On Fri, Oct 06, 2000 at 12:34:00AM +0200, Thomas Voss wrote:
> Anyway, even if bind would run on the firewall box, the problem would
> remain the same, i.e. bind would send a UDP packet which has to bring up
> the line (forcing a new IP for the interface), and which therefore leaves
> with the wr
Hi,
JLF> Maybe I'm missing the point here, but why do you think you need
JLF> to MASQ these packages? When a box from your internal network
JLF> do a lookup, it checks with BIND on your boundary/firewall box.
and exactly that's the point: There is no bind running on my firewall box.
Bind is
On Tue, Oct 03, 2000 at 03:37:00PM +0200, Thomas Voss wrote:
> Does anybody has an idea about that?
Maybe I'm missing the point here, but why do you think you need to
MASQ these packages? When a box from your internal network do a lookup,
it checks with BIND on your boundary/firewall box. BIND
Hello Phil,
PB> > the UDP packet is masqueraded
PB> > correctly and triggers the PPP dial-out to my ISP. But
PB> > finally, the UDP packet gets dropped out there because no
PB> > address rewriting is done for UDP packets
PB> If no address rewriting is done you need to check your ipchains
P
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
A long time ago, in a galaxy far, far way, someone said...
>
> Hello,
>
> I have a linux box (Debian 2.2, kernel 2.2.17) running as an ISDN dial-on-
> demand gateway to my ISP. The ISP is assigning dynamic IP adresses, and I
> have address rewriting
Hello,
I have a linux box (Debian 2.2, kernel 2.2.17) running as an ISDN dial-on-
demand gateway to my ISP. The ISP is assigning dynamic IP adresses, and I
have address rewriting enabled (echo 2 > /proc/sys/net/ipv4/ip_dynaddr).
UDP packets from my internal network arriving for port 53 of the NS
6 matches
Mail list logo