[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Sip] Comments to the TOTE draft
inline:
Hadriel Kaplan wrote:
>
>> -----Original Message----- From: Paul Kyzivat
>> [mailto:pkyzivat at cisco.com] Sent: Thursday, February 28, 2008 1:09
>> PM
>>
>> Advantages of using MSRP are: ... - it has some mechanisms to
>> mitigate the head of line blocking issue
>
> MSRP needs that because it creates the problem for itself (or rather
> allows it to happen), because it lets multiple logical use or user
> connections be aggregated into one TCP connection. I'm not saying
> that's good/bad, but for TOTE that wouldn't be an issue needing to be
> solved, afaict.
Correct. There is no sharing of connections between users, so any HOL
blocking is caused by a user's own problem. This is why the draft says
to, based on purpose, use different params. So basically tote lets the
transport layer solve all these problems.
>
>
>> The heavier weight of MSRP is an issue for UAs that won't otherwise
>> be implementing MSRP anyway. For those that do its probably not an
>> issue. OTOH, more uses for the protocol may serve to encourage
>> ubiquitous implementation.
>
> If the idea is to provide a media-path-channel alternative for some
> of the more media-applicable uses of INFO in the wild, the hurdle of
> MSRP would be immense. Making vendors choose between either
> implementing SIP+INFO, or SIP+MSRP? That's not a hard choice to
> choose from, and they'll choose poorly.
Yup. Thats why I wanted that "T" in TOTE to be "Trivial".
>
>
>> OTOH, the use of media intermediaries of one sort or another for
>> MSRP is likely to be required in many cases, at least in
>> enterprises, because they have a need to log the stuff for legal or
>> other business reasons. It may well be that the same people would
>> conclude that they need to log TOTE exchanges too, because it
>> otherwise might be used to tunnel things. So it may be that e2e
>> media connections won't be allowed in any case.
>
> I have no doubt that would be needed in some cases, maybe even most
> cases. There is, however, a HUGE difference between *monitoring* TCP
> stream data payload contents, and *modifying* TCP stream data payload
> contents. MSRP requires modifying, even just for NAT traversal
> (today). Some vendors have started to implement proprietary
> extensions, to avoid that. That's not a good thing.
TOTE would not; you could just insert in the middle of the tcp
connection and shovel data back and forth without ever looking at the
payload.
Again the idea is to do this in a way that, we can easily add new
applications in the future without infrastructure change. So having a
way to do e2e, SBCs, etc. really easy, and just ONCE, was important.
-Jonathan R.
--
Jonathan D. Rosenberg, Ph.D. 499 Thornall St.
Cisco Fellow Edison, NJ 08837
Cisco, Voice Technology Group
jdrosen at cisco.com
http://www.jdrosen.net PHONE: (408) 902-3084
http://www.cisco.com
_______________________________________________
Sip mailing list https://www.ietf.org/mailman/listinfo/sip
This list is for NEW development of the core SIP Protocol
Use sip-implementors at cs.columbia.edu for questions on current sip
Use sipping at ietf.org for new developments on the application of sip