Mechanisms for a multi-homed host to pick the best router
(My apologies, in advance, for the fact that this question is very long winded.) I have a server which is multi-homed to N routers as shown below: +---+ R1---| | | | R2---| | ... | S | | | Rn---| | +---+ This server is a host; it is not a router in the sense that it will never forward any packets (but it might run routing protocols as discussed below). Also, for the sake of simplicity in this discussion, let's say this server only receives inbound TCP connections; it never initiates outbound TCP connections. Finally, this server has a loopback address L. All traffic destined to the server uses address L as the destination address. All N routers have a static route to L and inject that route into their routing protocol (possibly as part of an aggregate). Now, imagine the server receives an inbound connection from another host whose address is A. Thus, the TCP SYN packet which S receives has source address A and destination address L. When the server sends TCP traffic for that same connection back to host A, it needs to pick one of the N routers, in other words, it needs to pick an outbound interface from its N interfaces. Traditionally, this is done by doing a best-match lookup for address A in the forwarding table of the server. One could install a ECMP default route which points to all N routers. In this case, the downstream router would essentially be picked at random (for each connection, assuming 5-tuple hashing). The problem is that some routers are "better" than other routers in the sense that they are closer to the final destination address A. (For example, each router could be connected to a different ISP.) One way for the server to pick the "optimal" downstream router, is to run "stub BGP" between the server and each of the routers. By "stub BGP" I mean that the server uses the BGP session only to learn routes. It advertises its own loopback L, but it never advertises any other routes, and it never propagates and routes from one BGP session to another BGP session. The server would have N BGP sessions and learn the full default-free BGP route table over each of those sessions. In other words, the server would end up with approximately N x 250,000 routes in its RIB and 250,000 routes in its FIB. While this approach would certainly allow the server to pick the optimal downstream router in all cases, I would prefer not to run routing protocols on this server for a number of reasons: - I don't want to the spend memory and CPU on such large RIBs and FIBs. - I'm afraid that other routers will attempt to forward traffic through the server (due to accidental misconfigurations) once it starts participating in the routing protocols. - Since there might be many of these servers (many more than the number of routers) I might end up stretching the routers beyond their scaling limits (number of BGP sessions, link state database size, etc.) and destabilizing the network. - I know there are good open source implementation of routing protocols, but still, I'm nervous that any instability or bugs on the servers could end up screwing up the routers (e.g. persistent BGP flaps). - One possible variation is that the server is a client of some route-reflectors which are not in the forwarding path (i.e. next-hop-self is not enabled). In that case, I might end up needing to do BGP next-hop resolution for a very large number of BGP next-hops. This, in turn, implies that the server might need to also run OSPF in a very large flat area 0. For all these reasons, I don't want to run BGP on the server. Someone suggested an idea to me which seems almost to simple to work, but I cannot find any good reason why it would not work. The idea is "the server simply sends all outbound traffic for the TCP connection out over the same interface over which the most recent TCP traffic for that connection was received". So, for example, if the server receives the SYN from router R3, it would send the SYN ACK and all subsequent packets for the TCP connection over that same interface R3. If the inbound packets for that same TCP connection start arriving from a different router (e.g. because of link failure), say R4, then the server also switches the outbound packets to that same router R4. I am aware that routing is not always symmetrical. In other words, I am aware that the best route from A to Z might be A->B->C->Z while the best route from Z to A might be Z->D->A. However, since the IP routing tables form an inverted tree, it seems to me that in realistic scenarios the traffic should still arrive at A (maybe over a non-optimal path in rare cases) if Z sends the reverse traffic to C instead of D. It seems unlikely (impossible?) that this would cause a routing loop. I can think of the following problems with this approach: (Problem 1) It only works for inbound TCP connections and not for outbound TCP connections. For outbound TCP connections we would not know which router to send the fi
Re: Mechanisms for a multi-homed host to pick the best router
Hi Paul, Thank you very much for the confirmation that the idea is sane and for the pointers to the additional information. -- Cayle On Wed, Sep 17, 2008 at 11:49 PM, Paul Vixie <[EMAIL PROTECTED]> wrote: > [EMAIL PROTECTED] ("Cayle Spandon") writes: > > > (My apologies, in advance, for the fact that this question is very long > > winded.) > > np. > > > I have a server which is multi-homed to N routers as shown below: > > > > +---+ > > R1---| | > > | | > > R2---| | > > ... | S | > > | | > > Rn---| | > > +---+ > > > > This server is a host; it is not a router in the sense that it will never > > forward any packets (but it might run routing protocols as discussed > below). > > > > Also, for the sake of simplicity in this discussion, let's say this > server > > only receives inbound TCP connections; it never initiates outbound TCP > > connections. > > > > Finally, this server has a loopback address L. All traffic destined to > the > > server uses address L as the destination address. All N routers have a > > static route to L and inject that route into their routing protocol > > (possibly as part of an aggregate). > > > > Now, imagine the server receives an inbound connection from another host > > whose address is A. Thus, the TCP SYN packet which S receives has source > > address A and destination address L. > ... > > For all these reasons, I don't want to run BGP on the server. > > "too many moving parts." > > > Someone suggested an idea to me which seems almost to simple to work, but > I > > cannot find any good reason why it would not work. > > > > The idea is "the server simply sends all outbound traffic for the TCP > > connection out over the same interface over which the most recent TCP > > traffic for that connection was received". > > > > So, for example, if the server receives the SYN from router R3, it would > > send the SYN ACK and all subsequent packets for the TCP connection over > that > > same interface R3. > > ... > > right idea. works great. see the following: > > http://www.academ.com/nanog/feb1997/multihoming.html > http://www.irbs.net/internet/nanog/9706/0232.html > http://gatekeeper.dec.com/pub/misc/vixie/ifdefault/ > > > ... > > I can think of the following problems with this approach: > > > > (Problem 1) It only works for inbound TCP connections and not for > outbound > > TCP connections. For outbound TCP connections we would not know which > router > > to send the first SYN packet to. > > you said above you only needed inbound. for outbound and udp: round robin. > > > ... > > My question for the NANOG community are these: > > > > (Question 1) Can you think of any additional problems with this approach? > > Specifically, I am interested in persistent failures in the absence of > > topology changes. > > topology change frequency is in a different order of magnitude than the > usual tcp session startup frequency, so unless you've got long running tcp > sessions which won't restart on a connection reset, you've got no problem, > and if you do have that kind of tcp session, you've already got problems. > > > (Question 2) Is there another mechanism for the server (a multi-homed > host) > > to pick a best router, short of running stub BGP? Are there any standards > > for this? > > there are a bazillion patented and/or ubersecret ways to do this. noone > has > ever demonstrated anything that works better than an undercommitted network > with undercommitted connections to other undercommitted first-hop networks. > > > (Question 3) If the answer to question 2 is "no", is there any interest > > in standardizing a protocol for a multi-homed host to pick a best > > next-hop router? One could think of this is a host-to-router routing > > protocol. One might call the existing routing protocols router-to-router > > protocols (because I think we are abusing them by running them on > > hosts). This is somewhat analogous to the multicast routing world where > > we use different protocols for router-to-router multicast (PIM) versus > > host-to-router (IGMP). > > sadly, this has been tried, but it always runs into least-cost routing > issues > whereby not only the predicted connection quality but also contract details > like whether this is over or under the per-period minima and how many > quatloos > per kilosegment it will cost all have to get exchanged at high speed with > low > latency and good accuracy. been there, did that, got no useful t-shirt > even. > -- > Paul Vixie > >
Re: Mechanisms for a multi-homed host to pick the best router
Hi Laurence, RE> why would you not sent the reply out the same spigot you go the request on? Yes, that exactly what I was trying to ask in the e-mail (in a much more verbose way than you :-). The problems I could think of are: - It only works for inbound TCP connections. - The TCP connections might be dropped after a topology change despite the existence of an alternative path. I was wondering if anyone else knows of any additional problems. -- Cayle
TRIP deployment?
I'm not sure if this is the right mailing list for this question: how widely is TRIP (Telephone Routing over IP [RFC3219]) deployed / used in current networks?
BGP on Mac OS X?
Does anyone known which open source BGP implementation I can get running on Mac OS X Leopard with a minimum of fuss? This is for experimentation only (not for a production environment) so I am not too concerned about scaling and performance. If any tweaking is needed to get it to compile / run on OS X, a pointer to a website with instructions would be highly appreciated. -- Cayle.
Re: Point to Point Ethernet
I frequently run into scenarios where two devices (two routers, or a router and a host) need a point-to-point connection to each other with a capacity of (much) more than 10 Gbps. For cost reasons, Ethernet is often used. Since more than 10 Gbps is needed, we end up with multiple parallel 10GE point-to-point connections. Because the devices often don't support LAG or have limitations on the number of links in a LAG, we often cannot use LAG at all or cannot put all 10GE links in a single LAG group. So, we end up with multiple parallel layer-3 point-to-point connections where each connections is either an Ethernet or a LAG group. Furthermore, in order to conserve IP addresses, there is a desire to make these interfaces unnumbered. The involved devices have a numbered loopback interface whose address is used as the "donor" for the unnumbered Ethernet / LAG interfaces. Most router vendors already support unnumbered point-to-point Ethernet, see for example: http://www.juniper.net/techpubs/en_US/junos9.5/information-products/topic-collections/config-guide-network-interfaces/interfaces-configuring-an-unnumbered-interface.html#id-10432956 However, there are some interoperability issues / open questions related to point-to-point unnumbered Ethernet, see for example: http://forums.juniper.net/jnet/board/message?board.id=JUNOS&message.id=130 http://forums.juniper.net/jnet/board/message?board.id=switch&thread.id=835 I would be very interested in some standards (i.e. an IETF BCP) to describe the best current practices for these applications of Ethernet. I am not particularly interested in re-inventing a new flavor of Ethernet for this application. All that is needed, in my opinion, is some clarifications or best practices on how to use the existing standards to create point-to-point unnumbered Ethernet connections. PS -- I am also aware of some esoteric BRAS applications of Ethernet where one side is numbered and the other side is unnumbered.