Date: Wed, 8 Sep 1999 22:11:01 -0700 (PDT) From: Joseph Lee <nugundam@nerv.nu> To: "D. Alex Neilson" <neilson@www.nugate.com>, "Nick A. Fikouras" <nick@dcs.shef.ac.uk>, freebsd-questions@freebsd.org Subject: Re: Static Route, need help (was: Re: Static Route (Correction)) Message-ID: <Pine.BSF.4.10.9909082150570.7622-100000@greenwood3.nerv.nu>
next in thread | raw e-mail | index | archive | help
nick@dcs.shef.ac.uk wrote on Mon, 02 Nov 1998 12:07:22 +0000: >D. Alex Neilson wrote > >> With this configuration, to get A to talk to B, one could say on A >> >> route add -host 192.168.67.67 -interface vx0 >> >> ------- >> | | >> | A | >> | | >> ------- >> | 192.168.66.66 (vx0) >> | >> ------------------------------------- ethernet (multiple nets) >> | >> | >> | 192.168.67.67 (le0) >> ------- >> | | >> | B | >> | | >> ------- >> >> I get this route, but the gateway ethernet address is for host A: >> >> Destination Gateway Flags Refs Use Netif >> 192.168.67.67 0:60:8:4:4:ed UHLS 0 0 vx0 >> >> Of course, it doesn't work; strangely, if I try telnetting from B to A, >> it'll pause for a moment, then let me in, with A giving this message >> >> myhost /kernel: arp: 192.168.67.67 moved from \ >> 00:60:08:04:04:ed to 08:00:20:73:87:89 >> >> and A now shows the route I wanted in the first place: >> >> Destination Gateway Flags Refs Use Netif >> 192.168.67.67 8:0:20:73:87:89 UHLS 1 55 vx0 >> >> Is there a way to get the latter route in the first place? >> >> Thanks, >> >> Alex >> [snip] >I have had the same problem, this is what I've found out from my experience. >The main problem is that whenever a static route to a host is created, for no >apparent reason, an ARP entry is created for the particular host indicating as >its MAC address the address of the interface that was passed as argumen in the >'route add -host' command (!). What makes things even worse is that static >links seem to create permanent arp entries. You can verify that by doing arp >-a. In my case I had two machines that could not exchange traffic due to the >folly arp table. All traffic transmitted by a node was destined at the link >layer for itself!!!!. I checked with a network analyser. >The way I've overcome this is, I've created I a file with all the proper arp >entries and I load it up manually after booting has completed (check arp >manual page for this). I know this not the most scientific way of doing it, if >anybody has any suggestions I'm willing to change. > >hope I've been of any help, > >nick I've found a solution short of hacking route itself. It's: route add -iface <ip> -link <interface>:<ether address> Or to restate the original example: route add -iface 192.168.67.67 -link vx0:8:0:20:73:87:89 -- Joseph nugundam =best=com==/==\=IIGS=/==\=Playstation=/==\=Civic HX CVT=/==\ # Anime Expo 1999 >> www.anime-expo.org/ > # FreeBSD: The Power to Serve >> www.freebsd.org > # EX: The Online World of Anime & Manga >> www.ex.org/ / To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.10.9909082150570.7622-100000>