nanog mailing list archives
Re: Anycast 101
From: Joe Shen <joe_hznm () yahoo com sg>
Date: Mon, 20 Dec 2004 21:53:49 +0800 (CST)
I don't think PPLB is compatible with anycast esp. in situation when we consider end-to-end communication with multiple packets. As PPLB may derive to out-of-sequence between TCP pacekets & different DNS server destination of the same UDP stream, it will broke anycast DNS service in some situation. So, if TCP based DNS requests is considered, flow-based load balancing should be considered which is total differnt from PPLB. Joe --- Iljitsch van Beijnum <iljitsch () muada com> wrote:
On 18-dec-04, at 22:31, Paul Vixie wrote:i would be interested in hearing from anybodyelse who thinks thatturning on pplb in a eyeball-centric isp that hasmultiple upstreampaths is a reasonable thing to do, even if therewere no anycastservices deployed anywhere in the world.so far, no takers. i've heard from rfc-writerswho say pplb was nevermeant to be used the way Iljitsch is describingit, and i've heard fromequipment vendors who say their customers don't dothat and that ifsome customer did that and asked for support theresponse would be "don't dothat!", and i've heard from network operators whosay they would neverdo that, and i've heard from customers of networkoperators who did thatwith notable bad effects.but so far nobody has said "yes, what Iljitsch isdescribing shouldwork."Apparently you also didn't get any pointers to RFCs or other authoritative sources that say "each and every packet injected into the internet must be delivered in sequence". You feel you get to decide what other people should and shouldn't do. I find that dangerous. As long as there is no standard or law that says something can't be done, people are free to do it. Apart from that, I'm not convinced per packet load balancing is as bad as people keep saying. In the absense of any research that I know of, my position is that per packet load balancing does have potential adverse effects, so per destination load balancing is preferred, but if there is a reason why pdlb doesn't fit the bill, pplb is a reasonable choice.let me summarize. Iljitsch says that pplb isincompatible withanycast,No. What I'm saying in general is that anycast isn't 100% problem free, so: 1. There should always be non-anycast alternatives 2. It would be good if we had a way (= BGP community) to make sure that anycasted routes aren't load balanced across I don't think either of these is unreasonable.since a pplb-using access router at the inner edgeof an ISP could heartwo different IGP routes to some destination,which ended up takingdifferent exits from the ISP and thus differentBGP paths. I'm not even sure if I understand this sentence, but it sure doesn't look like something I said. What I said was, that if you inject packets towards an anycasted address into two different routers within a certain AS, there is a very real possibility these two packets will end up at different anycast instances. I'm on very firm ground here as this follows directly from the BGP path selection rules. (Although in real life this wouldn't happen too often because customers tend to connect to two routers in the same or neighboring pops.)whereas pplb would normally only operate on equal-cost paths,the BGP->IGP pathwould hide the variance in BGP paths and make these"paths" eligible forpplb.Again: huh?i've said that pplb is only useful for turning twoOC3's into an "OC6"(or similar circuit bundling where a pair of routershas multipleconnections to eachother) and that even in this case, packetreordering is likelyto occur, which will make tcp-flow performance sufferacross this "link". But would the TCP performance over this "OC6 link" be better than that over a single OC3 link? That's the real question.i have also said that turning pplb on acrossnon-parallel links, suchas to multiple providers or through multiple tunnels orwhatever, wouldpretty much guaranty that a word rhyming with "massivesuckage" would occur.and i've made these claims independent of anycast --that is, life will bebad if you use pplb outside its intended purpose, evenif nobody anywherewas using anycast.Your argument is that since it's a bad idea to do this, nobody will, so making it even worse is ok. My argument is that even though it's a bad idea, some people will do it we shouldn't unnecessarily make things worse and/or make a reasonable effort to repair the damage.loath though i am to treat a "preponderance ofassertion" as equivilentto "proof", i see no alternative on this issue.noone is defendingthe use case Iljitsch is proposing. noone is even saying"i tried that and itwas OK". lots of people are saying various thingslike "don't do that!"and "are you crazy?"And we all know that when you tell people not to do something they don't, and there are no crazy people connected to the net.
__________________________________________________ Do You Yahoo!? Log on to Messenger with your mobile phone! http://sg.messenger.yahoo.com
Current thread:
- Re: Anycast 101, (continued)
- Re: Anycast 101 Iljitsch van Beijnum (Dec 17)
- Re: Anycast 101 Paul Vixie (Dec 17)
- Re: Anycast 101 Iljitsch van Beijnum (Dec 17)
- Re: Anycast 101 William Allen Simpson (Dec 17)
- Re: Anycast 101 Paul Vixie (Dec 18)
- Re: Anycast 101 Iljitsch van Beijnum (Dec 20)
- Re: Anycast 101 Stephane Bortzmeyer (Dec 20)
- Re: Anycast 101 Paul Vixie (Dec 20)
- Re: Anycast 101 Iljitsch van Beijnum (Dec 21)
- Message not available
- Re: Anycast 101 Christopher L. Morrow (Dec 21)
- Re: Anycast 101 Iljitsch van Beijnum (Dec 17)
- Re: Anycast 101 Joe Shen (Dec 20)
- Re: Anycast 101 bmanning (Dec 20)
- Re: Anycast 101 Paul Vixie (Dec 20)
- Re: Anycast 101 Petri Helenius (Dec 20)
- Re: Anycast 101 Paul Vixie (Dec 20)
- Re: Anycast 101 Richard Irving (Dec 17)
- Re: Anycast 101 vijay gill (Dec 17)
- Re: Anycast 101 Richard Irving (Dec 17)