So I guess my question is how do other methods proposed, for example SIPDHT2
work in this regard? How do they propose to solve the holy-grail nat
traversal problem?

-- 
"The greatest challenge to any thinker is stating the problem in a way that
will allow a solution"
- Bertrand Russell

Andy Savage
Cell Phone: +852 936 34341
Skype ID: andy_savage
Linked In: http://www.linkedin.com/in/andysavage


On Thu, Jun 10, 2010 at 4:08 PM, Adrian Georgescu <[email protected]>wrote:

>
> On Jun 10, 2010, at 9:59 AM, Andy Savage wrote:
>
> Yes, I can see that with a dedicated server with an external IP this
> wouldn't be a problem but obviously when the nodes are behind NAT it becomes
> a problem.
>
> So what are the possible solutions to this issue?
>
> Obviously I could imagine there is the idea of having supernodes in the
> same way that Skype does (nodes not behind nat) and using them as
> registrars.
>
>
> Yes this supernodes idea works well in practice. Your cassandra nodes must
> also have SIP registrar and Proxy functions and properly perform NAT
> traversal for their clients.
>
> Adrian
>
> --
> "The greatest challenge to any thinker is stating the problem in a way that
> will allow a solution"
> - Bertrand Russell
>
> Andy Savage
> Cell Phone: +852 936 34341
> Skype ID: andy_savage
> Linked In: http://www.linkedin.com/in/andysavage
>
>
> On Thu, Jun 10, 2010 at 3:31 PM, Adrian Georgescu <[email protected]>wrote:
>
>> Yes, you can use it to replace a dedicated registrar server, yet the real
>> problem is not storing the location information, is traversing the NAT for
>> SIP signaling and media. The distributed location server is just part of the
>> solution. #goodidea
>>
>> Adrian
>>
>> On Jun 10, 2010, at 9:18 AM, Andy Savage wrote:
>>
>> Hi there,
>>
>> I was thinking about using Cassandra for SIP location lookups.
>>
>> Does anybody have any ideas on whether this would be a good/bad idea? It
>> seems to be good for storing Key/Value pairs (and other data too like
>> presence) but I wasn't sure as it doesn't seem to be talked about much on
>> this list.
>>
>> Any thoughts from the experts?
>>
>> http://cassandra.apache.org/
>>
>> Cheers,
>> Andy
>> <http://cassandra.apache.org/>
>> --
>> "The greatest challenge to any thinker is stating the problem in a way
>> that will allow a solution"
>> - Bertrand Russell
>>
>> Andy Savage
>> Cell Phone: +852 936 34341
>> Skype ID: andy_savage
>> Linked In: http://www.linkedin.com/in/andysavage
>>  _______________________________________________
>> P2PSIP mailing list
>> [email protected]
>> https://www.ietf.org/mailman/listinfo/p2psip
>>
>>
>>
>
>
_______________________________________________
P2PSIP mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/p2psip

Reply via email to