Re: [rtcweb] SRTP not mandatory-to-use

[Eric Rescorla <[e-mail]>]

On Tue, Jan 3, 2012 at 4:28 PM, Bernard Aboba <[e-mail]> wrote:
>
> Ted Hardie said:
>
>> I'm a little lost. In a gateway implemented in a back-to-back user
>> agent, won't you end up with the same illusion?
>>
>> The case I think you're talking about is this:
>>
>> UA--1<-Connection1->B2BUA/Gateway<-Connection-2->UA-2
>>
>> Do you expect that the gateway would be refuse to use SRTP on one side
>> if it intended not to use it on the other?
>
>
> [BA] If the SBC needed to enable communication with a legacy endpoint, then
> it
> might want to negotiate security compatible with that endpoint.
>
> Today there are PSTN gateways that support SRTP with SDES, but interop is
> frequently an issue (I've had to debug interop issues countless times), so
> I've often had to advise customers to turn SRTP off until an issue was
> resolved.
>
> Few PSTN gateways support any flavor of end-to-end security today (e.g.
> ZRTP, DTLS/SRTP, etc.),
> so a failover option is even more likely in that case.
>
>> If the requirement is SRTP always for WEBRTC, then a b2bUA would have
>> to run SRTP on boths ides if both UA-1 and UA-2 were WEBRTC
>> applications, but that seems to be what we want.
>
> [BA] What you're missing is what legacy systems actually implement (see the
> SIPIt reports).

In the interest of clarity, when I say that I would require SRTP for WebRTC,
what I mean is that if you wish to have your WebRTC client talk to a PSTN
gateway that doesn't support SRTP, you would need to have a gateway
in between. E.g.,

UA-1 <--- SRTP ---> WebRTC GW <--- RTP --> PSTN GW <---- PSTN

I don't know what others have in their hands, but it's not my intent to levy
a requirement for crypto through the entire RTP pathway, let alone to
endpoints which don't even speak RTP. Of course, in the case above,
whatever your UA displayed about the identity of the target would be
somewhat disappointing.

-Ekr