[Cake] sometimes I worry about cobalt's effectiveness

[Cake] sometimes I worry about cobalt's effectiveness

[Dave Taht]

LONG string of tests on mikrotik at 19Mbit here:
https://forum.mikrotik.com/viewtopic.php?p=897554#p897554




-- 
I tried to build a better future, a few times:
https://wayforward.archive.org/?site=https%3A%2F%2Fwww.icei.org

Dave Täht CEO, TekLibre, LLC

Re: [Cake] sometimes I worry about cobalt's effectiveness

[Jonathan Morton]

> On 12 Dec, 2021, at 9:34 pm, Dave Taht <dave.taht@gmail.com> wrote:
> 
> LONG string of tests on mikrotik at 19Mbit here:
> https://forum.mikrotik.com/viewtopic.php?p=897554#p897554

I can't see any of the image attachments in that thread, since I'm not a member of the forum.  Is there anything that screams of being something I should be aware of?

 - Jonathan Morton

Re: [Cake] sometimes I worry about cobalt's effectiveness

[Dave Taht]

It turns out the host stack is doing something really unusual,  I
think. Not cake. Too many plots to post here, please join the forum if
you want to scratch your head with us.

On Mon, Dec 13, 2021 at 9:45 AM Jonathan Morton <chromatix99@gmail.com> wrote:
>
> > On 12 Dec, 2021, at 9:34 pm, Dave Taht <dave.taht@gmail.com> wrote:
> >
> > LONG string of tests on mikrotik at 19Mbit here:
> > https://forum.mikrotik.com/viewtopic.php?p=897554#p897554
>
> I can't see any of the image attachments in that thread, since I'm not a member of the forum.  Is there anything that screams of being something I should be aware of?
>
>  - Jonathan Morton



-- 
I tried to build a better future, a few times:
https://wayforward.archive.org/?site=https%3A%2F%2Fwww.icei.org

Dave Täht CEO, TekLibre, LLC

Re: [Cake] sometimes I worry about cobalt's effectiveness

[Dave Taht]

ok, it looks like ecn and perhaps dscp is busted on this mikrotik
release. Ton more plots, false starts, and packet captures here.

https://forum.mikrotik.com/viewtopic.php?p=897892#p897892

Also well, codel is doing better than cobalt, and SFQ at least at
these RTTs is doing really, really well.

On Mon, Dec 13, 2021 at 9:51 AM Dave Taht <dave.taht@gmail.com> wrote:
>
> It turns out the host stack is doing something really unusual,  I
> think. Not cake. Too many plots to post here, please join the forum if
> you want to scratch your head with us.
>
> On Mon, Dec 13, 2021 at 9:45 AM Jonathan Morton <chromatix99@gmail.com> wrote:
> >
> > > On 12 Dec, 2021, at 9:34 pm, Dave Taht <dave.taht@gmail.com> wrote:
> > >
> > > LONG string of tests on mikrotik at 19Mbit here:
> > > https://forum.mikrotik.com/viewtopic.php?p=897554#p897554
> >
> > I can't see any of the image attachments in that thread, since I'm not a member of the forum.  Is there anything that screams of being something I should be aware of?
> >
> >  - Jonathan Morton
>
>
>
> --
> I tried to build a better future, a few times:
> https://wayforward.archive.org/?site=https%3A%2F%2Fwww.icei.org
>
> Dave Täht CEO, TekLibre, LLC



-- 
I tried to build a better future, a few times:
https://wayforward.archive.org/?site=https%3A%2F%2Fwww.icei.org

Dave Täht CEO, TekLibre, LLC

Re: [Cake] sometimes I worry about cobalt's effectiveness

[Jonathan Morton]

> On 14 Dec, 2021, at 8:06 am, Dave Taht <dave.taht@gmail.com> wrote:
> 
> ok, it looks like ecn and perhaps dscp is busted on this mikrotik
> release. Ton more plots, false starts, and packet captures here.
> 
> https://forum.mikrotik.com/viewtopic.php?p=897892#p897892
> 
> Also well, codel is doing better than cobalt, and SFQ at least at
> these RTTs is doing really, really well.

Codel *with ECN disabled* is doing better under these conditions, based on what I can see via the Google Drive links.  This makes some sense if the ECN CE marks are being silently erased (which is *very* bad behaviour), rather than the packets carrying them being treated as dropped (as I'd expect from a wrong checksum).

Under this particular pathology, COBALT is still able to act via the BLUE algorithm, but in Cake this kicks in only when the queue first reads as full.  In other implementations of COBALT, it also triggers when the sojourn time reaches 400ms (by default).

Mikrotik - or whoever is responsible for this - needs to fix their crap so that the ECN field is processed correctly.  End of discussion.

 - Jonathan Morton

Re: [Cake] sometimes I worry about cobalt's effectiveness

[Sebastian Moeller]

Hi Jonathan

> On Dec 14, 2021, at 10:57, Jonathan Morton <chromatix99@gmail.com> wrote:
> 
>> On 14 Dec, 2021, at 8:06 am, Dave Taht <dave.taht@gmail.com> wrote:
>> 
>> ok, it looks like ecn and perhaps dscp is busted on this mikrotik
>> release. Ton more plots, false starts, and packet captures here.
>> 
>> https://forum.mikrotik.com/viewtopic.php?p=897892#p897892
>> 
>> Also well, codel is doing better than cobalt, and SFQ at least at
>> these RTTs is doing really, really well.
> 
> Codel *with ECN disabled* is doing better under these conditions, based on what I can see via the Google Drive links.  This makes some sense if the ECN CE marks are being silently erased (which is *very* bad behaviour), rather than the packets carrying them being treated as dropped (as I'd expect from a wrong checksum).
> 
> Under this particular pathology, COBALT is still able to act via the BLUE algorithm, but in Cake this kicks in only when the queue first reads as full.  In other implementations of COBALT, it also triggers when the sojourn time reaches 400ms (by default).
> 
> Mikrotik - or whoever is responsible for this - needs to fix their crap so that the ECN field is processed correctly.  End of discussion.

Could we maybe introduce a no-ecn keyword to switch cake to drop only mode? If only to help diagnose ECN issues?

Regards
	Sebastian


> 
> - Jonathan Morton
> _______________________________________________
> Cake mailing list
> Cake@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/cake

Re: [Cake] sometimes I worry about cobalt's effectiveness

[Jonathan Morton]

> On 14 Dec, 2021, at 11:59 am, Sebastian Moeller <moeller0@gmx.de> wrote:
> 
> Could we maybe introduce a no-ecn keyword to switch cake to drop only mode? If only to help diagnose ECN issues?

Inserting a firewall rule *before* Cake to force the ECN field to Not-ECT should suffice for this purpose.

 - Jonathan Morton

Re: [Cake] sometimes I worry about cobalt's effectiveness

[Jonathan Morton]

>> On 14 Dec, 2021, at 11:59 am, Sebastian Moeller <moeller0@gmx.de> wrote:
>> 
>> Could we maybe introduce a no-ecn keyword to switch cake to drop only mode? If only to help diagnose ECN issues?
> 
> Inserting a firewall rule *before* Cake to force the ECN field to Not-ECT should suffice for this purpose.

Addendum: either this firewall rule should ignore packets that are already CE-marked, *or* a second firewall rule which *drops* CE-marked packets should be inserted as well.  This avoids erasing congestion information from any upstream AQM.

 - Jonathan Morton