cannot access some popular websites from Linode, geolocation is wrong, ARIN is to blame?

Owen DeLong owen at delong.com
Sun Mar 3 09:35:20 UTC 2013


On Mar 2, 2013, at 16:21 , "Constantine A. Murenin" <mureninc at gmail.com> wrote:

> On 2 March 2013 15:45, Owen DeLong <owen at delong.com> wrote:
>>> Now, back to ARIN:  is Linode doing it right?  Is vr.org doing it
>>> wrong?  Are they both doing it correct, or are they both wrong?
>>> 
>> 
>> ARIN Policies do require Linode to SWIP their customer allocations,
>> so the fact that they are not doing so is actually a violation of policy.
> 
> 
> They have repeatedly disagreed, on two separate occasions, effectively
> claiming they themselves are the customers:
> 
>>>>> We now get all of our addresses assigned directly to us which is why our address is listed, however we still have some IP addresses in our pool with ...
> 
> So, the "new" addresses are all assigned directly to them.
> 

Not really...

The relevant section of the NRPM is 6.5.4 et. seq.

They are, for all practical purposes, acting as an LIR. They receive space from ARIN and then assign it to their customers (such as you).

According to 6.5.5.1, if they are assigning you a /64 or more, they must register that with ARIN through SWIP or RWHOIS.

If you are getting less than a /64, then, yes, you are out of luck.

>> Linode (and VR) probably shouldn't SWIP their datacenter blocks,
>> but this isn't really covered in policy one way or the other.
>> 
>>> Anyhow.  How do I get my geoloc to show Fremont, CA?  And to have yelp
>>> stop returning 403 Forbidden?
>> 
>> Get Linode to SWIP your blocks for starters.
> 
> Could work with IPv6, since I have a /56 from them, but I only have a
> single IPv4, so, per my understanding, an IPv4 SWIP is not possible.

Correct... For IPv4, you're kind of hosed.

Geolocation of IP addresses has always been an ill-conceived quagmire. This is not news.

However, in terms of getting the best you can out of a bad situation, the advice above is about as good as it gets.

Owen





More information about the NANOG mailing list