October 2009 - Test - lists.nanog.org

Verizon Southeast Network Map
by Jason Bertoch 05 Oct '09

05 Oct '09

We're considering adding a Verizon connection to our network in Florida, so I've been looking unsuccessfully for a map of Verizon's fiber network in the southeast to verify that I'll have diverse paths with my other providers. Does anyone know if such a map exists in a public location?

2 1

Re: Minimum IPv6 size
by Brandon Butterworth 05 Oct '09

05 Oct '09

> Is there some central repository for information on this? We usually > seem to find out about such changes out of the ARIN region a bit after > the fact. Have we not learnt from v4? If there are to be filters then they should be defined once and never changed as people will fail to update If a different allocation requirement arises then start a new prefix (v6 is big enough to handle this) and define its filter once. Do not allocate it from an existing range and expect people to adjust their filters. brandon

4 4

ICETel Costa Rica BGP Engineers
by Ken Gilmour 04 Oct '09

04 Oct '09

Hi there, Is there anyone from ICETel in Costa Rica who speaks English on this list? I need to discuss some BGP stuff and my Spanish is not the best. Failing that, anyone from RACSA or other major ISP in CR will do. Thanks! Ken

1 0

OT: iPhone Problems
by Clue Store 04 Oct '09

04 Oct '09

Mine's rebooted at leat 3 times a day sine the upgrade :( What ever happened to quality control.... http://discussions.apple.com/thread.jspa?threadID=2152619&tstart=0

2 1

Re: Minimum IPv6 size
by Brandon Butterworth 04 Oct '09

04 Oct '09

> > If there are to be filters then they should be defined once and never > > changed as people will fail to update > > Yay! We can return to classful routing again. That sure worked out well > for us the first time around. ^_^; We have already, all we're discussing now is if we do a better job of implementing it. Classful suffered from lack of space in each range and too coarse a set of fixed ranges, all fixable in v6 so perhaps it's not so bad? > So, if I need to break up my /32 into 4 /34s to cover different geographical > regions, I should instead renumber into a new range set aside for /34s > and give back the /32? And what if you need a few /48's, then you're open to others advertising some of yours too. With no organised mechanism of communicating acceptable prefix lengths to all routers [1] then we either do nothing and let anarchy rule, do something that overly constrains or do half the job and have both problems. > Perhaps we should give everyone an allocation out of each filter > range, so that they can simply number from the appropriately-classed > range; when you apply for space, you'd get a /32, a /33, a /34, a /35, > a /36, etc. all from the appropriate, statically defined ranges. > > *removes tongue from cheek* What would be most efficient for all? Semi classful or not I don't mind but it does seem pointless to have been going round the same circle for so many years. brandon [1] sbgp would be secure anarchy

1 0

Re: Dutch ISPs to collaborate and take responsibility for bottedclients
by deleskie＠gmail.com 04 Oct '09

04 Oct '09

Sounds great but who cover the costs? ------Original Message------ From: Peter Beckman To: Gadi Evron Cc: NANOG Subject: Re: Dutch ISPs to collaborate and take responsibility for bottedclients Sent: Oct 3, 2009 7:18 PM On Sat, 3 Oct 2009, Gadi Evron wrote: > The story is covered by PC mag: Thanks for the article Gadi. Honestly, I wish both my personal ISP and one of my business ISPs would do this. Though I have the technical ability to monitor my outgoing connections for such things, it's not a trivial task and requires resources I've decided not to invest in, namely a Linux PC running as my gateway with yet more software (OS, monitoring tools, etc) I need to secure and keep updated. For me to be thrilled about my ISPs monitoring my connection for "bad behavior," the ISP should: * Quickly notify the customer about the problem via email and phone * Offer the ability to view the evidence of the "bad behavior," accessible on the ISP network via the web so it can be viewed whether the connection is active or blocked, to help determine which host(s) is/are responsible * Clearly classify the type of "bad behavior" and offer both free and paid alternatives to potentially rectify the problem for those less technically inclined to self-solve the issue * Provide a short period of time (3 days) after notification and before disconnect to give an opportunity to fix the issue without service interruption * Offer a simple, automated way to get the connection re-tested and unblocked immediately (within 15 minutes) using a web service accessible even if the connection is blocked This would make me happy. What would make me angry is if they: * Simply turn the connection off with little or no notice * Provide no notification of what happened or why * Offer no evidence of why they turned the connection off to help debug * Force the customer to call customer service to ask for a retest or reconnect * Have the reconnect take multiple hours/days once approved If done right, such a treaty here in the US and elsewhere thing would be a major win for the Internet. Beckman --------------------------------------------------------------------------- Peter Beckman Internet Guy beckman(a)angryox.com http://www.angryox.com/ --------------------------------------------------------------------------- Sent from my BlackBerry device on the Rogers Wireless Network

2 1

Request for Contact - Verizon AS19262
by Shawn Marck 03 Oct '09

03 Oct '09

Can admin from Verizon AS19262 contact me off-list? Changes to our AS are not reaching Verizon AS19262. Thanks, -- Shawn Marck, shawn.marck(a)blacklotus.net

1 0

Request for Contact
by Shawn Somers 03 Oct '09

03 Oct '09

Can an EchoStar admin contact Me off-list? I'm seeing a routing loop at ECHOSTAR.edge1.SaltLakeCity1.Level3.net Shawn Somers (425) 405-0263

1 0

RE: Repeated Blacklisting / IP reputation
by Shawn Somers 03 Oct '09

03 Oct '09

I'd be more than happy to see this, with the added caveat that anyone that returned address space to ARIN that was subsequently marked as 'contaminated', should undergo a review process when attempting to obtain new address space. Charge them for the review process Anyone that intentionally uses address space in a manner that they know will cause it to become contaminated should be denied on any further address space requests. Another option, is to hit them where it matters. Assign fines and fees for churning address space and returning it as contaminated. Set the fee's on a sliding scale based on the amount of contamination and churn. the more contamination, the higher the fee. Shawn Somers Michiel Klaver wrote: --------- > > Message: 3 > Date: Tue, 15 Sep 2009 11:57:58 +0200 > From: Michiel Klaver <michiel(a)klaver.it> > Subject: RE: Repeated Blacklisting / IP reputation, replaced by > registered use > To: "Azinger, Marla" <marla.azinger(a)frontiercorp.com>, John Curran > <jcurran(a)arin.net>, "nanog(a)nanog.org" <nanog(a)nanog.org> > Message-ID: <4AAF6526.9000904(a)klaver.it> > Content-Type: text/plain; charset=UTF-8; format=flowed > > I think ARIN is no party to contact all RBL's and do any cleanup of > 'contaminated' address space. The only steps ARIN might do are: > > - When requesting address space, one should be able to indicate whether > receiving previous used address space would be unwanted or not. > > - When assigning address space, ARIN should notify receivers if it's > re-used or virgin address space. > > - When address space got returned to ARIN and there is evidence of > abuse, they have to mark that address space as 'contaminated' and only > re-assign that space to new end-users who have indicated to have no > problem with that. > > > > With kind regards, > > Michiel Klaver > IT Professional

11 15

Re: Minimum IPv6 size
by Brandon Butterworth 03 Oct '09

03 Oct '09

> > It might be worth relaxing filtering within 2001::/16. The RIPE NCC > > appears to be making /48 PI assignments from within 2001:678::/29 > > (e.g. the > > RIPE Meeting next week will be using 2001:67c:64::/48) > > Why the whole /16 rather than just that /29 and a few other blocks set > aside for /48s? Indeed, and why jumble these up, there's enough space to keep different allocation types separate and have no confusion with just a few trivial filters, universally deployed, which I suggest is the only way to stop degeneration. If one ISP deviates it creates pressure on others to accept the same. Then we're heading for another v4 mess as people will continuously push the boundary. > There are a lot of /48s in a /16, so protecting > against someone accidentally deaggregating their allocated /32 into / > 48s seems legitimate. And some will deaggregate to protect against others advertising more specifics brandon

1 0