Re: Are iptables rules absolutely neccessary?
From: Ed Murphy (emurphy42_at_socal.rr.com)
Date: 10/18/03
- Next message: Michael Munro: "Re: Debian network problem"
- Previous message: Gregory L. Hansen: "Re: Problems with .forward."
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Sat, 18 Oct 2003 03:15:49 GMT
[crosspost to c.o.l.networking added]
On Fri, 17 Oct 2003 21:59:28 -0400, Ben Logan wrote:
> If your MTU size is
> smaller on the Linux client than on the gateway, every packet will be
> fragmented. This wouldn't be a problem except that some webserver admins
> erroneously block all ICMP messages (in order to avoid things like the
> ping of death) and this prevents the server from sending a message back to
> your client to drop the MTU size.
I have this in my iptables. Is it erroneous?
-A RH-Lokkit-0-50-INPUT -p tcp -m tcp --tcp-flags SYN,RST,ACK SYN -j REJECT --reject-with icmp-port-unreachable
I had the opposite of the OP's config: RH9 acting as a gateway for
Win2K Pro. Worked fine for several weeks, then abruptly quit a couple
of days ago. I hadn't changed iptables nor the MTU size, and I don't
*think* anything had changed on the Windows side, but you never know...
Is this line (from the *nat section) erroneous?
-A POSTROUTING -o eth0 -j MASQUERADE
eth0 connects to the outside world; eth1 connects to the crossover
cable. I'm pretty sure this hadn't changed.
RH9 can ping the outside world and Windows; Windows can ping RH9,
but cannot nslookup or otherwise see the outside world.
Any and all suggestions welcome.
- Next message: Michael Munro: "Re: Debian network problem"
- Previous message: Gregory L. Hansen: "Re: Problems with .forward."
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|
|