[reSIProcate] Getting started

[Robert Sparks]

It turns out that stateless proxies are very hard to do.

You can't have them use DNS (because the 3263 rules force them to  
become stateful).
Having them add their own branch (but not keep real transaction  
state) introduces race conditions
and edge conditions that are probably can't be resolved.

In general, what people are really implementing are transaction  
stateful nodes that don't fork.
That's _probably_ what you're looking for anyhow.

What is it that you're trying to do that's making you look towards a  
stateless proxy?  If its
"I just want to get my feet wet with something simple", that's not  
the droid you're looking for.

RjS

On Mar 5, 2007, at 3:15 PM, Steve Hanka wrote:

> Regarding the flaws in stateless proxies that break the protocol:
>
> Does anyone know, in general, what these flaws are?  The RFCs
> specifically mention support for stateless proxies and I am curious as
> to what the problems are.
>
> Would anyone object if I investigated modifying the library to support
> them again?
>
> Best regards,
>
> Steve Hanka
>
>
> -----Original Message-----
> From: jason.fischl at gmail.com [mailto:jason.fischl at gmail.com] On Behalf
> Of Jason Fischl
> Sent: Monday, March 05, 2007 1:56 PM
> To: Byron Campen
> Cc: Steve Hanka; resiprocate-devel at list.resiprocate.org
> Subject: Re: [reSIProcate] Getting started
>
> The original reason that it was deprecated was for the reasons you
> listed. One of the consequences of deprecating it is that we stripped
> out the support for stateless proxies from the transaction layer of
> resip. As a result, it is no longer possible to implement a stateless
> proxy in resip without substantial changes to the library.
>
> Jason
>
>
> On 3/5/07, Byron Campen <bcampen at estacado.net> wrote:
>>
>>  As far as I know, the stuff in deprecated is the only stateless- 
>> proxy
> code
>> we have. Anyone remember why specifically this was deprecated? I am
> under
>> the impression it was because stateless proxies in general had flaws
> that
>> broke the protocol in some situations, and not any specific flaw in
> the
>> code, but I could be wrong.
>>
>> Best regards,
>> Byron Campen
>>
>>
>>
>>
>>
>> Hi,
>>
>>
>>
>> I am just getting started with reciprocate and am trying to get my
> mind
>> around the system.  For the last week I have been reading the source
> code,
>> looking at the test code and the API documentation, and so forth.  My
> goal
>> is to build a stateless proxy/message-relay server with some special
> routing
>> and message filtering capability for use in my employer's intelligent
>> network.
>>
>>
>>
>> I would like to find some sample code of a simple agent and a simple
> proxy.
>> I note that these pages are TBD in the documentation pages and that
> the
>> stateless proxy code in the source distribution has been deprecated.
> If
>> somebody will point me to such sample code, I will gladly write the
>> documentation pages once I figure out how the code works.
>>
>>
>>
>> Any suggestions?
>>
>>
>>
>> Best Regards,
>>
>>
>>
>> Steve Hanka
>>
>> Steve.Hanka at ucn.net
>>
>> shanka at ieee.org
>>
>>
>>
>>  ________________________________
>>
>>
>>  This message (including any attachments) may
>> contain confidential information intended
>> for a specific individual(s) and purpose,
>> and is protected by law. If you are not the
>> intended recipient, you should delete this
>> message.
>>
>>  Any disclosure, copying, or distribution of this
>> message, or the taking of any action based
>> on it, is strictly prohibited.
>> _______________________________________________
>> resiprocate-devel mailing list
>> resiprocate-devel at list.resiprocate.org
>> https://list.resiprocate.org/mailman/listinfo/resiprocate-devel
>>
>> _______________________________________________
>> resiprocate-devel mailing list
>> resiprocate-devel at list.resiprocate.org
>> https://list.resiprocate.org/mailman/listinfo/resiprocate-devel
>>
>>
>
>
>
>
>
> This message (including any attachments) may contain confidential  
> information intended for a specific individual(s) and purpose, and  
> is protected by law. If you are not the intended recipient, you  
> should delete this message.
>
> Any disclosure, copying, or distribution of this message, or the  
> taking of any action based on it, is strictly prohibited.
>
>
> _______________________________________________
> resiprocate-devel mailing list
> resiprocate-devel at list.resiprocate.org
> https://list.resiprocate.org/mailman/listinfo/resiprocate-devel