Re: Triple Play [was: CAUTION: Potentially Dumb Question...]
On Mon, 6 Feb 2006, Christian Kuhtz wrote: If you're near real time, you have lots of options actually. And I would contend that p2p can be efficient for broadcast distribution actually. There already are several startups doing exactly that for large scalability. Yep. Lots of startups have lots of ideas. If you are selling hammers, you can use the same hammer for lots of projects. But I'm not a true believer in the hammer religion. No actual end user (other than the geek crowd) will ever care that it's BitTorrent or whatever. Agreed. But that doesn't mean a bastardization of the idea won't run underneath. I'm a terrible forecaster. I have no idea how the future will turn out. Sometimes there are several ways to solve a problem.
NANOG36 PGP Key Signing
The key signing will be on Monday at 3pm in the State room. If you can't make it, feel free to submit keys as there will be a follow-up session during the Wednesday morning break. So get those keys in and I'll see you in Dallas! --msa -snip- Stickers for Your Name Badge When you stop by the registration desk at NANOG36, there will be colored stickers available for your name tag that indicate if you have an interest in signing PGP keys. If people keep trying to peer with you, you've picked up the wrong color sticker and should go back. How the Key Signing Works Those of you who plan to participate should email an ASCII extract of your public key to [EMAIL PROTECTED] by 10:00 p.m. CST on Sunday, February 12. Please include 'NANOG PGP KEY' in the subject, and if possible, don't send your key as a MIME attachment. I realize that some MUAs make this difficult, and I will attempt to fix any MIME-attached keys. Instructions for extracting your key to an ASCII file are below. After noon on the 13th, a complete key ring with all of the submitted keys will be available at puck.nether.net/~majdi/nanog36.pgp in binary form, and as an ASCII file at puck.nether.net/~majdi/nanog36.txt. Handouts with the details of each key submitted will be provided. All you should need to bring with you is: * Photo ID (driver's license, passport, etc.) * Your key ID, and its fingerprint * A pen Thank you, and I'm looking forward to seeing you all in Dallas! How to Extract Your Public Key to an ASCII File PGP 2.x: pgp -kxa your_email_address mykey.asc PGP 5.x: pgpk -xa your_email_address mykey.asc GnuPG: gpg --export --armor your_email_address mykey.asc PGP on Windows: Start the PGPkeys application, select your key in the list, click on the Keys menu, select Export, name the resulting file, and make sure that Include Private Keys is NOT checked. PGP on a Mac: I assume the procedure is similar to the one for Windows, but cannot confirm this. Hopefully it's easy enough to figure out.
Re: Triple Play [was: CAUTION: Potentially Dumb Question...]
On Feb 7, 2006, at 10:27 AM, Sean Donelan wrote: On Mon, 6 Feb 2006, Christian Kuhtz wrote: If you're near real time, you have lots of options actually. And I would contend that p2p can be efficient for broadcast distribution actually. There already are several startups doing exactly that for large scalability. Yep. Lots of startups have lots of ideas. If you are selling hammers, you can use the same hammer for lots of projects. But I'm not a true believer in the hammer religion. Argh. What I'm saying is that this is being worked on. And I know from the research perspective in a previous life that it can be made work. The fact that startups are working on commercializing wasn't supposed to suggest viability (it never does), but that products are on the way to market. I have my confirmation of viability of the concept from a different background altogether and I don't subscribe to startup=viability for anything.
Middle Eastern Exchange Points
I know of a Cairo IXP, and possibly one in the UAE. Is there one in Kuwait as yet?
Re: Middle Eastern Exchange Points
There is one in Pakistan, and maybe Dubai. I would address this question to the SANOG list. Regards Marshall On Feb 7, 2006, at 12:48 PM, Howard C. Berkowitz wrote: I know of a Cairo IXP, and possibly one in the UAE. Is there one in Kuwait as yet?
RE: Triple Play [was: CAUTION: Potentially Dumb Question...]
I think the main challenge in making this type of media distribution a reality is not the technology, we mostly know how to make it work. The real challenge is the content owners' willingness to make the content available while preserving their IP rights. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Christian Kuhtz Sent: Tuesday, February 07, 2006 7:46 AM To: Sean Donelan Cc: nanog@merit.edu Subject: Re: Triple Play [was: CAUTION: Potentially Dumb Question...] On Feb 7, 2006, at 10:27 AM, Sean Donelan wrote: On Mon, 6 Feb 2006, Christian Kuhtz wrote: If you're near real time, you have lots of options actually. And I would contend that p2p can be efficient for broadcast distribution actually. There already are several startups doing exactly that for large scalability. Yep. Lots of startups have lots of ideas. If you are selling hammers, you can use the same hammer for lots of projects. But I'm not a true believer in the hammer religion. Argh. What I'm saying is that this is being worked on. And I know from the research perspective in a previous life that it can be made work. The fact that startups are working on commercializing wasn't supposed to suggest viability (it never does), but that products are on the way to market. I have my confirmation of viability of the concept from a different background altogether and I don't subscribe to startup=viability for anything.
Re: Interesting netflow entry
On Mon, 6 Feb 2006, Wil Schultz wrote: Incidentally (because I ask everyone this), what's your flow volume (flows per second)? Cannot get ahold of the machine until tomorrow. I did a 'wc' on 4 devices for 5 minutes and it comes out to just under 3600, about 11-12 per second... Erm, that seems kind of low. Flow volume for two 6509s in what I consider a small to medium size hosting site, with about 6+ gigs of differentiated egress generates more than 8 to 9 *thousand* flows per second, and that's after discard incomplete tcp flows (port scans, half open syns, etc.) Are you sure you're getting everything? - billn
Re: Interesting netflow entry
On Tue, 7 Feb 2006, Bill Nash wrote: Erm, that seems kind of low. Flow volume for two 6509s in what I consider a small to medium size hosting site, with about 6+ gigs of differentiated egress generates more than 8 to 9 *thousand* flows per second, and that's after discard incomplete tcp flows (port scans, half open syns, etc.) Are you sure you're getting everything? he did previously state he was only using about 120mbps... and it'd depend upon his/your sample rates as well...
Re: Middle Eastern Exchange Points
Howard C. Berkowitz wrote: I know of a Cairo IXP, and possibly one in the UAE. Is there one in Kuwait as yet? ISOC-IL is running the IIX for Israel.
Re: Interesting netflow entry
On Tue, 7 Feb 2006, Christopher L. Morrow wrote: Are you sure you're getting everything? he did previously state he was only using about 120mbps... and it'd depend upon his/your sample rates as well... Missed that part. Even so, 120mbps of actual usage, I would expect to see a higher volume. Sampling would definitely bring this down a bit, but for a volume that small, why bother sampling? You'll miss too much. One problem I had while checking out various packages, flow-tools specifically, is that some can't handle differing flow versions. Also, flow generation from a routing-capable 6509 is configured in two different places, so the potential to lose flow traffic due to poor documentation (of both the collector and the generator) definitely exists. Flow-tools picks which version it processes based on the version of the first flow packet it receives, and then discards all else. - billn
Re: Middle Eastern Exchange Points
On 2/7/06, Howard C. Berkowitz [EMAIL PROTECTED] wrote: I know of a Cairo IXP, and possibly one in the UAE. Is there one in Kuwait as yet? http://www.emix.net.ae/ it's flash heavy fyi
Re: Middle Eastern Exchange Points
I know of a Cairo IXP, and possibly one in the UAE. Is there one in Kuwait as yet? Yes, KIX. Note, there's CIX and CRIX. If you are trying to reach African users, there's also KIX ala Kenya. -M -- Martin Hannigan(c) 617-388-2663 Renesys Corporation(w) 617-395-8574 Member of the Technical Staff Network Operations [EMAIL PROTECTED]
eastern Nebraska regional service problem (Sprint, ATT, Qwest transport?)
I'm talking to an ISP in eastern Nebraska who has a DS3 to Sprint. They've got a peer they tie to with private fiber. That peer has a DS3 from ATT. Both normally see 20ms response times on pings from their border routers to the carrier router. Since last Thursday the Sprint connected ISP has been seeing this time bounce between 20ms and 1000ms or more. Customer experience with regards to latency mirrors the ping response times. Sprint has done intrusive testing on the DS3 physical layer/link layer - no problems found. The Sprint connected ISP has good bandwidth management practices - control of usage via an Allot Netenforcer and netflow export from their border router collected by a Manage Enginer Netflow Analyzer box. They see nothing out of the ordinary. I have less contact with the ATT ISP but I'm told they've got some sort of netflow collection going and they've not seeing anything unusual in terms of customer traffic. Is anyone else seeing this? We're guessing there might be some shared MPLS transport for ATT and Sprint in this mix since its affecting both ISPs in the same fashion, but we don't know the gritty details about the path. If this is an MPLS issue I believe Qwest transport could be involved. -- mailto:[EMAIL PROTECTED] // IM:layer3arts voice: 402 408 5951 cell : 402 301 9555 fax : 402 408 6902
Re: Did anyone else notice the CAIDA skitter poster in the background of George Bush's speech at the NSA?
[warning NOT operationally relevant, just need to clarify] re http://news.yahoo.com/photos/ss/events/ts/122805nsaspying/im:/060125/480/dcev10301252131;_ylt=Ag51RnYLYcMpHtd_Cq9ZJCNiWscF;_ylu=X3oDMTA3dmhrOGVvBHNlYwNzc20- thanx to those who forwarded this, it was news to us. note that the AScore image has been used often by government agencies over the past 5 years, and its use doesn't imply any relationship. but in interest of full disclosure (though i believe it's entirely unrelated to the photo-prop): in 2005 NSA did supplement (about $100k, and we hope more this year) an NSF grant, in order to keep skitter (macroscopic topology measurement) on life support. Joel's DARPA and NSF URLs are from skitter.infancy -- before NSA's rescue maneuver, skitter hadn't had any earmarked funding in over 5 years (Cisco has also helped fund some analysis, but funding raw measurement is hard). the NSA funding allowed us to keep it going one more year, while we try to ascertain its role, if any, in some future, if any, community-oriented measurement infrastructure: http://www.caida.org/outreach/papers/2005/conmi/ (comments/feedback More than welcome.) in any event, there is nothing classified or covert in the skitter project or measurement, and in particular we have never made any data available to NSA that isn't available to any other researchers who can stomach filling out the required web forms: http://www.caida.org/tools/measurement/skitter/research.xml and in case of commercial use, join caida as a member, though we kind of relaxed that requirement when all the commercially interested customers were non-profits according to wall street's (apparently even less validated) measurement methodology. we have different access control policies for passive (tapped) data, since that data belongs to the owner of the tapped link, so the owner controls what happens to it. (afaik, DHS is doing the most to advance this ball on scaling sensitive data sharing ( http://www.predict.org ) in a way that ISPs support, but it's also severely underfunded within USG and thus slow to get off the ground.) not to drive this into the ground (unless it helps), but it bears recognizing/remembering that NSA has two missions, only one of which seems to get media attention: http://www.nsa.gov/about/about3.cfm publically available macroscopic Internet topology data falls into the first. anyone who thinks NSA needs caida for its SIGINT mission has never been brought up to speed on the capabilities of either.. (caida is not currently getting any commercial data (network upgrade - funding new monitor, building new monitor, installing new monitor, etc. same cooperative trick, different decade. we don't have it down yet..) and finally, while we appreciate the international affinity for ASporn, our real goal (all along) is to improve the integrity of empirically grounded Internet science, e.g., http://www.caida.org/analysis/topology/rank_as/ http://as-rank.caida.org/ which additional funding is certainly necessary but not sufficient to improve. and no, no existing USG agency has financially embraced the (open, anyway) Internet measurement mission. which is arguably not the worst case scenario, but i [re]read brin's 'transparent society' [ http://www.davidbrin.com/tschp1.html ] last month, and it reads much more eerily with 8 years of empirical data to support his analysis. eeps. sorry for latency, length, tangents, and operational sub-relevance, but felt a need to clarify, k
Re: Middle Eastern Exchange Points
On 7-Feb-2006, at 11:27, Aaron Glenn wrote: On 2/7/06, Howard C. Berkowitz [EMAIL PROTECTED] wrote: I know of a Cairo IXP, and possibly one in the UAE. Is there one in Kuwait as yet? http://www.emix.net.ae/ it's flash heavy fyi Note that EMIX is a transit service, not really peering. (It's peering in the same way that once upon a time the Australian for peering was buy transit from Telstra.) Joe
Re: Middle Eastern Exchange Points
On 7-Feb-2006, at 11:54, Martin Hannigan wrote: I know of a Cairo IXP, and possibly one in the UAE. Is there one in Kuwait as yet? Yes, KIX. Note, there's CIX and CRIX. If you are trying to reach African users, there's also KIX ala Kenya. The exchange point in Nairobi is called KIXP, not KIX, in case it helps avoid that confusion. The KIXP is The Place to reach Kenyan users, but no ISPs from parts of Africa outside Kenya participate in it, as far as I know. http://www.kixp.net/. Terrestrial paths between adjacent African countries are still somewhat rare. I don't have science to back this up, but I would not be surprised if the toplogical centre of today's African Internet turned out to be the LINX. Joe
Re: Interesting netflow entry
Got my hands on the box today, looks like it is Skype. Below is a support article from their site: http://support.skype.com/index.php?_a=knowledgebase_j=questiondetails_i=148
Re: ml hacks for goodmail
* Randy Bush: so, anyone working on the majordomo and mailman hacks for goodmail? i am sorry, but you can not subscribe to this list from an aol.com address. don't ask us to explain, ask [EMAIL PROTECTED] or am i missing something here? clue-bat if so, please. I don't expect the existing filters will change significantly. If you've got problems routing mail to AOL customers, you are just offered another option. I would be surprised if AOL intends to make money off that service; it's probably just an experiment if this helps to curb misuse of the bypass facilities (which have already existed). What's the response of the solicited bulk mailers? Do they welcome this move? If they are too happy about it, maybe we should be worried. 8-) As far as I can tell, the filters at AOL are far less problematic than crude filters at smaller sites which simply use SORBS or bl.spamcop.net.
Re: Middle Eastern Exchange Points
At 04:11 PM 2/7/2006, Joe Abley wrote: On 7-Feb-2006, at 11:54, Martin Hannigan wrote: I know of a Cairo IXP, and possibly one in the UAE. Is there one in Kuwait as yet? Yes, KIX. Note, there's CIX and CRIX. If you are trying to reach African users, there's also KIX ala Kenya. The exchange point in Nairobi is called KIXP, not KIX, in case it helps avoid that confusion. The KIXP is The Place to reach Kenyan users, but no ISPs from parts of Africa outside Kenya participate in it, as far as I know. http://www.kixp.net/. Terrestrial paths between adjacent African countries are still somewhat rare. I don't have science to back this up, but I would not be surprised if the toplogical centre of today's African Internet turned out to be the LINX. Yes, and double checking, I believe CIX/CRIX are one in the same with a distinction being telco colo vs. IP colo. It's not specifically clear. CRIX is run by the Egyptian MCIT. There are other options in Egypt depending upon what you are doing. The incumbent IP provider is Xceed. -M Martin Hannigan(c) 617-388-2663 Renesys Corporation(w) 617-395-8574 Member of Technical Staff Network Operations [EMAIL PROTECTED]
RE: Middle Eastern Exchange Points
A look at Telegeography's bandwidth maps suggest that the African routes are predominantly coastal. http://www.afridigital.net/downloads/DFIDinfrastructurerep.doc adds some more detail. Frank -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Joe Abley Sent: Tuesday, February 07, 2006 3:12 PM To: Martin Hannigan Cc: Howard C. Berkowitz; nanog@merit.edu Subject: Re: Middle Eastern Exchange Points On 7-Feb-2006, at 11:54, Martin Hannigan wrote: I know of a Cairo IXP, and possibly one in the UAE. Is there one in Kuwait as yet? Yes, KIX. Note, there's CIX and CRIX. If you are trying to reach African users, there's also KIX ala Kenya. The exchange point in Nairobi is called KIXP, not KIX, in case it helps avoid that confusion. The KIXP is The Place to reach Kenyan users, but no ISPs from parts of Africa outside Kenya participate in it, as far as I know. http://www.kixp.net/. Terrestrial paths between adjacent African countries are still somewhat rare. I don't have science to back this up, but I would not be surprised if the toplogical centre of today's African Internet turned out to be the LINX. Joe
Re: NANOG36 PGP Key Signing
--On February 7, 2006 7:29:56 AM -0800 Majdi S. Abbas [EMAIL PROTECTED] wrote: PGP on a Mac: I assume the procedure is similar to the one for Windows, but cannot confirm this. Hopefully it's easy enough to figure out. Depends on what you're using. GPG instructions are the same, there's also a utility called GPG Keychain Access, click on the correct key, click on export, check ASCII Armored and give it a file name and a place to store it. But, hopefully, anyone using OS X has already figured these out ;)
Re: Interesting netflow entry
Apparently not, this looks more like it: Time window: Feb 05 2006 22:56:57 - Feb 07 2006 16:58:10 Flows analysed: 202925 matched: 202925, Bytes read: 10028280 Sys: 0.500s flows/second: 405167.7 Wall: 1.293s flows/second: 156923.1 Just a few more than 11 -Wil
FYI - RFC 4367 on What\'s in a Name: False Assumptions about DNS Names
I think some of the people here may want to read this new RFC: http://www.rfc-editor.org/rfc/rfc4367.txt RFC 4367 Title: What\'s in a Name: False Assumptions about DNS Names Author: J. Rosenberg, Ed., IAB Status: Informational Date: February 2006 Mailbox:[EMAIL PROTECTED] Pages: 17 Characters: 41724 -- William Leibzon Elan Networks [EMAIL PROTECTED]
Re: FYI - RFC 4367 on What\'s in a Name: False Assumptions about DNS Names
On Tue, 7 Feb 2006, william(at)elan.net wrote: I think some of the people here may want to read this new RFC: http://www.rfc-editor.org/rfc/rfc4367.txt Small comment - its probably not the people here that need to read it but people at http://www.icann.org But then again it doesnt appear that ICANN has any public discussion/comments forum so no good way to reach them -- William Leibzon Elan Networks [EMAIL PROTECTED]
Re: CAUTION: Potentially Dumb Question...
On Mon, 6 Feb 2006, Randy Bush wrote: I'm interested in responses to this ... MPLS is still a four letter word .. :) http://www.potaroo.net/ispcol/2006-02/converged.html here's me hiding this article from 'management' who are again chasing the 'converged' network :( In some cases it appears convergence makes some sense, I think often though (in my very humble experience) it's more of a buzzword-compliance test than anything else. In the case which kicked off this discussion I was struck that perhaps an older and simpler solution (ipsec vpn and some strict firewalling) would provide the seperation necessary over a single network connection. Oh the fun of converged networks, mpls private vpn's :)
Re: CAUTION: Potentially Dumb Question...
http://www.potaroo.net/ispcol/2006-02/converged.html here's me hiding this article from 'management' who are again chasing the 'converged' network :( In some cases it appears convergence makes some sense, I think often though (in my very humble experience) it's more of a buzzword-compliance test than anything else. actually, it is commonly an internal power play. we'll deploy mpls over our lambdas and then take over the ip business unit, the frame relay business unit [0], the voice folk, ... randy --- [0] - aka how to turn the fr cash cow into a bleeder
Re: Middle Eastern Exchange Points
On Tue, 7 Feb 2006, Joe Abley wrote: I would not be surprised if the toplogical centre of today's African Internet turned out to be the LINX. Yep, with 111 8th close behind. On Tue, 7 Feb 2006, Frank Bulk wrote: A look at Telegeography's bandwidth maps suggest that the African routes are predominantly coastal. Effectively, there's no connection between North Africa and the rest of Africa... North Africa is relatively well connected to Europe via multiple cables across the Mediterranean. The western coast of Africa, wrapping around down to Cape Town, is served by SAT3/WASC, which is a consortium cable with a strict noncompete, so there's no market pricing available anywhere along there... Fiber is just as expensive as satellite, but with the additional cost and hassle of monopoly backhaul from the landing. East Africa and the land-locked central African countries are unserved. Since Nigeria is a huge market and generates a fair amount of cash relative to other markets in Africa, there are a couple of new cables which may soon introduce competition on the relatively short route from Lisbon down to Lagos and Abuja. Also, there's been talk forever, but no action, on an East African cable which would close the loop down from Djibouti to Cape Town, serving Mombasa and Dar and Maputo. The population on the east coast is smaller and less densely packed, though, and the fact that SAT3/WASC is effectively running without a safety net (unless anybody's bothering to patch a protect loop through SAFE to KL and back again through FLAG, which I doubt) doesn't seem to bother anybody, since the cable is priced out of the market anyway, and is thus virtually empty. Anyway, back to the conversation at hand: On Tue, 7 Feb 2006, Howard C. Berkowitz wrote: Middle Eastern Exchange Points I know of a Cairo IXP, and possibly one in the UAE. Is there one in Kuwait as yet? All the ISPs I've talked to in Egypt claim that the Cairo IX was a failed experiment and that they haven't heard anything about it in the last two years. Which roughly corresponds with the last time I heard anyone talking about it in the present tense. But I'll defer to Joe if he has other information. As Joe's pointed out, what's available in the UAE, Saudi Arabia, and Kuwait are governmental monopoly incumbent transit services, a la STIX, as opposed to Internet exchanges where peering takes place. There are several private colocation facilities which sell transit, but are not IXes, in Dubai and Kuwait. There has been a Bahrain governmental effort to get an actual neutral IX going, which has been taking a while to get up to speed, and isn't out of the weeds yet... They've been talking to all the right people, have a site, have commitments from all of the cable systems, have ISP customers who've signed letters of intent and have cash waiting, but they don't have a building yet, just a bunch of cargo containers sitting on the lot in Manama, and a satellite dish farm. Nothing else I know of. -Bill
Re: So -- what did happen to Panix?
Martin Hannigan wrote: My answer, in short, was to say that I see it as more of an enterprise play because it's a managed service and the hardest part of provisioning is typically the order cycle. If you are an ISP, you are theoretically multi homed by definition and your providers are going to remain fairly stable (you hope) based on your own needs. My point remains: designs based on such assumptions are not a good idea, since these assumptions are by no means fundamental and could certainly change. People get creative with how they announce prefixes, change upstreams, etc., and you can't assume that things like this would stay the way they are. As an aside, another question occurred to me about delaying unusual announcements. Boeing Connexion offers another example of unorthodox prefix announcements. Wouldn't the tactic of delaying unusual announcements would cause problems for this service? -Nick
Re: So -- what did happen to Panix?
On Tue, 7 Feb 2006, Nick Feamster wrote: As an aside, another question occurred to me about delaying unusual announcements. Boeing Connexion offers another example of unorthodox prefix announcements. Wouldn't the tactic of delaying unusual announcements would cause problems for this service? I had thought Josh's paper (or maybe not josh, whomever it was) said something along the lines of: 1) if more than one announcement prefer 'longer term', 'older', 'more usual' route 2) if only one route take it and run! So.. provided Connexion withdraws from 'as-germany' and announces in 'as-atlantic ocean', and so on there would only be 1 route, and you'd fall to step 2. (yes, the paper was more detailed and there were more steps...)
Re: Middle Eastern Exchange Points
At 10:30 PM 2/7/2006, Bill Woodcock wrote: [ SNIP ] Anyway, back to the conversation at hand: On Tue, 7 Feb 2006, Howard C. Berkowitz wrote: Middle Eastern Exchange Points I know of a Cairo IXP, and possibly one in the UAE. Is there one in Kuwait as yet? All the ISPs I've talked to in Egypt claim that the Cairo IX was a failed experiment and that they haven't heard anything about it in the last two years. Which roughly corresponds with the last time I heard anyone talking about it in the present tense. But I'll defer to Joe if he has other information. As Joe's pointed out, what's available in the UAE, Saudi Arabia, and Kuwait are governmental monopoly incumbent transit services, a la STIX, as opposed to Internet exchanges where peering takes place. There are several private colocation facilities which sell transit, but are not IXes, in Dubai and Kuwait. Bill: UAE ISC has equipment out here. 192.228.85.0/24 is being announced out of emirates.net can't be that bad. :-) they are downstream of a whole bunch of net and I see what looks like an IX. (corrections welcome) UAE looks interesting network wise. It's too bad they can't get it together as you noted. I don't see it as bad as you...interconnecting in a government exchange is still peering. It may not be exactly the same, but I've found in some cases you can't be too picky if you can peer with even a few regionals. KIX: 3 lg. upstreams, 4 regional isp down, interconnected to UAE IX Cairo: CRIX is dead in name, but MCIT is running some exchange space refer to it crix xor cix xor mcit possibly by simple legacy and they will talk to anyone about it. Xceed is the incumbent, renamed IIRC. The terminology and sexy colo's built to Telcordia standards and NEBS compliance may not be out there, but they are peering, even if it isn't by our definitions. Howard, contact info for each out of band. -M Martin Hannigan(c) 617-388-2663 Renesys Corporation(w) 617-395-8574 Member of Technical Staff Network Operations [EMAIL PROTECTED]
Re: So -- what did happen to Panix?
At 11:27 PM 2/7/2006, Nick Feamster wrote: Martin Hannigan wrote: My answer, in short, was to say that I see it as more of an enterprise play because it's a managed service and the hardest part of provisioning is typically the order cycle. If you are an ISP, you are theoretically multi homed by definition and your providers are going to remain fairly stable (you hope) based on your own needs. My point remains: designs based on such assumptions are not a good idea, since these assumptions are by no means fundamental and could certainly change. People get creative with how they announce prefixes, change upstreams, etc., and you can't assume that things like this would stay the way they are. Nick: I wouldn't call them assumptions. I would call them engineering decisions in operational environments. I guess I fail to see where a commodity market with a broker adding a vig resolves a real network problem. I'm think tier1? They aren't buying service from anyone on Equinix direct and move/add/drop is just another day on the Internet. I really can't see any provider doing it, but perhaps smaller ones. *shrug*. I don't know why you wouldn't make temporary arrangements via peering fabric, PNI, or transit and eliminate the middle man (point of failure). As an aside, another question occurred to me about delaying unusual announcements. Boeing Connexion offers another example of unorthodox prefix announcements. Wouldn't the tactic of delaying unusual announcements would cause problems for this service? [ snip ] -M -Nick Martin Hannigan(c) 617-388-2663 Renesys Corporation(w) 617-395-8574 Member of Technical Staff Network Operations [EMAIL PROTECTED]
Re: Middle Eastern Exchange Points
On Tue, 7 Feb 2006, Martin Hannigan wrote: Interconnecting in a government exchange is still peering. Uh, not if it's buying transit. They are peering, even if it isn't by our definitions. Uh, Marty... the difference between peering and transit is that they have different definitions. If you say transit is peering, just not by our definitions, then you're into 1984 territory. War is Peace, Freedom is Slavery, Ignorance is Strength. For me, however, peering is peering, and transit is transit, and my world works better when I use words in accord with, rather than in contravention to, their definitions. -Bill
Re: So -- what did happen to Panix?
Chris has it! And to be clear, we only require a slow (1 day) provider changeover in the case that you want to announce your old provider's sub-prefix at a new provider. For instance, if you are an ATT customer using a 12/8 sub-prefix and change providers but keep the prefix, the prefix will look funny coming from another originator for the first day and be delayed. All other methods of changing providers will not be interfered with. Josh I had thought Josh's paper (or maybe not josh, whomever it was) said something along the lines of: 1) if more than one announcement prefer 'longer term', 'older', 'more usual' route 2) if only one route take it and run! So.. provided Connexion withdraws from 'as-germany' and announces in 'as-atlantic ocean', and so on there would only be 1 route, and you'd fall to step 2. (yes, the paper was more detailed and there were more steps...)
Re: Middle Eastern Exchange Points
At 11:55 PM 2/7/2006, Bill Woodcock wrote: On Tue, 7 Feb 2006, Martin Hannigan wrote: Interconnecting in a government exchange is still peering. Uh, not if it's buying transit. They are peering, even if it isn't by our definitions. Uh, Marty... the difference between peering and transit is that they have different definitions. If you say transit is peering, just not by our definitions, then you're into 1984 territory. Bill: I'm pretty sure you know that I know the difference between paid transit and peering. If I were buying transit, I would've had a different comment. I think we may have a difference though, I don't think jumping on a big switch and saying yes to every peering request is peering, and I think this is a better discussion at the peering bof or in person. Anyhow, I'll be happy to tell you as much as I can (NDA) in Montreal. I don't see you listed for Dallas. Is Howard going to find peering in, er, Egypt? It depends what his value proposition is. -M Martin Hannigan(c) 617-388-2663 Renesys Corporation(w) 617-395-8574 Member of Technical Staff Network Operations [EMAIL PROTECTED]
Re: Middle Eastern Exchange Points
On 7-Feb-2006, at 20:50, Martin Hannigan wrote: As Joe's pointed out, what's available in the UAE, Saudi Arabia, and Kuwait are governmental monopoly incumbent transit services, a la STIX, as opposed to Internet exchanges where peering takes place. There are several private colocation facilities which sell transit, but are not IXes, in Dubai and Kuwait. ISC has equipment out here. 192.228.85.0/24 is being announced out of emirates.net can't be that bad. :-) The F-root node in Dubai is facilitated by Emirates Telecom/Etisalat/ EMIX, as per http://f.root-servers.org/. At the time we installed there was no facility available for peering or other multi-point interconnect with operators in UAE. I am not aware that this has changed. Woody's comparison with the STIX is spot on, as far as I know. In pragmatic terms, due to the local regulatory environment and in the absence of a neutral exchange point, obtaining transit from EMIX in Dubai is the best approximation to a comprehensive set of bilateral peering arrangements with local ISPs. However, it's not peering in a topological/routing policy sense. The fact that F-root's covering prefix doesn't propagate beyond the region is due to special handling of that prefix by our colleagues in AS 8966. ISC's intention in Dubai, as in all regions, was to provide the best access possible to F-root within the immediate surrounding region. I believe we achieved that goal. Joe
Re: Middle Eastern Exchange Points
On Tue, 7 Feb 2006, Bill Woodcock wrote: different definitions. If you say transit is peering, just not by our definitions, then you're into 1984 territory. So what exactly is definition of transit that does not make it peering? And when ISP A buys access from ISP B for purpose of getting to ISP C is that peering or transit? -- William Leibzon Elan Networks [EMAIL PROTECTED]
Re: Middle Eastern Exchange Points
At 01:11 AM 2/8/2006, Joe Abley wrote: On 7-Feb-2006, at 20:50, Martin Hannigan wrote: As Joe's pointed out, what's available in the UAE, Saudi Arabia, and Kuwait are governmental monopoly incumbent transit services, a la STIX, as opposed to Internet exchanges where peering takes place. There are several private colocation facilities which sell transit, but are not IXes, in Dubai and Kuwait. ISC has equipment out here. 192.228.85.0/24 is being announced out of emirates.net can't be that bad. :-) The F-root node in Dubai is facilitated by Emirates Telecom/Etisalat/ EMIX, as per http://f.root-servers.org/. At the time we installed there was no facility available for peering or other multi-point interconnect with operators in UAE. I am not aware that this has changed. Woody's comparison with the STIX is spot on, as far as I know. Guys, are you being semantic? I'm *agreeing with you and Woody here. Just not re: Kuwait and Egypt. You keep saying EMIX and you're confusing me. Peering or no? IX naturally insinuates yes regardless of neutrality. In pragmatic terms, due to the local regulatory environment and in the absence of a neutral exchange point, obtaining transit from EMIX in Dubai is the best approximation to a comprehensive set of bilateral peering arrangements with local ISPs. However, it's not peering in a topological/routing policy sense. The fact that F-root's covering prefix doesn't propagate beyond the region is due to special handling of that prefix by our colleagues in AS 8966. That's what I was interested in, and found. I appreciate the political explanation. I saw ASN 8966 and behind that ASN 5384 w/55 prefixes. 5384 looks like a choke point. ISC's intention in Dubai, as in all regions, was to provide the best access possible to F-root within the immediate surrounding region. I believe we achieved that goal. What is the benchmark of speedy resolution vs. application i.e. how fast do you resolve before it's irrelevant, if at all? -M -- Martin Hannigan(c) 617-388-2663 Renesys Corporation(w) 617-395-8574 Member of Technical Staff Network Operations [EMAIL PROTECTED]
