- NANOG - lists.nanog.org

L3 ECMP over links with different RTT
by Dikkema, Michael (Business Technology) 10 May '11

10 May '11

Hi, Is it foolish to build a L3 ECMP connection between 2 iBGP routers with one of the links having a 50% longer RTT? We're building fiber between 2 locations with the option of going both north and south around the great lakes. One link will have a 20ms RTT and one with around 35ms. The majority of our peering and transit will be on one side, with the enterprise datacenter on the other. The design right now would use OSPF as an IGP over these links with iBGP peering on loopbacks. Both links are 1G. Any thoughts on doing this different/better would be appreciated. Thanks.

3 2

Re: Cent OS migration
by Jay Ashworth 10 May '11

10 May '11

----- Original Message ----- > From: "Walter Vaughan" <wvaughan(a)steelerubber.com> > You most definately will want to make sure your user id's are > identical between the two systems, otherwise stuff like @CB will have wrong information. Excellent point. > Also, do you have any expertise maintaing a linux box? If you want > something closer > to SCO in mentality, FreeBSD and SCO have the same grandparents. Linux > is like > the cute girl that moved into town. Stuff isn't always where you > expect > to find it, > and you may get a surprise if you reach into the wrong place. Oh, don't *even* send him to BSD. CentOS and SuSE 11 are the only rational free Linuces for business use. *Any* of the BSDs are so much less well supported that they'll drive you straight up a wall. Cheers, -- jra

9 8

Suspecious anycast prefixes
by Yaoqing(Joey) Liu 10 May '11

10 May '11

Hi all, I found the following prefixes are often originated by many ASNs more than five, wonder if they provide global anycast service, if so what specific service they provide? 12.64.255.0/24 70.37.135.0/24 198.32.176.0/24 199.7.49.0/24 199.7.80.0/24 199.16.93.0/24 199.16.94.0/24 199.16.95.0/24 206.223.115.0/24 Thanks, Yaoqing

11 32

Re: Yahoo and IPv6
by Cameron Byrne 10 May '11

10 May '11

On May 9, 2011 6:11 PM, "Owen DeLong" <owen(a)delong.com> wrote: > > > On May 9, 2011, at 9:04 AM, Cameron Byrne wrote: > > > On Mon, May 9, 2011 at 8:16 AM, Arie Vayner <ariev(a)vayner.net> wrote: > >> Actually, I have just noticed a slightly more disturbing thing on the Yahoo > >> IPv6 help page... > >> > >> I have IPv6 connectivity through a HE tunnel, and I can reach IPv6 services > >> (the only issue is that my ISP's DNS is not IPv6 enabled), but I tried to > >> run the "Start IPv6 Test" tool at http://help.yahoo.com/l/us/yahoo/ipv6/ and > >> it says: > >> "We detected an issue with your IPv6 configuration. On World IPv6 Day, you > >> will have issues reaching Yahoo!, as well as your other favorite web sites. > >> We recommend disabling > >> IPv6< http://us.lrd.yahoo.com/_ylt=ArHGqIAYvt_4fpp3N3vLzmNRJ3tG/SIG=11vv8jc1f/**h… >, > >> or seeking assistance in order to fix your system's IPv6 configuration > >> through your ISP or computer manufacturer." > >> > >> What disturbs me is the piece saying "We recommend disabling > >> IPv6< http://us.lrd.yahoo.com/_ylt=ArHGqIAYvt_4fpp3N3vLzmNRJ3tG/SIG=11vv8jc1f/**h… > > >> ", with a very easy link... > >> > > > > No IPv6 is better than broken* IPv6. > > > For production, yes. However, in terms of recommendations to prepare for > IPv6 day, uh, no... The better recommendation would be to explain the > exact issue detected and suggest ways for the user to resolve it. > I hope you are not insinuating that ipv6 is not production today, many real prod sites are ds today :) If one properly working website fails over ipv6, they will all fail for that user (modulo routing table fragmentation). I would really like the world to treat v6 as real production, and not doing so is one of the major hurtles to v6 deployment. Agreed. It would be nice to diagnose and fix, there are web sites for that, but that is not yahoo's issue or scope to "the masses" who don't know and should not care about L3 protocols. The only practical solution at scale is to turn it off at the client to restore service (web page loads ... ). The other practical solution, at scale, is keep v6 off at the server .... which is, IMHO, worse. Cb > Owen >

2 1

aster.pl unwise abuse policy
by goemon＠anime.net 09 May '11

09 May '11

Anyone with contacts at aster.pl advise them of their unwise policies? Thanks. -Dan From: Abuse ASTER <abuse(a)aster.pl> ======================================================= This email was send automatically ! Do not reply to this email. ------------------------------------------------------- Dear Sir or Madam, We kindly inform you that reports of violations of the Rules and Regulations of detailed benefits of internet access by ASTER Sp. z o.o based in Warsaw made by ASTER subscribers can only be sent by the help of the form available on the website: http://abuse.aster.pl. Reports sent via E-Mail will not be processed. Sincerely, Departament of Customer Service ASTER http://www.aster.pl/ebok/ tel: 0-801-014-014 lub 022 4-014-014

2 1

Apologies, NANOG
by Jay Ashworth 09 May '11

09 May '11

I appear to have somehow cross-posted a reply that was meant to go to a different list entirely; no clue how I did that. Now I understand the vehemence of the BSD-supporters. :-) Nothing to see here. Move along. Cheers, -- jra

1 1

Re: Cent OS migration
by Jay Ashworth 09 May '11

09 May '11

---- Original Message ----- > From: "Brant I. Stevens" <branto(a)networking-architecture.com> > >CentOS and SuSE 11 are the only rational free Linuces for business > >use. > > Don't forget Scientific Linux as well. Same heritage as CentOS. "Scientific Linux? What's that?" :-) And while there's probably nothing wrong with the OS or its pedigree, that (semi-amusing) question is pertinent, I think, for someone in his position: he doesn't know much Linux, and is willing to do some of his own legwork on the support front, in exchange for not having to pay a fairly sizable yearly support charge. In that situation, the more public resources there are on which you can draw, the better off you are. For you and I, something like SL (which I do remember seeing go by on LWN a few months ago) would be a good bet. I'm not sure that's true for someone like our OP here. Cheers, -- jr 'deja vu all over again' a

1 0

Current recommendations for 2 x full bgp feed
by Support 09 May '11

09 May '11

Can anyone give me their recommendation for current hardware to take 2 x full BGP feeds over 1Gb/s ports with a third Gb port for the local network? I did this about 6/7 years ago with a Cisco 7200VXR NPE300 256MB RAM but I'm guessing things have moved on??? Thanks, Chris

10 10

Rwhois not serving all records - it is almost working though.
by Landon Stewart 08 May '11

08 May '11

Hello NANOG, I sent this information to the rwhoisd mailing list originally but I've been informed that the mailing list is mostly dead now. I hope this is not too far off-topic for NANOG. One person replied to me off-list from the rwhois mailing list and had some help but I haven't found a solution yet. Scrapping our entire rwhois implementation and starting from scratch would be a shame since I don't have that many free cycles these days. If anyone has any info or can offer some information/help that'd be super duper appreciated. Someone else installed this copy of rwhoisd and then that service was moved to a new server. The rwhois server we maintain is rwhois.hopone.net on port 4321. All of this used to work until the rwhois service and its directories were moved to a new server a few months ago. It hasn't worked quite right since then. I'm really not sure how it all works - I'm jumping into the middle of this since the person who set it up isn't around any more. I've checked for things as simple as permissions and file formatting but I can't find any problems. Anyway - The data files are exported from our customer database and look OK. By "OK" I mean they are getting exported, I'm not sure if there's a formatting issue or not. The files are generated regularly on another server that hasn't had any changes made to it so they should, in theory, be fine. For this example I'll use 66.36.235.19 as the IP address in question. This is our address. When doing a lookup I see the following which is incomplete. Actually I'd like it to not display this at all personally. I'd like the customer information to be displayed instead but multiple records is fine (ours for the netblock and theirs for their allocation). See below for more information about the data files. # whois 66.36.235.19 [Querying whois.arin.net] [Redirected to rwhois.hopone.net:4321] [Querying rwhois.hopone.net] [rwhois.hopone.net] %rwhois V-1.5:003fff:00 rwhois.hopone.net (by Network Solutions, Inc. V-1.5.9.5) network:Class-Name:network network:ID:66.36.224.0/19 network:Auth-Area:66.36.224.0/19 network:Network-Name:Superb Internet Corporation network:IP-Network:66.36.224.0/19 network:Organization;I:Superb_Internet_Corporation network:Tech-Contact:hostmaster@superb.net network:Admin-Contact:hostmaster@superb.net network:Created:20050124 network:IP-Total:8192 network:IP-Used:3578 network:IP-Available:4614 network:IP-Usage:43.68 network:Updated:20110317 network:Updated-By:rwhois@hopone.com %referral rwhois://root.rwhois.net:4321/auth-area=. %ok The data files for this 66.36.231.147 are in a directory called /usr/local/rwhoisd/net-66.36.224.0-19. The directory looks like this: # pwd /usr/local/rwhoisd/net-66.36.224.0-19 # ls -la total 40 drwxr-xr-x 3 www wheel 4096 Mar 17 00:03 . drwxr-xr-x 26 rwhois rwhois 4096 Mar 17 00:03 .. drwxr-xr-x 3 www wheel 4096 Mar 17 00:05 data -rw-r--r-- 1 www wheel 125 Mar 17 00:03 schema -rw-r--r-- 1 www wheel 181 Mar 17 00:03 soa When changing into /usr/local/rwhoisd/net-66.36.224.0-19/data/network I see the data file that contains the information that should be served which is: # cat 1230-bakertrg.txt ID: 1230.66.36.224.0/19 Auth-Area: 66.36.224.0/19 Network-Name: bakertrg Network-Block: 66.36.235.19-66.36.235.19 Organization: Baker_TRG__Inc. IP-Used: 1 Created: 20050124 Updated: 20110317 Updated-By: rwhois(a)hopone.net The local.db file at /usr/local/rwhoisd/net-66.36.224.0-19/data/network/local.db contains the following for this which I assume is correct type:DATA file:net-66.36.224.0-19/data/network/1230-bakertrg.txt file_no:0 size:220 num_recs:1 lock:OFF Does anyone on this list have any idea why this data is not being served as expected through rwhois anymore? Some ideas on what to check would be helpful. I'd like to avoid resetting this up from scratch if there's an easy fix somewhere since it seems like its *almost* working. Thanks for taking the time to read this. I appreciate any help anyone can suggest on or off list. -- Landon Stewart <LStewart(a)SUPERB.NET> SuperbHosting.Net by Superb Internet Corp. Toll Free (US/Canada): 888-354-6128 x 4199 Direct: 206-438-5879 Web hosting and more "Ahead of the Rest": http://www.superbhosting.net

3 2

open source DPI suggestions?
by Rogelio 07 May '11

07 May '11

Can anyone suggest any open source DPI (deep packet inspection) projects? I am working on various telco projects in emerging markets, but can't quite justify the price for the bigger and more well known players. :/ (Until then, I'll have to rely on some of the more well known Linux and BSD traffic shaping tools) -- Also on LinkedIn? Feel free to connect if you too are an open networker: scubacuda(a)gmail.com

8 7