February 2013 - Test - lists.nanog.org

2013.02.04 NANOG57 day 1 afternoon notes
by Matthew Petach 05 Feb '13

05 Feb '13

Notes from the afternoon session, including the community meeting, but minus most of the BCOP presentation have been posted: http://kestrel3.netflight.com/2013.02.04-NANOG57-day1-afternoon-session.txt it looks like apache2 serves up about 100 connections, then wedges. :( No time to troubleshoot it, but i'll try to check back once an hour and kick it in the head. Thank you Dave, our hero of the day for making sure the WCIT talk contents did *not* get sealed behind closed doors! round two begins tomorrow morning, bright and early. ^_^ Matt

2 2

Is Google Fiber a model for Municipal Networks?
by Leo Bicknell 04 Feb '13

04 Feb '13

I've been searching for a few days on information about Google Fiber's Kansas City deployment. While I wouldn't call Google secretive in this particular case, they haven't been very outgoing on some of the technologies. Based on the equipment they have deployed there is speculation they are doing both GPON and active thernet (point2point). I found this presentation: http://static.googleusercontent.com/external_content/untrusted_dlcp/researc… It has a very good summary of the tradeoffs we've been discussing regarding home run fibers with active ethernet compared with GPON, including costs of the eletronics compared to trenching, the space required in the CO, and many of the other issues we've touched on so far. Here's an article with some economics from several different deployments: http://fastnetnews.com/fiber-news/175-d/4835-fiber-economics-quick-and-dirty Looks like $500-$700 in capex per residence is the current gold standard. Note that the major factor is the take rate; if there are two providers doing FTTH they are both going to max at about a 50% take rate. By having one provider, a 70-80% take rate can be driven. Even with us a 4%, 10 year government bond, a muni network could finance out a $700/prem build for $7.09 per month! Add in some overhead and there's no reason a muni-network couldn't lease FTTH on a cost recovery bases to all takers for $10-$12 a month (no Internet or other services included). Anyone know of more info about the Google Fiber deployment? -- Leo Bicknell - bicknell(a)ufp.org - CCIE 3440 PGP keys at http://www.ufp.org/~bicknell/

7 15

2013.02.04 NANOG57 day 1 morning session notes posted
by Matthew Petach 04 Feb '13

04 Feb '13

I jotted down some notes from this morning's session in case people are curious; they should be up visible at http://kestrel3.netflight.com/2013.02.04-day1-morning-session.txt apache's kinda flakey on the box; if it stops responding, ping me and I'll bounce the server. ^_^; Thanks! Matt

1 0

RE: Global caches
by Kyle Camilleri 04 Feb '13

04 Feb '13

Hi Alex, We already have Google Cache although we don't peer with them directly. As said earlier, what I'm after is to cache content to provide better performance to our customers and alleviating some bandwidth towards provider. Taking an example of Global Google Cache, they provide and manage the servers themselves, absolutely no huge effort needed from the ISP. Making it very convenient. I know of Akamai and Netfix that does this in a fairly similar way, but would like to know if there are any other CDNs that do this. Regards, Kyle From: askoorb(a)gmail.com [mailto:askoorb@gmail.com] On Behalf Of Alex Brooks Sent: 04 February 2013 16:13 To: Kyle Camilleri Subject: Re: Global caches Hi, On Mon, Feb 4, 2013 at 2:03 PM, Kyle Camilleri <Kyle.Camilleri(a)melitaplc.com<mailto:Kyle.Camilleri@melitaplc.com>> wrote: Dear Nanog Community, Some CDN providers such as Akamai and Google (often called Global Google Cache) are offering caches to ISPs. It is very convenient for small ISPs to alleviate bandwidth towards the provider, but also the CDN provider benefits by putting source of data closer to the user resulting in far better performance. Does anybody know of any other CDN providers that offer similar caches? I think I should give a note of caution first. Normally, CDNs will only consider you for caching if you already peer with them - in most cases the majority of any performance or cost improvements come from direct peering with a CDN first. Unfortunately, I can't find an entry for you on PeeringDB, but assuming that you are AS12709, you only seem to be peering with a few 'major' networks (Vodafone Malta (AS33874) WIND Italy (AS1267), Global Crossing (AS3549) and Level3 (AS3356)). If you don't (or can't) peer with, say, Google or Akamai, they are unlikely to invite you to join their caching systems. I'm guessing that because you are on an Island in the middle of the Mediterranean; with nobody wanting to peer on the Island, you are looking to cache as much as possible on the Island, as undersea cables are expensive. Is this correct? If you give the list a bit more information about what the problem you are trying to solve is, the people here are pretty good at coming up with solutions or putting you in contact with people who may be able to help. Best wishes, Alex

1 0

WhoisGuard
by Joshua William Klubi 04 Feb '13

04 Feb '13

Is there any person from WhoisGuard . com on this group. or any one can help me with an effective contact, all mails to their ' support(a)whoisguard.com' is not been acknowledged or answered. Joshua

1 0

Re: Muni fiber: L1 or L2?
by Scott Helms 03 Feb '13

03 Feb '13

Eric, Lol, yeah should have been Gig-E :) On Sun, Feb 3, 2013 at 5:49 PM, Eric Brunner-Williams <brunner(a)nic-naa.net>wrote: > On 2/3/13 12:33 PM, Scott Helms wrote: > > PON is worse > > in every performance related way to PON > > typo??? > -- Scott Helms Vice President of Technology ZCorum (678) 507-5000 -------------------------------- http://twitter.com/kscotthelms --------------------------------

1 0

Announcing a reserved ASN?
by Suresh Ramasubramanian 03 Feb '13

03 Feb '13

AS23456 is currently announcing a good few netblocks (which don't have a very good smtp reputation, by the way). Funny thing is, that's a special use ASN as per rfc4893, something about two octet ASNs that don't have a four octet representation. Only one upstream (airtelbroadband-as-ap, as24560) that I can see > > 103.7.204.0/22 > > 103.14.208.0/22 > > 103.23.124.0/22 > > 103.30.12.0/22 > > 103.245.112.0/22 > > 111.235.148.0/22 > > 177.55.249.0/24 > > 186.251.192.0/21 --srs (htc one x)

3 3

Re: Announcing a reserved ASN?
by Dave Pooser 03 Feb '13

03 Feb '13

On 2/3/13 9:04 AM, "Rich Kulawiec" <rsk(a)gsp.org> wrote: >On Sun, Feb 03, 2013 at 06:12:32PM +0530, Suresh Ramasubramanian wrote: >> AS23456 is currently announcing a good few netblocks (which don't have a >> very good smtp reputation, by the way). > >To say the least. A quick rDNS scan reveals that those netblocks include: > > 8448 addresses > 6932 return nxdomain > 512 return servfail > 1004 with rDNS entries > >Those 1004 hosts with rDNS account for 36 domains: <snip long list of spammy domains> Just as another data point, the domain names you listed hit on enough URL blacklists that Spamassassin quarantined the message for me (and would have rejected it during the SMTP transaction had the NANOG server not been listed on DNSWL-High). Spam hosts plus fake ASN = paging the Spamhaus DROP maintainers to the white courtesy phone.... -- Dave Pooser Manager of Information Services Alford Media http://www.alfordmedia.com

4 3

Rollup: Small City Municipal Broadband
by Jay Ashworth 03 Feb '13

03 Feb '13

Ok, here's a rough plan assembled from everyone's helpful contributions and arguing all week, based on the City with which, if I'm lucky, I might get a job Sometime Soon. :-) (I'm sure some of you can speculate which city it might be, but Please Don't.) It's about 3 square miles, and has about 8000 passings, the majority of which are single or double family residential; a sprinkling of multi-tenant, about a dozen city facilities, and a bunch of retail multi-unit business. Oh, and a college campus, commuter. My goal is to fiber the entire city, with a 3-pr tail on each single-family residence (or unit of a duplex/triplex), and N*1.5 on multi-tenant business buildings, and probably about N*1.1 or so on large multi-unit residences. Empty lots, if we have any, will also get a 3-pr tail, in a box. My plan is for the city to contract out the design and build of the physical plant, with each individual pair home-run to a Master Distributing Frame in a city building. Since the diameter of the city is so small, this can be a single building, and it need not be centrally located -- since we are a coastal city, I want it at the other end. :-) I propose to offer to clients, generally ISP, but also property owner/ renters, L1 connectivity, either between two buildings, or to a properly equipped ISP, and also to equip for and offer L2 aggregated connectivity to ISPs, where the city, instead of the ISP, will provide the necessary CPE termination gear (ONT). The entire L0 fiber build, and all L2 aggregation equipment (except potential GPON splitters mentioned next) will be the property of the City. Assuming that the optical math pans out, we will hang GPON splitter frames in the MDT, and cross connect subscriber ports to the front of them, and the back of them to Provider equipment in an associated colo, in rooms or cages; we'll also probably do this for our L2 subscribers, using our own GPON splitters. Those will then be groomed into Ethernet handoffs for whatever providers want to take it that way, at a higher MRC. Splitters installed for Providers who take L1 handoffs will be their property, though installed in our MDF room. We will do all M-A-C work on the MDF, into which Provider employees will generally not be admitted, at least unescorted, on a daily basis, except in "emergencies", for which an extra NRC will be levied. The cost we will charge the Providers, per subscriber, will be a fixed MRC, similar to a 'tariffed' rate, which is published, and all Providers pay the same rate, which is subject by contract to occasional adjustment in either direction, and which is set to recover our costs to provide the service, based on take rates and depreciation periods which I have not yet determined. I'm assuming I can get 30 year depreciation out of the fiber plant with no problems, probably 40... maybe 50 if it's built to high enough standards -- I do not expect passive glass fiber to become obsolete in 50 years. Active equipment, a much shorter period, of course, probably between 4 and 7 years, depending on how far up the S-curve of terminal equipment design it proves that we've already traveled. At the moment, my comparison device is the Calix E7-20, with either 24-port AE or the GPON cards; either 836GE interior ONTs, or their equivalent exterior ones (since the power module has to be inside anyway, I'm not sure you gain that much by putting the ONT outside, but...) My motivation for not doing L3 is that it is said to greatly improve the chances for competition at the ISP level, a fact not yet in evidence. My motivation for not doing GPON in the field is that it's thoroughly impractical to do that in an environment where an unknown number of multiple providers will be competing for the subscribers, and anyway it breaks point to point, which the city will need for itself, and which I want to offer to residents as well. My motivation for doing L2 is that it takes a lost of the front-end cost burden off of potential smaller 'boutique' ISPs specializing in various disciplines (very low cost/lifeline service, very high speed, 'has a big local usenet spool', or what have you); such providers will have to pay (and recover) a higher per-subscriber MRC, in exchange for not having to themselves provision and install GPON splitters and something like a Calix E7 -- such hardware will be installed by the City, and cost-shared; if/when such a provider gets big enough, they can install their own, and we'll cut them over. I propose to take the project to the council for funding and approval having in my pocket a letter of intent from a local 2nd tier ISP of long standing to become our launch provider, with no incentives over the published rates except the guarantee of additional subscribers. My underlying motivation, which is intended to answer any tradeoff queries which I haven't explicitly addresses before this point, is to increase the City's position as being "full service" (as small as it is, it does it's own fire, police, garbage and water already), and improve it's chances of selection by people who are deciding where to move. The City already has a relatively good image, within its target market, but as time marches ever forwards, the maximum available broadband in its footprint will become less and less acceptable, and I expect that there are a significant number of people around the country for whom "I can get Gigabit in my house? Bidirectional? I'm moving" is a valid viewpoint. I know already that "what kind of broadband can I get" is a top-5, and sometimes top-3 selection issue for people contemplating a move. Things, therefore, which improve the city's image with potential immigrants, be they residents or small businesses, are a Good Thing, whether because those people actually want or need those services, or whether it's merely because they like to bask in the reflected glow there of. These things will likely reduce the city's vacancy rate, and thus increase property tax revenue and hence the city's budget, in addition to slowly improving the city's socioeconomic demographics, which will itself likely have a salutary effect on the small businesses already here, and in the decision processes of people thinking to move one here or start one. That's my thinking so far. Now comes the hard part: assembling enough other budgetary numbers to determine how much it will cost, how much we'll have to charge, and whether people will *pay* that much. I don't have any illusions that the wholesale charges will be a revenue stream for the City, and I won't let the council get any such ideas either; the benefits to the city (aside from dark fiber to all our own buildings) are a bit deeper than that, and will require sufficient time to come to fruition. I wrote this as a summary for all the helpful NANOGers who chimed in this week, and as a clarification for those who weren't quite sure where *I* was trying to go -- all muni builds are sui generis, and this one moreso than most. If any of you see anything we've already said, but I left out, please let me know... And have a Whacky Weekend. If any of you pass through the west coast enroute to ORL, let me know. :-) Cheers, -- jra -- Jay R. Ashworth Baylink jra(a)baylink.com Designer The Things I Think RFC 2100 Ashworth & Associates http://baylink.pitas.com 2000 Land Rover DII St Petersburg FL USA #natog +1 727 647 1274

7 15

Re: Followup: Small City Municipal Broadband
by Jay Ashworth 03 Feb '13

03 Feb '13

----- Original Message ----- > From: "Brandon Ross" <bross(a)pobox.com> > > No; I wouldn't offer it retail; I'd offer it to all provider-comers > > wholesale, at cost plus, just like everything else. > > It sure seems like just pushing the competition (or lack thereof) up > the stack. Could be. To compete with Roadrunner, people will have to do triple play, and the CATV is the hard part. If someone else is already doing the aggregation, I'm good with that. > >> Running a decent layer 3 service is "hard" too. Isn't the whole > >> point to > >> let these service providers compete with each other on the quality > >> and > >> cost of their services? > > > > You could say the same thing about the uplink, > > Which uplink is that? I'm a little confused. My colo's uplinks to the world, which were one of three things I proposed offering at wholesale to ISPs. > > though; I note you didn't throw a flag at that, or at Akamai; is the > > IPTV issue different to you? > > If you were to open your colo to all comers that have similar models > to Akamai, that seems fair. After all, it's not the city selling Akamai > services to either the ISPs or end-users, the city is just providing a > convenient way for the providers that are there to interconnect with > content providers that care to show up. Precisely. Akamai's business model is that they just show up? Me and my ISPs don't have to pay them? > Now if you were to encourage an IPTV services provider that WASN'T the > city to co-locate at the facility, that seems reasonable as long as terms > were even if another one wanted to show up. I could imagine that some > might sell service direct retail, others might go wholesale with one of > the other service providers. Maybe both? Perhaps; yeah. > This whole thing is the highway analogy to me. The fiber is the road. > The city MIGHT build a rest stop (layer 2), but shouldn't be allowed > to either be in the trucking business (layer 3), nor in the > business of manufacturing the products that get shipped over the road > (IPTV, VOIP, etc.), and the same should apply to the company that > maintains the fiber, if it's outsourced. Ok, fair point. My goal in IX and Akamai was "unload my uplinks". The bigger my downlinks are, the more I will care. Cheers, -- jra -- Jay R. Ashworth Baylink jra(a)baylink.com Designer The Things I Think RFC 2100 Ashworth & Associates http://baylink.pitas.com 2000 Land Rover DII St Petersburg FL USA #natog +1 727 647 1274

2 1