Re: [tcpPrague] TCP Prague under-utilising capacity when RTT scaling turned on

[Ashutosh Srivastava <[e-mail]>]

Hi Olivier,


>
> Could you confirm that this is also happening with the tip of the current
> Branch? I.e., 3b63cc0 is the "future commits" you are referring to above?
>

Yes, this is also happening with the tip of the current branch.

>
> In addition to the ss plots, could you:
> - Confirm that this happens with both the internal TCP pacing and the fq
> qdisc as pacer on the data sender
>

Yes, this happens with both internal TCP pacing and with the fq qdisc as
pacer at the sender.


> - Confirm that this happens both with and without gro/gso/tso/lro on the
> endhosts (data sender and receiver, and also the aqm node as fq does not do
> gro splitting prior to enqueue)?
>

There is still underutilisation after I turn off offloading (gro/gso/tso)
on all interfaces on all hosts but it does not look as bad as earlier.

- Confirm that the problem persists if you increase the base RTT--a simple
> netem qdisc on the reverse path to add 2-5ms should be sufficient, do not
> forget to disable gro/gso as that poorly interacts with netem.
>

At higher RTTs, there is some underutilisation at the beginning but TCP
Prague slowly reaches full bandwidth  and operates as expected thereafter.

- Log the reported CE marks by the AQM, as well as the
> delivered_ce/received_ce counters throughout the experiment on the data
> sender/receiver?
>

I do not observe any CE marks / delivered_ce packets as the link always
remains underutilised.

- Log eventual drops (dropped counter on AQM, and retransmission
> counters on the data sender).
>
>
I have uploaded additional data for experiments with commit 3b63cc0 on the
google drive I shared  earlier :
https://drive.google.com/drive/folders/1O6uEngxrDX5ipY71sjqr36lXCQBoZV--?usp=sharing

The columns of the AQM csv file are : timestamp, dropped packets, queue
backlog (bytes), queue backlog (packets)
Columns of the ss.csv file at the sender are : timestamp, ID, cwnd, SRTT
(ms), delivered, delivered_ce, pacing rate (Mbps)

I have also added the raw data files. The experiment settings were the same
as in my last mail.

Thank you,
Ashutosh.