March 2016 - Test - lists.nanog.org

Cisco Fabricpath
by Nicolas V 15 Mar '16

15 Mar '16

Hello, Does anyone already played with cisco fabricpath feature ? I want to use it on my nexus 5548 Is it working as easy as it seems ? No bugs / particular nx-os version... ? Thanks ! Nicolas

3 2

Juniper MX Sizing
by Graham Johnston 15 Mar '16

15 Mar '16

I am wondering if anyone can provide their real world experience about sizing Juniper MX routers as it relates to BGP. I am needing a device that has a mix of layer 2 and 3 features, including MPLS, that will have a very low port count requirement that will primarily be used at a remote POP site to connect to the local IX as well as one or two full route transit providers. The MX104 has what I need from a physical standpoint and a data plane standpoint, as well as power consumption figures. My only concern is whether the REs have enough horsepower to churn through the convergence calculations at a rate that operators in this situation would find acceptable. I realize that 'acceptable' is a moving target so I would happily accept feedback from people using them as to how long it takes and their happiness with the product. For those of you that deem the MX104 unacceptable in this kind of role and moved up to the MX240, what RE did you elect to use? Thanks, Graham Johnston Network Planner Westman Communications Group 204.717.2829 johnstong(a)westmancom.com<mailto:johnstong@westmancom.com> P think green; don't print this email.

10 15

Re: Cogent - Google - HE Fun
by Mark Tinka 15 Mar '16

15 Mar '16

On 10/Mar/16 17:51, Owen DeLong wrote: > I think it’s a little different from what you say… > > I think Google already reaches Cogent for IPv4 via transit. > > Google, long ago, announced that they would not be purchasing IPv6 transit and that they have an open peering policy for anyone who wishes to reach them. In order to avoid significant disruption, they haven’t terminated their IPv4 transit contracts, but it certainly makes sense that they would not be pursuing IPv6 transit contracts. Understandable, but sort of moot if interconnect ports are dual-stack and there are no discrete charges for IPv4 or IPv6. Of course, if removal of IPv6 traffic saves bandwidth and operational costs for Google interconnect with Cogent, then that makes sense to do. Mark.

12 16

L-Root IPv6 address renumbering
by David Soltero 15 Mar '16

15 Mar '16

This is advance notice that there is a scheduled change to the IPv6 addresses in the Root Zone for the L root-server, also known as L.ROOT-SERVERS.NET, which is administered by the ICANN. The current IP addresses for the L.ROOT-SERVERS.NET service are: 199.7.83.42 2001:500:3::42 As of March 23, 2016, the new IP addresses for the L.ROOT-SERVERS.NET service will be: 199.7.83.42 2001:500:9f::42 The change will be implemented on the root zone on March 23, 2016 2100UTC, however the new address is already operational. We encourage DNS infrastructure operators to update their DNS resolvers root "hints” file. New hints files will be available at the following URLs once the change has been formally executed on March 23, 2016: * http://www.internic.net/domain/named.root * http://www.internic.net/domain/named.cache

2 2

Re: Why the US Government has so many data centers
by Scott Weeks 14 Mar '16

14 Mar '16

--- sean(a)donelan.com wrote: From: Sean Donelan <sean(a)donelan.com> : But if a majority of the "data centers" are a single server : in a room, the cost savings of moving it to a different : room may not save billions of dollars. But no one will : remember. Many are not one, rather several. For example, in a job there were two M$ servers in a room. The guides required an AD setup (with a backup AD), 2 NTP servers with Stratum 1 (for, you know the logs in the syslog server; they need to be accurate), a router, a LAN switch (can't put a LAN switch card in the router) and a firewall. All for *two* servers. I'm not even mentioning the administrative requirements for all the boxes. It's all phunny money. Real economics are not even considered. At all. : Prediction, there will be a glowing report in a year or so : about the huge cost savings, and then a couple years later : will be an Inspector General report about problems counting : things. And he will get his money, so the counting will get done. Bigger budget, more responsibility, more scope of authority, higher pay. Not the IE just gov't managers in general. : If that's what taxpayers want, that's what they'll get. Taxpayers get no say in the process. We can chant 'more efficiency' all we want. It's just water on a duck's back. Nothing will change. Like an ant trying to turn a tank. scott

5 4

Re: Why the US Government has so many data centers
by Scott Weeks 14 Mar '16

14 Mar '16

--- sean(a)donelan.com wrote: From: Sean Donelan <sean(a)donelan.com> On Mon, 14 Mar 2016, Scott Weeks wrote: > It's all phunny money. Real economics are not even > considered. At all. : And what makes your think the Data Center Optimization : Initiative is any different, when they are counting : single servers instead of data centers? I don't think it's different. It's more of the same insanity. : If it was a rational, coherent plan; that would be great : Instead I see lots of people spending years looking for : servers, and writing reports about counting servers, and : moving servers from on room to another room. That's the world of top-heavy management. Not just in the gov't, but in many orgs. It's just worse in the gov't due to the sheer size of the workforce there. It's not rational and there're just too many middle management people fighting for their own increase in scope of authority for coherency. : What's the return on investment counting paperclips? This is the phunny money comment. In the real world we deal with real ROI or the business fails. In their world, there is no such thing. : Declare victory now. They did that a long time ago in other areas. Apparently, this is just the next one... ;-) scott

1 0

AT&T E-mail Admin
by Mike Hammett 14 Mar '16

14 Mar '16

Could someone forward me over to someone in AT&T's e-mail department? I've got an IP that's seemingly mistakenly on their RBL. It isn't on any blacklists listed at MXToolbox and the logs don't show any recent SPAM like activity. I filled out their form, but no response yet and I suspect they won't take too kindly to my "There's nothing wrong." response in it. ----- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com Midwest Internet Exchange http://www.midwest-ix.com

1 0

HP HSR Routers
by Colton Conor 14 Mar '16

14 Mar '16

Does anyone deploy HP HSR routers for full BGP routing? Looks like they have couple of routers that can hold 4 Million IPv4 routes, and do full BGP routing. I did no even know that HP had routers of this size. I read this post on Reddit, and it said the following: I would suggest looking at the HP routing line, in North America for some reason people over look them (HP's ability to get the message out is not stellar). The HSR 6602-XG will push 15 Mpps with routing table sizes of 4mil (ipv4) and 2mil (ipv6) there is no additional licensing for any feature you want to use. With respect to implementation I have always felt if you understand the protocol who gives a damn about the syntax... The MSR 4060 will handle 36 Mpps with table sizes of 1mil (ipv4) and 1mil (ipv6). Either solution will be cost effective. The person I heard about HP from manages a direct peer as a transit AS to hurricane electric with dual 10G Ethernet with a HSR 6800 (420Mpps) the throughput and feature set on their product is unreal. In a municipality's network for peering I purchased an ASR for the main site prior to learning about the MSR/HSR line and just put in a 4060 for the secondary and tertiary site they work like a charm. I think the total cost per 4060 with redundant MPU's / Power Supplies / 4 port Gig T HIMM card and 5years of support was like $15k (CDN) so like $25 USD... and not only do they go toe to toe with the QFP in the ASR for performance but I can terminate ipsec tunnels without shelling out an addition $20k!! or I can redistribute into MPBGP from my IGP without shelling out an additional 20K for the IP Enterprise liscense!! :D I would at least check em' out. Oh last thing the routers support IRF which is the HP spin on RSMLT for fabric creation (think VSS without the arbitrary limitations on which line cards will be a/a or a/p (looking at you FWSM and IDSM) so you can effectively have millisecond convergence across the routers... Also Comware is modular so the OS is identical across all products, which is kind of nice because with an ASR you have crap like VASI groups which only exist in the ASR so ya that was fun.... https://www.reddit.com/r/networking/comments/347e74/costefficient_peering_r…

3 4

IETF RFC 7707: Network Reconnaissance in IPv6 Networks
by Fernando Gont 13 Mar '16

13 Mar '16

Folks, Tim Chown and me have published RFC7707 on "Network Reconnaissance in IPv6 Networks". The RFC is available at: <https://www.rfc-editor.org/rfc/rfc7707.txt>. You can find some context for this RFC here: <http://blog.si6networks.com/2016/03/the-ietf-has-just-published-rfc-7707_12…> Thanks! Best regards, -- Fernando Gont SI6 Networks e-mail: fgont(a)si6networks.com PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492

1 0

SFP Cost Variation
by Nicholas Warren 12 Mar '16

12 Mar '16

Quick question for the experts. Why when looking at SFPs, some sites list them as $800 when the same part number can be found on places like amazon for $30-$40. What is the difference in them? Why would I buy them from a place like CDW with what appears to be a 2,000% markup. https://www.cdw.com/shop/products/Brocade-SFP-mini-GBIC-transceiver-module-G igabit-Ethernet/1411743.aspx http://www.amazon.com/gp/product/B0076Q1CTY Thanks, Nich

6 5