Messages in this thread | | | Date | Sat, 26 Dec 1998 12:21:34 -0500 (EST) | From | Tom Oehser <> | Subject | Re: MTU/MRU, always-defrag, masq not working for some URLs |
| |
However, they *DO* work over the MTU/MRU=296 PPP link.
They just don't work over masquerading.
Meaning, the fragments are coming through OK, and *my* kernel is *supposed* to "ALWAYS_DEFRAGMENT", which is *supposed* to mean (as I understand it), that if it works at all (over 296), it should work (over the 1500 eth0 forwarding). Maybe it does defragment to *larger* than 1500, and then just doesn't *re-* fragment it down to under 1500? I dunno until I spend some time in tcpdump and ip_input.c or whatever, I guess...
> In my experience this is most often caused by misconfiguration at the > "other end" of your connection. meaning: nothing much you can do about it; > some sysadmin has set up a firewall between you and the webserver which > doesn't allow the ICMP unreachable messages to reach the webserver. > > Only thing you can do is a) set your MRU to 1500 so the problem doesn't > occur or b) contact affected websites on a case by case basis and get them > to fix their setup by EITHER alloing ICMP unreachable packets to reach > their webserver OR turning off MTU Path discovery on the websever (for > linux it's a compile-time option of the kernel).
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |