpenter[SMTP:sjc@delphi.com]
Sent: Thursday, March 26, 1998 12:26 AM
To: debian-user@lists.debian.org
Cc: The recipient's address is unknown.
Subject: a stupid idea, a stupid test
Ok...I got a bright idea earlier....
I have a linux machine at work (where we don't really have linux
machines...its one of the
3 that I know of in existance on our entire network (of at least 10 000
users) )
Anyway...mine is on DHCP like most of the network and a new one which
was setup
is also on DHCP (mostly because im a tech and he is at the helpdesk and
neither of us
can justify why we need network engineering to give us static IPs)
One of the Linux machines I know of has a static IP (this guy is some
important doctor
and I am sure I coul dget him to let me use his linux machine or one of
his slowaris runnin sun stations
if I asked )
Here was my idea:
I want my machine and my friend at the helpdesks computer to be able to
communicate
(maybe even share soem NFS mounts ;) )
The problem is we are on DHCP...dynamic adressing which seems to change
at least every few days
My idea was ...can I setup a "virtual network" layered on top of our
ethernet / TCP/IP network?
My idea was to use netcat and pppd to make a ppp connection through a
tcp/ip socket...
then assign my own Private IP adresses (10.*) to the ends of the ppp
connection
then I could have the machine with the static IP as a central hub for
the other 2 systems to connect too
thus the other two machines could have "Static IPs" for talking to
eachother
ok...I know there must be a better way to do this...I think its called
tunneling??
and I believe there is some kernel level support fo rit...
I wanted to find a way to do it my way...
the idea was this:
use netcat on both systems to open the equivalent of a "pipe" and then
attach pppd to it
I tried this on the local machine...maybe someone can say why this setup
didn't work
first I made 4 named fifos's
in1 out1 in2 out2
then
cat in1 | nc -l -p 1555 | cat out1
cat in2 | nc 127.0.0.1 1555 | cat out2
...both connected....and netstat showed a conenction between the two
then I tried this...
I cated a file into in1
then I did "cat out2" on another VT
nothing came out... I would have thought the end result would have
been...
data in -> in1 -> nc -> tcpIP -> nc -> out2 -> data out
any ideas why did didn't work
(suffice it to say "pppd <out1 >in1" and the same for in2 and out2 ...
they didn't connect....
any ideas? (and yes I know this is a stupid way to do it but....
I would think it should work! )
-Steve
--
To UNSUBSCRIBE, email to debian-user-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
--
To UNSUBSCRIBE, email to debian-user-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: