Package: linux-image-3.0.0-rc4-amd64 Severity: normal Tags: experimental
-- System Information: Debian Release: squeeze APT prefers proposed-updates APT policy: (500, 'proposed-updates'), (500, 'unstable'), (500, 'stable'), (1, 'experimental') Architecture: amd64 (x86_64) Kernel: Linux 2.6.38-2-amd64 (SMP w/1 CPU core) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Hello, I'm noticing a problem with clients behind my linux system that acts as a NAT gateway. When doing a traceroute from the client, to say, google.com, the first hop is showing up as the final destination rather then the IP address of the gateway. ----- traceroute to 74.125.224.48 (74.125.224.48), 64 hops max, 52 byte packets 1 74.125.224.48 (74.125.224.48) 0.658 ms 0.106 ms 0.123 ms 2 boid-dsl-gw07-199.boid.qwest.net (184.99.64.199) 42.359 ms 41.447 ms 42.161 ms 3 boid-agw1.inet.qwest.net (184.99.65.49) 41.549 ms 40.989 ms 42.573 ms 4 sea-edge-12.inet.qwest.net (67.14.41.22) 86.234 ms 57.025 ms 55.848 ms 5 65.122.121.66 (65.122.121.66) 58.344 ms 56.292 ms 55.848 ms ----- Proper traceroute from unaffected kernel (in this case, 2.6.38-2) ----- traceroute to 74.125.224.48 (74.125.224.48), 64 hops max, 52 byte packets 1 gateway (10.11.1.1) 1.431 ms 0.458 ms 0.445 ms 2 boid-dsl-gw07-199.boid.qwest.net (184.99.64.199) 42.727 ms 43.713 ms 43.292 ms 3 184-99-65-49.boid.qwest.net (184.99.65.49) 43.442 ms 46.488 ms 43.461 ms 4 sea-edge-12.inet.qwest.net (67.14.41.22) 59.208 ms 58.577 ms 56.627 ms 5 65.122.121.66 (65.122.121.66) 56.779 ms 59.540 ms 121.465 ms ----- Same exact firewalling rules are in place with iptables on both traceroutes (SNAT, not messing with traceroutes, icmp in/out). Even with a completely clean boot with no firewalling rules and no nat rules, traceroute shows exactly the same thing, first hop shows up as the destination. So, I'm not entirely sure its related to netfilter. I do have packet dumps from the tests if they might be useful. I can confirm that 2.6.32-5 kernels from squeeze also do not exhibit this issue. Thanks for looking into this! -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110628202442.6059.37176.report...@noc.lulz.nu
