This is a discussion on Iptables checksum question within the Linux Networking forums, part of the Linux Forums category; Hi, Iptables question: we've managed to get ip tables working on the ingress router to the extent that it ...
|
|||||||
| FAQ | Members List | Calendar | Search | Today's Posts | Mark Forums Read |
|
|||
|
Hi,
Iptables question: we've managed to get ip tables working on the ingress router to the extent that it modifies QOS bits on the IP header as desired if the destination port is a match to the iptables command. We'd like to have this work both ways - i.e. put communiction over a particular port in a special diffserv class. However, when we add the same iptables command to the egress router, the checksum is incorrect when it arrives at the end host (Ethereal tell us this). Question: what are we doing wrong? Is this a bug in iptables, or more likely a lack of understanding on our part? Any answers/help much appreciated. Best Regards, Sam90 |
|
|||
|
samhunt90@hotmail.com (Sam) wrote in message news:<e20518e0.0407261730.347d1835@posting.google. com>...
> Hi, > > Iptables question: we've managed to get ip tables working on the > ingress router to the extent that it modifies QOS bits on the IP > header as desired if the destination port is a match to the iptables > command. We'd like to have this work both ways - i.e. put communiction > over a particular port in a special diffserv class. However, when we > add the same iptables command to the egress router, the checksum is > incorrect when it arrives at the end host (Ethereal tell us this). > > Question: what are we doing wrong? Is this a bug in iptables, or more > likely a lack of understanding on our part? Any answers/help much > appreciated. > > Best Regards, > Sam90 I still don't have an answer - however, I think I can avoid the packets from getting processed twice by iptables (on in each router) simply by specifying the interface, i.e., they should only be processed by the ingress router, and no other. Hopefully that will do the trick. Sam90 |