- Test - lists.nanog.org

RE: Verizon 1xRTT/EVDO for OOB
by David Hubbard 04 May '12

04 May '12

We've recently deployed our first Opengear ACM5004-G http://opengear.com/product-acm5000-g.html and its working great. It has cellular (3G), ethernet, USB and four serial interfaces; you could easily daisy chain a serial interface to a terminal server if you need more OOB serial ports. The setup was very easy; our Verizon rep just required the ESN off the bottom of the unit since Opengear's already gone through Verizon's certification process so their devices' ESN's are known to Verizon as valid for their network. He asked what plan we wanted, I told him I wanted a static public IP, two days later we were up and running. Since this is considered a M2M (machine to machine) device, similar to alarm systems and other low bandwidth devices, you can get monthly plans down to $7 for a few megabytes and up to normal consumer-level 5 gigs for $50/month type plans. The static IP didn't add any cost. The device has a cool unique feature where you can set it to keep the cellular interface 'down' for data as its normal state so no one outside is trying to break into it or wasting your bandwidth. Then you send a text message to the phone number assigned to it and it will bring up the data interface so you can SSH in. You can configure firewall rules on it, use key-based auth, change the ssh port number, etc. David > -----Original Message----- > From: Christopher J. Pilkington [mailto:cjp@0x1.net] > Sent: Friday, May 04, 2012 10:53 AM > To: nanog(a)nanog.org > Subject: Verizon 1xRTT/EVDO for OOB > > Is anyone using Verizon 1xRTT/EVDO ("3G") for OOB work? I'm trying to > sort out how exactly to order a compatible service from them. > Unfortunately I don't manage our Verizon Wireless relationship, so I > need to be specific. > > Is there a service code or name they refer to this service as? > Looking for low bandwidth, static IP. > > -cjp > > >

1 0

NUD- ipV6.
by S, Somasundaram (Somasundaram) 04 May '12

04 May '12

Hi Everyone, Would like to hear from you on the significance of IPV6 "Neighbor Unreachability detection (NUD)" specifically on the Router-Router link. While quick failure detection protocols like BFD are already present to detect the liveliness of the neighbor, does the providers/operators find NUD to be useful? Rgds/ Somasundaram

3 2

Peering in Brazil
by James Bensley 03 May '12

03 May '12

Hi all, I have posted this to the LacNOG list but it seems pretty quite and it involves North America also, so I hope no one minds me reposting here also; I am looking for any guidance and advice people have regarding first time peerings in South America. Currently I am doing some work with a content provider in North America and I want to get them better routers into South America, to South American ISPs. I am looking to get them an interconnect from their NA location to SA, into a PTT IX location in Sao Paulo (they seem to be up there for top IXs in SA, or am I mistaken?). I'm new to the PTT IXs so does anyone have any recommendations about any aspect here, support horror stories, reason to prefer one location for peering over another etc? I see Level3 are in the PTT Sao Paulo location, and we have access to Level 3 already. Is there someone else I should be looking at who is especially good at private routes down to SA, enough to digress from Level 3, or are Level 3 a good choice here? Again, any past experiences are welcome, and recommendations for a different IX or provider any why. Many thanks, James.

2 1

Re: [lacnog] Peering in Brazil
by Rubens Kuhl 03 May '12

03 May '12

> I am looking for any guidance and advice people have regarding first > time peerings in South America. Currently I am doing some work with a > content provider in North America and I want to get them better > routers into South America, to South American ISPs. I am looking to > get them an interconnect from their NA location to SA, into a PTT IX > location in Sao Paulo (they seem to be up there for top IXs in SA, or > am I mistaken?). First a clarification: PTT stands for "traffic exchange point" in Portuguese, so people familiar with this acronym related to telco companies or mobile/ http://ptt.br will only give you routes to Brazil, so you would still need other locations to reach other south american countries. If the content provider is already at Miami NOTA (Nap of The Americas), it's already at one of the best locations to be to reach all South America, but with if a latency penalty. > I'm new to the PTT IXs so does anyone have any recommendations about > any aspect here, support horror stories, reason to prefer one location > for peering over another etc? I see Level3 are in the PTT Sao Paulo > location, and we have access to Level 3 already. Is there someone else > I should be looking at who is especially good at private routes down > to SA, enough to digress from Level 3, or are Level 3 a good choice > here? Again, any past experiences are welcome, and recommendations for > a different IX or provider any why. PTT.br is a Layer-2 traffic exchange with route-servers, so you need either to colo a router here and back-haul your traffic (preferred solution) or to contract a lan-2-lan over MPLS circuit from a provider. Level 3 is probably your best bet, but one other option could be the BrT/Oi facility where Globenet (http://globenet.net/) has its Sao Paulo POP. You would probably prefer to deal with Globenet and have them hire colo and cross-connect from their sister company (Globenet is part of the Oi group). The only fiber cables with available capacity to get to South America are the ones from Level 3, Globenet and Telefónica, but Telefónica International is not a PTT.br location (although being a member and provide transit services) and Telefónica Brazil is so not helpful for such services. LANautilus has some capacity swap with Level 3 as well. In short: if you will be installing a router here but not servers, Level 3 (#1) or Globenet (#2). If you are not installing anything, Telefónica and LANautilus are also choices. If you would be sending a CDN node with servers and will only need Internet uplink, then you have more choices, but that's not my reading of your strategy, am I right ? Disclaimer: I work for the organization who maintains PTT.br. Rubens

1 0

Network diagram app that shows realtime link utilizatin
by Hank Disuko 03 May '12

03 May '12

Hi folks, I wonder if anyone can recommend a network diagram tool that can show realtime link utilization via snmp? Mikrotik's "The Dude" app actually does exactly what I'm looking for, but the snmp support for non-RouterOS devices seems to be lacking, as it simply won't enumerate my switch interfaces in order to capture utilization. I've downloaded several trial tools (WhatsUp, NetCure, Solarwinds LANsurveyor etc.) but they don't serve this very basic need of mine to see the realtime link util in the diagram. Thanks, Hank Disuko

8 7

Re: Network diagram app that shows realtime link utilizatin
by Nathan Eisenberg 03 May '12

03 May '12

Php network weathermap works well for me. The configuration language is pretty straightforward, and it's easy to consume data from either the usual suspects (mrtg), or to write a plugin that uses a custom (sql or other) datasource. This gets you within 60 seconds of realtime, and the price (time) is right. Nathan Eisenberg Sent from my HTC on the Now Network from Sprint! ----- Reply message ----- From: "Hank Disuko" <gourmetcisco(a)hotmail.com> Date: Tue, May 1, 2012 9:42 am Subject: Network diagram app that shows realtime link utilizatin To: "NANOG" <nanog(a)nanog.org> Hi folks, I wonder if anyone can recommend a network diagram tool that can show realtime link utilization via snmp? Mikrotik's "The Dude" app actually does exactly what I'm looking for, but the snmp support for non-RouterOS devices seems to be lacking, as it simply won't enumerate my switch interfaces in order to capture utilization. I've downloaded several trial tools (WhatsUp, NetCure, Solarwinds LANsurveyor etc.) but they don't serve this very basic need of mine to see the realtime link util in the diagram. Thanks, Hank Disuko

1 0

Re: Operation Ghost Click
by A. Pishdadi 02 May '12

02 May '12

At some point in like 10 years when all the computer illiterate people are gone there will be no more excuses for not being educated on malware and viruses. While I understand the ISP doesn't want to possibly cut into there profit margins they could easily put in place monitoring tools that can detect network traffic that is malware bound and reach out to the customer by email, phone and if need be by person. How much of tax payer money is spent to pay these FEDERAL (F.B.I.) agents to sit here and baby sit these computer ignorant and illiterate people for 6 months? So for the big ISPs like comcast i should pay out of my tax money because they cannot properly enforce a network policy that would require them to actually give a crap what is coming out of there network? There is always going to be viruses and malware, they will find ways to get them through but for heavens sake why would we if identified leave millions of compromised machines online with an attempt to do a cleanup? YOU as a network operator have a responsiblity to the other 40,000 AUTONOMOUS network to make sure your not polluting our private network infrastructure with garbage coming from your users and network. Clean up your mess. Like we will not tolerate spammers being housed on 'hosting' networks why should tolerate malware and infections coming from ISP's??? How much money is spent cleaning up hacked word press servers and udp.pl scripts... This is much bigger issue then at any cost making sure a user can get on to facebook to upload a picture of there cat sleeping upside down. If we enforced a proper policy and held network activity to certain standards the ISP's would fix the issue of ignorant users themselves by #1 educating there users , #2 implementing network monitoring on there outbound traffic to identify sources of infected and compromised machines, #3 implementing a cleanup policy, #4 letting the end user know they have a responsibility to make sure the machines they access the network from are clean and to do checks and to do there antivirus updates and os updates. Oh yah, and if we got all these 'supporting' DNS servers up why not just direct ALL users of it, who are clearly infected to a temporary page that will enlighten the customer that they are infected and give them instructions on clean up and give them a deadline of when there service will stop......... How hard is that? On Fri, Apr 27, 2012 at 10:55 PM, <Valdis.Kletnieks(a)vt.edu> wrote: > On Fri, 27 Apr 2012 21:39:20 -0500, you said: > > > Is it not detected by the common anti-virus software vendors? If the > > This assumes that the computer hasn't been hit by something *else* that > disables the user's AV software. Remember, multiple infections are > *common*. > > > internet stopped working on my computer i would reach out to someone who > > knew how to fix it, keeping these people online and spreading the malware > > helps how?? > > The point is that the internet *didn't* stop working, so they have no > reason to > reach out yet. > > And no, you can't just blindly cut the users off and make them call the > ISP for > several reasons: > > 1) At that point, the ISP incurs an expense to fix a problem they didn't > cause. > Remember that margins on most consumer-grade Internet accounts are pretty > thin, > and one long support call can wipe out the profit. So explain why the ISP > wants to cut off a user who makes them $10/year profit, and spend $30 or > more > handling the support call, when they aren't in the business of providing > security services to end users? > > 2) If the user has no POTS, cutting them off may have just cut off their > 911 > service. You want to take that risk? > > 3) Many times, there are multiple customer computers behind a NAT. Do you > really want the hassle of an irate user calling in because you just broke > the > dad's VPN to work, because one of their kids has some cruft on their > computer? > (And no, don't try to tell them they should have bought business class > service > or similar crap, that *will* lose you a customer). > > So explain why the ISP wants to cut off the user, when it will cost them > money, and possibly a customer? >

2 1

Re: CDNs should pay eyeball networks, too.
by Aleksi Suhonen 02 May '12

02 May '12

Morning, I have no idea what's really going on at LLNW, but I thought I'd still share an alternative view on this matter: My understanding is that LLNW is spending tons of money to upgrade some of their IXP connections to 100GbE in Europe. With that in mind, I'm not that surprised if they wish to get some new income to cover those costs. While content is king, eye balls are kings too. Go figure. -- Aleksi Suhonen / Axu TM Oy Internetworking Consulting Cellular: +358 45 670 2048 World Wide Web: www.axu.tm

4 3

IPv6 monitoring...
by Vytautas V Grigaliunas 02 May '12

02 May '12

Greetings... What are people using for IPv6 monitoring - in particular, for monitoring services such as DNS, Web, E-mail, etc. ? Nagios seems the people's choice. Any others...open source or commercial ? TIA... Vyto

5 4

Problems getting to Verisign's whois server on IPv6
by TR Shaw 02 May '12

02 May '12

Anyone else having problems getting to Verisign's whois server on IPv6? $ host com.whois-servers.net com.whois-servers.net is an alias for whois.verisign-grs.com. whois.verisign-grs.com has address 199.7.59.74 whois.verisign-grs.com has IPv6 address 2001:503:3227:1060::74 $ traceroute6 2001:503:3227:1060::74 traceroute6 to 2001:503:3227:1060::74 (2001:503:3227:1060::74) from 2001:470:5:4ed:cabc:c8ff:fea1:560c, 64 hops max, 12 byte packets 1 2001:470:5:4ed:226:bbff:fe6d:426e 0.311 ms 0.374 ms 0.260 ms 2 ipv6oitc-1.tunnel.tserv12.mia1.ipv6.he.net 21.128 ms 21.052 ms 17.389 ms 3 gige-g2-3.core1.mia1.he.net 20.055 ms 16.198 ms 22.699 ms 4 10gigabitethernet4-3.core1.atl1.he.net 40.166 ms 33.887 ms 32.547 ms 5 10gigabitethernet6-4.core1.ash1.he.net 49.821 ms 45.999 ms 52.751 ms 6 2001:504:0:2::2641:1 47.197 ms 46.748 ms 47.289 ms 7 xe-1-2-0.r1.bb-fo.chi2.vrsn.net 65.094 ms xe-0-2-0.r2.bb-fo.chi2.vrsn.net 66.441 ms xe-1-2-0.r1.bb-fo.chi2.vrsn.net 66.320 ms 8 2001:503:3227:14ff::2 66.448 ms 2001:503:3227:13ff::2 101.761 ms 86.864 ms 9 2001:503:3227:13ff::2 69.818 ms !P 2001:503:3227:14ff::2 69.311 ms !P 2001:503:3227:13ff::2 68.662 ms !P

8 9