Re: [tsvwg] Prague requirements survey

[Ingemar Johansson S <[e-mail]>]

Hi

 

Inline [IJ3]

/Ingemar

 

From: Bob Briscoe <[e-mail]> 
Sent: den 23 februari 2021 17:50
To: Ingemar Johansson S <[e-mail]>; De Schepper,
Koen (Nokia - BE/Antwerp) <[e-mail]>; tsvwg IETF
list <[e-mail]>
Subject: Re: [tsvwg] Prague requirements survey

 

Ingemar, - see inline [BB2]

On 23/02/2021 08:45, Ingemar Johansson S wrote:

Hi Bob

Please see inline [IJ2]

 

From: Bob Briscoe  <mailto:[e-mail]> <[e-mail]> 
Sent: den 22 februari 2021 21:35
To: Ingemar Johansson S  <mailto:[e-mail]>
<[e-mail]>; De Schepper, Koen (Nokia - BE/Antwerp)
<mailto:[e-mail]>
<[e-mail]>; tsvwg IETF list
<mailto:[e-mail]> <[e-mail]>
Subject: Re: [tsvwg] Prague requirements survey

 

Ingemar,

On 10/02/2021 09:09, Ingemar Johansson S wrote:

Hi

Please see inline

/Ingemar

 

From: De Schepper, Koen (Nokia - BE/Antwerp)
<mailto:[e-mail]>
<[e-mail]> 
Sent: den 8 februari 2021 14:37
To: Ingemar Johansson S  <mailto:[e-mail]>
<[e-mail]>; tsvwg IETF list  <mailto:[e-mail]>
<[e-mail]>
Subject: RE: Prague requirements survey

 

Hi Ingemar,

 

Thanks for your contributions. I linked your doc to the
https://l4steam.github.io/#prague-requirements-compliance
<https://protect2.fireeye.com/v1/url?k=ae271499-f1bc2db2-ae275402-866038973a
15-a6d18987ee63d9e2&q=1&e=4b98878d-2d51-48b6-9518-92696e72501d&u=https%3A%2F
%2Fl4steam.github.io%2F%23prague-requirements-compliance>  web page (and
will do so for others).

[IJ] OK, great!

 

I didn't see any issues or objections mentioned to the current requirements
as specified in the draft. Does this mean you think they are all reasonable,
valid and feasible?

[IJ] There is a slight possibility that I misunderstood parts of the layout.
Even though SCReAM is only partially compliant in many cases I believe that
requirements are reasonable.  

 

Interesting observation (related to the performance optimization topic 1)
that for the control packets "RTCP is likely not using ECT(1)". Why is this
not likely? I assume this will impact the performance? Do we need to
recommend the use of ECT(1) on RTCP packets in the draft?

[IJ] Upon first glance I don't see that it is beneficial that RTCP packets
are ECT(1). But of course there is a possibility that RTCP packets go into a
queue with higher latency and that may affect performance. So. perhaps it is
reasonable that RTCP packets are ECT(1) too, but these are then to be
regarded as non queue building as it can be hard to rate control RTCP.


[BB] My knowledge is outdated, but RTCP used to be rate controlled in
multicast to prevent implosion (that was in the early days on shared
multicast, not single source). Are you saying all rate control mechanisms
have been removed? Is the problem that there's no feedback channel for
congestion indications on feedback? In 2-way RTP at least, couldn't you
infer the congestion of RTCP datagrams from the ECN on data running
alongside it? Or might they be disjoint paths?

[IJ2] Yes. It would of course be possible to set ECT(1) on RTCP/UDP/IP
packets and determine how large share of these packets that are marked. The
problem is then that there is not (currently) any RTCP feedback(on feedback)
format to carry that information back to the RTCP sender (==streaming
client) to make some rate control of the RTCP feedback possible. 
The RTCP bandwidth is indeed limited, the rule of thumb is that it should be
less than 5% of the RTP media bitrate scaled down by the number of
receivers. SCReAM is just one receiver and the RTCP bitrate is roughly 2% of
the medial bitrate. 


[BB2] Maybe what I said above didn't make sense, so I'll try saying it a
different way... 
Imagine this bidirectional scenario where X and Y are sending r-t data to
each other:

1a. X --RTP--> Y
1b. X <-RTCP-- Y
2a. X <--RTP-- Y
2b. X --RTCP-> Y

Why can't Y use the proportion of ECN on the arriving RTP stream (1a) to
infer the likely ECN on RTCP stream 2b? and
Why can't X use the proportion of ECN on the arriving RTP stream (2a) to
infer the likely ECN on RTCP stream 1b?
you
Then I thought maybe one can't be sure that 1a and 2b are actually
traversing the same path. 
Similarly for 1b and 2a.

 

[IJ3] RTP and RTCP is quite likely to traverse the same path, the difference
may of course be if RTP is L4S capable and RTCP is not, for this reason it
can make sense to have RTCP L4S capable as well. And you are right, as they
traverse the same path then the RTCP traffic is equally likely to become
marked as the RTP traffic. The question I have is how one congestion control
RTCP and if it is needed, it is after all a small fraction of the media
bitrate. Currently there is no available mechanism for this in the IETF
standards that cover RTCP (at least it is unknown to me).




Bob






Incidentally, the attitude being taken to ECT on TCP Pure ACKs in AccECN is
in the AccECN draft. Basically, the info is now there for Ack CC if needed,
but no need to use it currently.


Bob





 

Thanks,

Koen.

 

From: Ingemar Johansson S <[e-mail]
<mailto:[e-mail]> > 
Sent: Monday, February 8, 2021 10:59 AM
To: De Schepper, Koen (Nokia - BE/Antwerp)
<[e-mail]
<mailto:[e-mail]> >; tsvwg IETF list
<[e-mail] <mailto:[e-mail]> >
Cc: Ingemar Johansson S <[e-mail]
<mailto:[e-mail]> >
Subject: RE: Prague requirements survey

 

Hi

Please find attached (hopefully) a Prague requirements survey applied to
SCReAM (RFC8298 std + running code)

 

Regards
Ingemar

 

From: tsvwg <[e-mail] <mailto:[e-mail]> > On
Behalf Of De Schepper, Koen (Nokia - BE/Antwerp)
Sent: den 6 februari 2021 23:20
To: tsvwg IETF list <[e-mail] <mailto:[e-mail]> >
Subject: [tsvwg] Prague requirements survey

 

Hi all,

 

To get a better understanding on the level of consensus on the Prague
requirements, we prepared an overview document listing the L4S-ID draft
requirements specific to the CC (wider Prague requirements), as a
questionnaire towards potential CC developers. If you are developing or have
developed an L4S congestion control, you can describe the status of your
ongoing development in the second last column. If you cannot share status,
or plan-to/would implement an L4S CC, you can list what you would want to
support (see feasible). In the last column you can put any
description/limitations/remarks/explanations related to evaluations,
implementations and/or plans (will implement or will not implement). Any
expected or experienced issues and any objections/disagreements to the
requirement can be explained and colored appropriately.

 

The document can be found on following link:
https://raw.githubusercontent.com/L4STeam/l4steam.github.io/master/PragueReq
s/Prague_requirements_Compliance_and_Objections_template.docx
<https://protect2.fireeye.com/v1/url?k=d16bc960-8ef0f066-d16b89fb-86ee86bd51
07-080c65bfd839440d&q=1&e=7dbb7494-67c3-4315-88a6-325f32e4e8b1&u=https%3A%2F
%2Fraw.githubusercontent.com%2FL4STeam%2Fl4steam.github.io%2Fmaster%2FPrague
Reqs%2FPrague_requirements_Compliance_and_Objections_template.docx> 

 

As an example I filled it for the Linux TCP-Prague implementation on
following link:
https://l4steam.github.io/PragueReqs/Prague_requirements_Compliance_and_Obje
ctions_Linux_TCP-Prague.docx
<https://protect2.fireeye.com/v1/url?k=f839c5f7-a7a2fcf1-f839856c-86ee86bd51
07-29dabadc5d0e673d&q=1&e=7dbb7494-67c3-4315-88a6-325f32e4e8b1&u=https%3A%2F
%2Fl4steam.github.io%2FPragueReqs%2FPrague_requirements_Compliance_and_Objec
tions_Linux_TCP-Prague.docx> 

 

Please send your filled document to the list (Not sure if an attachment will
work, so I assume you also need to store it somewhere and send a link to it,
or send to me directly). 

 

We hope to collect many answers, understanding the position of the different
(potential) implementers and come faster to consensus.

 

Thanks,

Koen.






-- 
________________________________________________________________
Bob Briscoe                               http://bobbriscoe.net/
<https://protect2.fireeye.com/v1/url?k=9e93ec25-c108d525-9e93acbe-86fc6812c3
61-6ce26fbc5b7d3fe4&q=1&e=9dde995b-c278-4be1-a612-c70ef0067fc2&u=http%3A%2F%
2Fbobbriscoe.net%2F> 





-- 
________________________________________________________________
Bob Briscoe                               http://bobbriscoe.net/
<https://protect2.fireeye.com/v1/url?k=9c972e29-c30c1729-9c976eb2-86fc6812c3
61-770acfd11988f54b&q=1&e=67990d9c-a717-4edd-bfdc-2431e035aa8b&u=http%3A%2F%
2Fbobbriscoe.net%2F>