kre@cs.mu.OZ.AU (Robert Elz) (01/21/91)
I sent a comment to Paul Tsuchiya wrt his internet draft (draft-tsuchiya-subnetnos-00.txt) on subnet number allocation policies. This doesn't touch on the main subject of his draft, but a side issue he mentions. Paul says (in mail to me) ... Could you put your message on all of tcp-ip, so we can see what other people have to say about it? So, here it is ... In there you say (wrt two subnets on one cable) ... In fact, this is not such a bad solution, because assuming that the gateway is capable of recognizing multiple subnet numbers on the same subnet, the gateway will simply send the host an ICMP Redirect [4], and subsequent packets will go directly to the host [1]. I don't think that can be true can it? That would require the ICMP redirect to contain an ethernet address. The sending host is in no doubt of the destination's IP address, sending a redirect that contains that address can do no more than confuse it, if its routing table has it believe that to reach that address it must route through a gateway. Oh - do you mean send the host a redirect, containing its own address as the gateway? I guess that might work, assuming that the host's software understands the BSD type convention "if I am the gateway, I send directly out on my ethernet", if not, almost anything might happen. If that's it, I think I'd be more explicit about it. To provide a little context - this is related to a cable being used to carry two subnets, with a router somewhere on the cable configured to forward between the two - and perhaps send redirects to hosts sending through the router, when, with a little more intelligence, the host would be able to send directly to the destination. In a reply to that message, Paul indicated that the first of my two scenarios was the one he intended, with the idea that on receiving a redirect, (containing the hosts IP address as the gateway to send through) the host receiving the redirect would not look inside it (much), but would simply ARP for the "gateway" address drom the ICMP, and end up sending directly to the destination. Personally, I would expect that a host receiving a redirect like that (if any gateway sent one) would simply ignore it - as a defence against bogus redirects. Anyway - comments? kre
ejm@ejmmips.NOC.Vitalink.COM (Erik J. Murrey) (01/22/91)
In article <6490@munnari.oz.au>, kre@cs.mu.OZ.AU (Robert Elz) writes: > In there you say (wrt two subnets on one cable) ... > > In fact, this is not such a bad solution, > because assuming that the gateway is capable of recognizing multiple > subnet numbers on the same subnet, the gateway will simply send the > host an ICMP Redirect [4], and subsequent packets will go directly to > the host [1]. > > I don't think that can be true can it? That would require the ICMP > redirect to contain an ethernet address. The sending host is in no > doubt of the destination's IP address, sending a redirect that contains > that address can do no more than confuse it, if its routing table has > it believe that to reach that address it must route through a gateway. > > Oh - do you mean send the host a redirect, containing its own address > as the gateway? I guess that might work, assuming that the host's > software understands the BSD type convention "if I am the gateway, > I send directly out on my ethernet", if not, almost anything might > happen. > > If that's it, I think I'd be more explicit about it. > Actually, using two subnets on a cable is very common when more than one link-layer protocol is being used. (i.e. Ethernet v2 and 802.3/SNAP) In this case, an ICMP isn't even practical since we don't have a way to specify link-layer information in an ICMP message. In cases where the subnets share the same frame type, it is still impractical to send ICMP's since the "offending" host will know nothing about host specified in the ICMP redirect. Most IP implementations based on the 4.3Tahoe stack only allow one IP address per interface. Maybe the interface handling code could be extended to support multiple addresses per protocol? --- Erik J. Murrey Vitalink Communications NOC ejm@NOC.Vitalink.COM ...!uunet!NOC.Vitalink.COM!ejm
raj@hpindwa.cup.hp.com (Rick Jones) (01/23/91)
Instead of relying on redirects, why not use proxy-ARP? It would probably be more robust in the face of various implementations. If those hosts that can use it, and the router routes for those who can't, then all should be well. Or have I missed something? rick jones ___ _ ___ |__) /_\ | Richard Anders Jones | This space undergoing | \_/ \_/ Hewlett-Packard Co. | renovations - results soon ;-) ------------------------------------------------------------------------ Being an employee of a Standards Company, all Standard Disclaimers Apply
jbvb@FTP.COM (James B. Van Bokkelen) (01/23/91)
To work efficiently in "multiple IP nets or subnets on one cable" situations, you can't be discriminating about ICMP redirects. This opens up the possibility of bogus redirects taking a host off the net. Currently the world seems willing to accept the risk to get the convenience. James B. VanBokkelen 26 Princess St., Wakefield, MA 01880 FTP Software Inc. voice: (617) 246-0900 fax: (617) 246-0901