thank you for so rapid answer.
 
i still can't understand . in the host  when we add a tap device 192.168.0.1, uml set 192.168.0.2
 
after that we can reach 192.168.0.2 ? we don't need to add route .  is that i  understand this as  if we add a tap device , just think that tap device is a vitural 
 
network interface of the host ?? i have taken a tumble.
 


 
2006/6/19, Stefano Melchior <stefano.melchior@openlabs.it>:
On Mon, Jun 19, 2006 at 02:58:14PM +0800, paul john wrote:
Dear paul john,
> i just wondering about the network setup .  as i know , if the host act as a
> gate way it also should has an internal IP  so it can ping the umls .
>
> if use a bridge ,  the bridge should be set of the external ip , and the
> umls will be setup the internal ip , and umls will use tapN,
>
> but now in the host  if ping umls  obversily can't reach it (is that right
> just add a route to the host , umls can be reached ?
>
> like that , tap0 address is 192.168.0.1,  uml1 address is 192.168.0.2 , and
> in uml1 set network as  ip 192.168.0.2/24 gateway is 192.168.0.2
>
> and add route in host like that   addroute 192.168.0.0/24 tap0 [this command
> maybe wrong , you just know the meaning]
>
> is that right ?

pratical example of conf:
Host
====
iface tap0 inet static
address 192.168.0.1
netmask 255.255.255.0
network 192.168.0.0
pre-up tunctl -u <your-user>
pre-up bash -c 'echo 1 > /proc/sys/net/ipv4/ip_forward'
pre-up iptables -t nat -A POSTROUTING -o <your-internal-interface> -j MASQUERADE
post-down tunctl -d <your-user>

Uml
===
auto eth0
iface eth0 inet static
address 192.168.0.2
netmask 255.255.255.0
up route add -net 0.0.0.0 dev eth0
up route add -net 0.0.0.0 gw <your-public-IP>
>
>
> ).   i just don't understant  how tap works . so if you give me a practical
> setup ,will be very appreciated.

tap is a p-t-p connection between host and uml.
>
> also "exec linux ubd0=... umid=mytest eth0=..." , if edit the .bash_profile
> like that ,  it means the user can't login the host  ? as he login the host
> ,will be redirect to uml ?

yes, it is automatically "redirect" to uml image that is launched after
the login on host: I mean that you execute the login on host, then once
authenticated the first thing the the system does is to launch the uml
image (with that command). Then you will be on the uml.

> my understood is that , if the user login just start a uml instance for him
> . but if the uml instance is already there , how should i do ?

it was only an alternative way to reach it. You can somehow get the
/dev/pts/X serial connection on uml and then connect on the previously
launched uml by using screen or minicom.
 
this i still can't understand . if i already setup umls , and i want they login to redirct to the umls (assume the uml is 192.168.0.2),
"exec linux ubd0=... umid=mytest eth0=..." ,this cann't do that ,
may add a command in the users .bash_profile that ,ssh 192.168.0.2 when he login ,the first command is loging in the uml 192.168.0.2?
 

> thank you for your help .
Cheers

SteX
--
Stefano Melchior, GPG key = D52DF829 - < stefano.melchior@openlabs.it>
http://etinarcadiaego.dyndns.org    --     http://www.stex.name
Skype ID "stefanomelchior"


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (GNU/Linux)

iQCVAwUBRJZPl4lnkxrVLfgpAQIjPAQAoxuub9i9e77BaCkKY1qNFtor4kYKzuEg
i1fv1CH/I+dFg7qLkADb3kggIrOD6u8afCam3h+eEp4TPQmH5yKbHf4wcoIgbvOv
SYaj9rEO+S1WQftjAC5hip1JTfMu4xtfNjjCo/PYRIYIhXpp7aOc4K9y2D0bqPEO
Gdy8RIg/bwg=
=cc8P
-----END PGP SIGNATURE-----