On Thu, Jun 4, 2015 at 9:19 AM, Kenneth Klette Jonassen
wrote:
>> why not just call tcp_init_cwnd_reduction()?
>
> I deferred considering the ECN implications of doing so. The code to
> start PRR was based on tcp_enter_cwr()/tcp_init_cwnd_reduction(), save
> that both of these functions ensure a c
> why not just call tcp_init_cwnd_reduction()?
I deferred considering the ECN implications of doing so. The code to
start PRR was based on tcp_enter_cwr()/tcp_init_cwnd_reduction(), save
that both of these functions ensure a call to tcp_ecn_queue_cwr().
The upcoming patch set instead exports tcp_
>> +static int window __read_mostly = 8;
>
> This is a (readonly) module parameter, but really is a constant (if you
> ever try to change it dynamically, bad things will happen in current
> code)
>
> const int window = 8;
It cannot be adjusted dynamically, but we can override the initial
window p
On Tue, May 19, 2015 at 12:16 AM, Stephen Hemminger
wrote:
>
> I am really happy to see more congestion control development.
:)
>
> Module parameters make for rather ugly tuning.
> They are ok for testing but difficult to use in real life.
CDG's throughput-delay tradeoff is affected by the para
I am really happy to see more congestion control development.
> +static int window __read_mostly = 8;
> +static bool use_shadow __read_mostly = true;
> +static bool use_tolerance __read_mostly;
> +static uint backoff_beta __read_mostly = 0.70 * 1024;
> +static uint backoff_factor __read_mostly =
On Sun, May 17, 2015 at 1:31 PM, Kenneth Klette Jonassen
wrote:
> This is a request for comments.
>
> CAIA Delay-Gradient (CDG) is a TCP congestion control that modifies
> the TCP sender in order to [1]:
>
> o Use the delay gradient as a congestion signal.
> o Back off with an average probabil
