Why do some companies get depeered and some don't?

Patrick W. Gilmore patrick at ianai.net
Wed Nov 5 15:05:55 UTC 2008


On Nov 5, 2008, at 6:14 AM, Jasper Bryant-Greene wrote:

> Isn't it because the receiver is more likely to backhaul the traffic  
> further, due to hot-potato routing - at least in the case of large  
> networks with multiple points of interconnect?

That's the reason given.  One can argue over whether it is the "real"  
reason.

Since we just had a long and thorough discussion of this in the last  
few days on this very list, perhaps the people who are wondering about  
this could read the archives and not bother the other 10K of us on  
something we answered _yesterday_?

-- 
TTFN,
patrick



> On 5/11/2008, at 10:15 PM, Mark Foster <blakjak at blakjak.net> wrote:
>
>> I'm sure someone else must've seen it before.
>>
>> Surely even assymetric peering agreements are mutually  
>> beneficial... ISPs are also content providers, either directly or  
>> through their customers... peering is going to have a flow-on  
>> effect in terms of reducing the cost of offering content to the  
>> people you peer with too, right?
>>
>> Why all the focus on even or non-even-ness of up/down ratios in the  
>> first place?
>>
>> Mark.
>>
>> On Tue, 4 Nov 2008, Mike Lyon wrote:
>>
>>> Those with bad or uneven ratios then purchase transit and don't let
>>> themselves get depeered...
>>>
>>> On 11/1/08, Nelson Lai <nelson.lai at indiatimes.com> wrote:
>>>> What I mean is, how come networks like Teleglobe, Limelight, etc.  
>>>> don't get
>>>> depeered by others, but Cogent does? I'm sure Cogent isn't the  
>>>> only one with
>>>> bad ratios.
>>>>
>>>>
>>>> --
>>>> Hyundai to launch the i20 in India. Catch the exclusive preview on
>>>> ZigWheels.com
>>>> http://www.zigwheels.com/b2cam/newsDetails.action?name=Emb11_20080731&path=/INDT/News/Emb11_20080731&page=1&pagecount=2&utm_source=indmail&utm_medium=footer&utm_content=tracking&utm_campaign=Nletter_07oct2008_ZW
>>>>
>>>>
>>>
>>> -- 
>>> Sent from my mobile device
>>>
>>>
>>
>





More information about the NANOG mailing list