[Cake] Re: vpn as wan interface on freshtomato, other fiber I control as wan interface for freshtomato (which uses cake)

[Sebastian Moeller <moeller0@gmx.de>]

Hi,

since you seem to be concerned about a legal issue, you really should be talking to a lawyer versed in the respective law of your country, I have a hunch, that copy right infringement is still copy right infringement independent of rate.
About the technical issue, you would need some DPI to semi-reliably detect p2p applications (sure well behaved ones might use specific known ports, but you seem to be after not-well-behaved ones as well) and then some marking/QoS trickery to steer such flows into suitable capacity tiers, and you would need to do this for ingress traffic. Side stepping the legal issues, I would look into qosify/luci-app-qosify and QoSmateb as starting points for OpenWrt, or given you seem to be acting as ISP maybe even LibreQoS (which will not solve the special-casing p2p issue by itself, but might be a decent way to manage your low/mid hundreds of users).


> On Aug 24, 2026, at 01:04, le berger des photons <thejoff@gmail.com> wrote:
> 
> Hello,
> 
> I've been using tomato for 20 years.  Just started using cake with
> freshtomato.  Most of my routers are distributing 200 to 300 gigabytes per
> day to 40 or 50 clients.
> 
> I have 9 different fibre optics, each capable of 1 gigabit/second.  I use
> less than 100 megabits per second for each ot them.
> 
> It occurs to me that I could perhaps benefit by putting a fast computer on
> one of the fibers and having the other eight fibers' routers pull their wan
> interface from that one fiber.
> 
> I saw that using cake with 8 qos priorities still left people with huge
> bandwidth on priority 8.
> 
> When I was using SQM, I set the speed limits on each class.  I would use a
> rule to limit all p2p protocols to that low speed.
> 
> Now using cake, I'm concerned that an illegal downloader could cause me
> problems.
> 
> When I knew it would take a week to get a film,  I didn't worry about it.
> 
> Please tell me if you think what I've written represents the real world at
> all as opposed to a likely massive grouping of bad assumptions and
> misunderstandings on my part.
> 
> thanks
> 
> On Mon, Aug 24, 2026 at 12:25 AM <cake-request@lists.bufferbloat.net> wrote:
> 
>> Send Cake mailing list submissions to
>>        cake@lists.bufferbloat.net
>> 
>> To subscribe or unsubscribe via email, send a message with subject or
>> body 'help' to
>>        cake-request@lists.bufferbloat.net
>> 
>> You can reach the person managing the list at
>>        cake-owner@lists.bufferbloat.net
>> 
>> When replying, please edit your Subject line so it is more specific
>> than "Re: Contents of Cake digest..."
>> 
>> Today's Topics:
>> 
>>   1. "I was able to implement AQL and ATF in ath11k. AQL specifically
>> allowed me to almost eliminate bufferbloat over Wi-Fi"
>>      (Frantisek Borsik)
>>   2. Re: "I was able to implement AQL and ATF in ath11k. AQL specifically
>> allowed me to almost eliminate bufferbloat over Wi-Fi"
>>      (Tom Herbert)
>>   3. Re: "I was able to implement AQL and ATF in ath11k. AQL specifically
>> allowed me to almost eliminate bufferbloat over Wi-Fi"
>>      (Frantisek Borsik)
>>   4. Re: "I was able to implement AQL and ATF in ath11k. AQL specifically
>> allowed me to almost eliminate bufferbloat over Wi-Fi"
>>      (Frantisek Borsik)
>> 
>> 
>> ----------------------------------------------------------------------
>> 
>> Message: 1
>> Date: Sun, 23 Aug 2026 23:54:14 +0200
>> From: Frantisek Borsik <frantisek.borsik@gmail.com>
>> Subject: [Cake] "I was able to implement AQL and ATF in ath11k.  AQL
>>        specifically allowed me to almost eliminate bufferbloat over Wi-Fi"
>> To: bloat <bloat@lists.bufferbloat.net>, Cake List
>>        <cake@lists.bufferbloat.net>,        codel@lists.bufferbloat.net,
>>        Make-Wifi-fast <make-wifi-fast@lists.bufferbloat.net>
>> Message-ID:
>>        <
>> CAJUtOOgcb1bhwCAtx3yLg3umNR9ws_NHxfpchfyRBtCZYv2Leg@mail.gmail.com>
>> Content-Type: text/plain; charset="UTF-8"
>> 
>> "Let it be in memory of Dave Taht, he would have been happy."
>> 
>> https://forum.openwrt.org/t/qualcommax-nss-build/148529/6624
>> https://forum.openwrt.org/t/aql-and-the-ath10k-is-lovely/59002/948
>> 
>> AMAZING. This is for you, Dave <3
>> 
>> All the best,
>> 
>> Frank
>> 
>> Frantisek (Frank) Borsik
>> 
>> 
>> *In loving memory of Dave Täht: *1965-2025
>> 
>> https://libreqos.io/2025/04/01/in-loving-memory-of-dave/
>> 
>> 
>> https://www.linkedin.com/in/frantisekborsik
>> 
>> Signal, Telegram, WhatsApp: +421919416714
>> 
>> iMessage, mobile: +420775230885
>> 
>> Skype: casioa5302ca
>> 
>> frantisek.borsik@gmail.com
>> 
>> 
>> ---------- Forwarded message ---------
>> From: St Pavl <mail@forum.openwrt.org>
>> Date: Sun, Aug 23, 2026 at 11:35 PM
>> Subject: OpenWrt Forum: AQL and the ath10k is *lovely*
>> To: <frantisek.borsik@gmail.com>
>> 
>> 
>> Someone replied to a topic you are Watching.
>> St Pavl <https://forum.openwrt.org/u/sppmaster> sppmaster
>> August 23
>> 
>> ka2107:
>> 
>> Is WiFi AQL supported in ath11k driver
>> 
>> vochong:
>> 
>> No, it’s not yet supported in ath11k driver.
>> 
>> It's just around the corner, thankfully to @JuliusBairaktaris
>> <https://forum.openwrt.org/u/juliusbairaktaris>.
>> 
>> Qualcommax NSS Build
>> <https://forum.openwrt.org/t/qualcommax-nss-build/148529/6624> Community
>> Builds, Projects & Packages
>> <https://forum.openwrt.org/c/community-builds-projects-packages/11>
>> 
>> I was able to implement AQL and ATF in ath11k. AQL specifically allowed me
>> to almost eliminate bufferbloat over wifi: In the PPE (hardware): Wi-Fi
>> download rate — shaped by the CPU port's token bucket via cpu_port_rate,
>> because every frame to a Wi-Fi client leaves the switch by port 0. Wi-Fi
>> upload rate — shaped by the WAN egress tbf, same as wired. Classification —
>> the ACL marks small frames into the priority band for everything the switch
>> forwards. Not in the PPE, and can't be: Which…
>> 
>> Let it be in memory of @dtaht <https://forum.openwrt.org/u/dtaht>, he
>> would
>> have been happy.
>> ------------------------------
>> 
>> Visit Topic
>> <https://forum.openwrt.org/t/aql-and-the-ath10k-is-lovely/59002/948> or
>> reply to this email to respond.
>> 
>> To unsubscribe from these emails, click here
>> <
>> https://forum.openwrt.org/email/unsubscribe/63176a97a41cd337964dd910126136ea6582eb2df8d36cb858ecae37bbf44efb
>>> 
>> .
>> 
>> ------------------------------
>> 
>> Message: 2
>> Date: Sun, 23 Aug 2026 14:57:59 -0700
>> From: Tom Herbert <tom@herbertland.com>
>> Subject: [Cake] Re: "I was able to implement AQL and ATF in ath11k.
>>        AQL specifically allowed me to almost eliminate bufferbloat over
>>        Wi-Fi"
>> To: Frantisek Borsik <frantisek.borsik@gmail.com>
>> Cc: bloat <bloat@lists.bufferbloat.net>, Cake List
>>        <cake@lists.bufferbloat.net>,        codel@lists.bufferbloat.net,
>>        Make-Wifi-fast <make-wifi-fast@lists.bufferbloat.net>
>> Message-ID:
>>        <
>> CALx6S34Qvuer6Zt19JYHYOsG-JYHozGphGNTd9B1Gk9GTimP+w@mail.gmail.com>
>> Content-Type: text/plain; charset="UTF-8"
>> 
>> Fantastic!
>> 
>> On Sun, Aug 23, 2026, 2:43 PM Frantisek Borsik <frantisek.borsik@gmail.com
>>> 
>> wrote:
>> 
>>> "Let it be in memory of Dave Taht, he would have been happy."
>>> 
>>> https://forum.openwrt.org/t/qualcommax-nss-build/148529/6624
>>> https://forum.openwrt.org/t/aql-and-the-ath10k-is-lovely/59002/948
>>> 
>>> AMAZING. This is for you, Dave <3
>>> 
>>> All the best,
>>> 
>>> Frank
>>> 
>>> Frantisek (Frank) Borsik
>>> 
>>> 
>>> *In loving memory of Dave Täht: *1965-2025
>>> 
>>> https://libreqos.io/2025/04/01/in-loving-memory-of-dave/
>>> 
>>> 
>>> https://www.linkedin.com/in/frantisekborsik
>>> 
>>> Signal, Telegram, WhatsApp: +421919416714
>>> 
>>> iMessage, mobile: +420775230885
>>> 
>>> Skype: casioa5302ca
>>> 
>>> frantisek.borsik@gmail.com
>>> 
>>> 
>>> ---------- Forwarded message ---------
>>> From: St Pavl <mail@forum.openwrt.org>
>>> Date: Sun, Aug 23, 2026 at 11:35 PM
>>> Subject: OpenWrt Forum: AQL and the ath10k is *lovely*
>>> To: <frantisek.borsik@gmail.com>
>>> 
>>> 
>>> Someone replied to a topic you are Watching.
>>> St Pavl <https://forum.openwrt.org/u/sppmaster> sppmaster
>>> August 23
>>> 
>>> ka2107:
>>> 
>>> Is WiFi AQL supported in ath11k driver
>>> 
>>> vochong:
>>> 
>>> No, it’s not yet supported in ath11k driver.
>>> 
>>> It's just around the corner, thankfully to @JuliusBairaktaris
>>> <https://forum.openwrt.org/u/juliusbairaktaris>.
>>> 
>>> Qualcommax NSS Build
>>> <https://forum.openwrt.org/t/qualcommax-nss-build/148529/6624> Community
>>> Builds, Projects & Packages
>>> <https://forum.openwrt.org/c/community-builds-projects-packages/11>
>>> 
>>> I was able to implement AQL and ATF in ath11k. AQL specifically allowed
>> me
>>> to almost eliminate bufferbloat over wifi: In the PPE (hardware): Wi-Fi
>>> download rate — shaped by the CPU port's token bucket via cpu_port_rate,
>>> because every frame to a Wi-Fi client leaves the switch by port 0. Wi-Fi
>>> upload rate — shaped by the WAN egress tbf, same as wired.
>> Classification —
>>> the ACL marks small frames into the priority band for everything the
>> switch
>>> forwards. Not in the PPE, and can't be: Which…
>>> 
>>> Let it be in memory of @dtaht <https://forum.openwrt.org/u/dtaht>, he
>>> would
>>> have been happy.
>>> ------------------------------
>>> 
>>> Visit Topic
>>> <https://forum.openwrt.org/t/aql-and-the-ath10k-is-lovely/59002/948> or
>>> reply to this email to respond.
>>> 
>>> To unsubscribe from these emails, click here
>>> <
>>> 
>> https://forum.openwrt.org/email/unsubscribe/63176a97a41cd337964dd910126136ea6582eb2df8d36cb858ecae37bbf44efb
>>>> 
>>> .
>>> _______________________________________________
>>> Cake mailing list -- cake@lists.bufferbloat.net
>>> To unsubscribe send an email to cake-leave@lists.bufferbloat.net
>>> 
>> 
>> ------------------------------
>> 
>> Message: 3
>> Date: Mon, 24 Aug 2026 00:16:32 +0200
>> From: Frantisek Borsik <frantisek.borsik@gmail.com>
>> Subject: [Cake] Re: "I was able to implement AQL and ATF in ath11k.
>>        AQL specifically allowed me to almost eliminate bufferbloat over
>>        Wi-Fi"
>> To: Tom Herbert <tom@herbertland.com>
>> Cc: bloat <bloat@lists.bufferbloat.net>, Cake List
>>        <cake@lists.bufferbloat.net>,        codel@lists.bufferbloat.net,
>>        Make-Wifi-fast <make-wifi-fast@lists.bufferbloat.net>
>> Message-ID:
>>        <CAJUtOOj1t+hX_nrx4HGFxivvshD=
>> yCC7sPsk1wm5GSnudKcBYQ@mail.gmail.com>
>> Content-Type: text/plain; charset="UTF-8"
>> 
>> And I found the guy https://github.com/JuliusBairaktaris - great stuff!
>> 
>> All the best,
>> 
>> Frank
>> 
>> Frantisek (Frank) Borsik
>> 
>> 
>> *In loving memory of Dave Täht: *1965-2025
>> 
>> https://libreqos.io/2025/04/01/in-loving-memory-of-dave/
>> 
>> 
>> https://www.linkedin.com/in/frantisekborsik
>> 
>> Signal, Telegram, WhatsApp: +421919416714
>> 
>> iMessage, mobile: +420775230885
>> 
>> Skype: casioa5302ca
>> 
>> frantisek.borsik@gmail.com
>> 
>> 
>> On Sun, Aug 23, 2026 at 11:58 PM Tom Herbert <tom@herbertland.com> wrote:
>> 
>>> Fantastic!
>>> 
>>> On Sun, Aug 23, 2026, 2:43 PM Frantisek Borsik <
>> frantisek.borsik@gmail.com>
>>> wrote:
>>> 
>>>> "Let it be in memory of Dave Taht, he would have been happy."
>>>> 
>>>> https://forum.openwrt.org/t/qualcommax-nss-build/148529/6624
>>>> https://forum.openwrt.org/t/aql-and-the-ath10k-is-lovely/59002/948
>>>> 
>>>> AMAZING. This is for you, Dave <3
>>>> 
>>>> All the best,
>>>> 
>>>> Frank
>>>> 
>>>> Frantisek (Frank) Borsik
>>>> 
>>>> 
>>>> *In loving memory of Dave Täht: *1965-2025
>>>> 
>>>> https://libreqos.io/2025/04/01/in-loving-memory-of-dave/
>>>> 
>>>> 
>>>> https://www.linkedin.com/in/frantisekborsik
>>>> 
>>>> Signal, Telegram, WhatsApp: +421919416714
>>>> 
>>>> iMessage, mobile: +420775230885
>>>> 
>>>> Skype: casioa5302ca
>>>> 
>>>> frantisek.borsik@gmail.com
>>>> 
>>>> 
>>>> ---------- Forwarded message ---------
>>>> From: St Pavl <mail@forum.openwrt.org>
>>>> Date: Sun, Aug 23, 2026 at 11:35 PM
>>>> Subject: OpenWrt Forum: AQL and the ath10k is *lovely*
>>>> To: <frantisek.borsik@gmail.com>
>>>> 
>>>> 
>>>> Someone replied to a topic you are Watching.
>>>> St Pavl <https://forum.openwrt.org/u/sppmaster> sppmaster
>>>> August 23
>>>> 
>>>> ka2107:
>>>> 
>>>> Is WiFi AQL supported in ath11k driver
>>>> 
>>>> vochong:
>>>> 
>>>> No, it’s not yet supported in ath11k driver.
>>>> 
>>>> It's just around the corner, thankfully to @JuliusBairaktaris
>>>> <https://forum.openwrt.org/u/juliusbairaktaris>.
>>>> 
>>>> Qualcommax NSS Build
>>>> <https://forum.openwrt.org/t/qualcommax-nss-build/148529/6624>
>> Community
>>>> Builds, Projects & Packages
>>>> <https://forum.openwrt.org/c/community-builds-projects-packages/11>
>>>> 
>>>> I was able to implement AQL and ATF in ath11k. AQL specifically allowed
>> me
>>>> to almost eliminate bufferbloat over wifi: In the PPE (hardware): Wi-Fi
>>>> download rate — shaped by the CPU port's token bucket via cpu_port_rate,
>>>> because every frame to a Wi-Fi client leaves the switch by port 0. Wi-Fi
>>>> upload rate — shaped by the WAN egress tbf, same as wired.
>> Classification
>>>> —
>>>> the ACL marks small frames into the priority band for everything the
>>>> switch
>>>> forwards. Not in the PPE, and can't be: Which…
>>>> 
>>>> Let it be in memory of @dtaht <https://forum.openwrt.org/u/dtaht>, he
>>>> would
>>>> have been happy.
>>>> ------------------------------
>>>> 
>>>> Visit Topic
>>>> <https://forum.openwrt.org/t/aql-and-the-ath10k-is-lovely/59002/948> or
>>>> reply to this email to respond.
>>>> 
>>>> To unsubscribe from these emails, click here
>>>> <
>>>> 
>> https://forum.openwrt.org/email/unsubscribe/63176a97a41cd337964dd910126136ea6582eb2df8d36cb858ecae37bbf44efb
>>>>> 
>>>> .
>>>> _______________________________________________
>>>> Cake mailing list -- cake@lists.bufferbloat.net
>>>> To unsubscribe send an email to cake-leave@lists.bufferbloat.net
>>>> 
>>> 
>> 
>> ------------------------------
>> 
>> Message: 4
>> Date: Mon, 24 Aug 2026 00:35:29 +0200
>> From: Frantisek Borsik <frantisek.borsik@gmail.com>
>> Subject: [Cake] Re: "I was able to implement AQL and ATF in ath11k.
>>        AQL specifically allowed me to almost eliminate bufferbloat over
>>        Wi-Fi"
>> To: Tom Herbert <tom@herbertland.com>
>> Cc: bloat <bloat@lists.bufferbloat.net>, Cake List
>>        <cake@lists.bufferbloat.net>,        codel@lists.bufferbloat.net,
>>        Make-Wifi-fast <make-wifi-fast@lists.bufferbloat.net>
>> Message-ID:
>>        <CAJUtOOjX7L98En0brnm4fiK2CZxhBntznv0k=
>> 6PB02GX+xyZQQ@mail.gmail.com>
>> Content-Type: text/plain; charset="UTF-8"
>> 
>> This guy is s!ck: https://github.com/openwrt/openwrt/pull/24806
>> Nice.
>> 
>> All the best,
>> 
>> Frank
>> 
>> Frantisek (Frank) Borsik
>> 
>> 
>> *In loving memory of Dave Täht: *1965-2025
>> 
>> https://libreqos.io/2025/04/01/in-loving-memory-of-dave/
>> 
>> 
>> https://www.linkedin.com/in/frantisekborsik
>> 
>> Signal, Telegram, WhatsApp: +421919416714
>> 
>> iMessage, mobile: +420775230885
>> 
>> Skype: casioa5302ca
>> 
>> frantisek.borsik@gmail.com
>> 
>> 
>> On Mon, Aug 24, 2026 at 12:16 AM Frantisek Borsik <
>> frantisek.borsik@gmail.com> wrote:
>> 
>>> And I found the guy https://github.com/JuliusBairaktaris - great stuff!
>>> 
>>> All the best,
>>> 
>>> Frank
>>> 
>>> Frantisek (Frank) Borsik
>>> 
>>> 
>>> *In loving memory of Dave Täht: *1965-2025
>>> 
>>> https://libreqos.io/2025/04/01/in-loving-memory-of-dave/
>>> 
>>> 
>>> https://www.linkedin.com/in/frantisekborsik
>>> 
>>> Signal, Telegram, WhatsApp: +421919416714
>>> 
>>> iMessage, mobile: +420775230885
>>> 
>>> Skype: casioa5302ca
>>> 
>>> frantisek.borsik@gmail.com
>>> 
>>> 
>>> On Sun, Aug 23, 2026 at 11:58 PM Tom Herbert <tom@herbertland.com>
>> wrote:
>>> 
>>>> Fantastic!
>>>> 
>>>> On Sun, Aug 23, 2026, 2:43 PM Frantisek Borsik <
>>>> frantisek.borsik@gmail.com> wrote:
>>>> 
>>>>> "Let it be in memory of Dave Taht, he would have been happy."
>>>>> 
>>>>> https://forum.openwrt.org/t/qualcommax-nss-build/148529/6624
>>>>> https://forum.openwrt.org/t/aql-and-the-ath10k-is-lovely/59002/948
>>>>> 
>>>>> AMAZING. This is for you, Dave <3
>>>>> 
>>>>> All the best,
>>>>> 
>>>>> Frank
>>>>> 
>>>>> Frantisek (Frank) Borsik
>>>>> 
>>>>> 
>>>>> *In loving memory of Dave Täht: *1965-2025
>>>>> 
>>>>> https://libreqos.io/2025/04/01/in-loving-memory-of-dave/
>>>>> 
>>>>> 
>>>>> https://www.linkedin.com/in/frantisekborsik
>>>>> 
>>>>> Signal, Telegram, WhatsApp: +421919416714
>>>>> 
>>>>> iMessage, mobile: +420775230885
>>>>> 
>>>>> Skype: casioa5302ca
>>>>> 
>>>>> frantisek.borsik@gmail.com
>>>>> 
>>>>> 
>>>>> ---------- Forwarded message ---------
>>>>> From: St Pavl <mail@forum.openwrt.org>
>>>>> Date: Sun, Aug 23, 2026 at 11:35 PM
>>>>> Subject: OpenWrt Forum: AQL and the ath10k is *lovely*
>>>>> To: <frantisek.borsik@gmail.com>
>>>>> 
>>>>> 
>>>>> Someone replied to a topic you are Watching.
>>>>> St Pavl <https://forum.openwrt.org/u/sppmaster> sppmaster
>>>>> August 23
>>>>> 
>>>>> ka2107:
>>>>> 
>>>>> Is WiFi AQL supported in ath11k driver
>>>>> 
>>>>> vochong:
>>>>> 
>>>>> No, it’s not yet supported in ath11k driver.
>>>>> 
>>>>> It's just around the corner, thankfully to @JuliusBairaktaris
>>>>> <https://forum.openwrt.org/u/juliusbairaktaris>.
>>>>> 
>>>>> Qualcommax NSS Build
>>>>> <https://forum.openwrt.org/t/qualcommax-nss-build/148529/6624>
>> Community
>>>>> Builds, Projects & Packages
>>>>> <https://forum.openwrt.org/c/community-builds-projects-packages/11>
>>>>> 
>>>>> I was able to implement AQL and ATF in ath11k. AQL specifically allowed
>>>>> me
>>>>> to almost eliminate bufferbloat over wifi: In the PPE (hardware): Wi-Fi
>>>>> download rate — shaped by the CPU port's token bucket via
>> cpu_port_rate,
>>>>> because every frame to a Wi-Fi client leaves the switch by port 0.
>> Wi-Fi
>>>>> upload rate — shaped by the WAN egress tbf, same as wired.
>>>>> Classification —
>>>>> the ACL marks small frames into the priority band for everything the
>>>>> switch
>>>>> forwards. Not in the PPE, and can't be: Which…
>>>>> 
>>>>> Let it be in memory of @dtaht <https://forum.openwrt.org/u/dtaht>, he
>>>>> would
>>>>> have been happy.
>>>>> ------------------------------
>>>>> 
>>>>> Visit Topic
>>>>> <https://forum.openwrt.org/t/aql-and-the-ath10k-is-lovely/59002/948>
>> or
>>>>> reply to this email to respond.
>>>>> 
>>>>> To unsubscribe from these emails, click here
>>>>> <
>>>>> 
>> https://forum.openwrt.org/email/unsubscribe/63176a97a41cd337964dd910126136ea6582eb2df8d36cb858ecae37bbf44efb
>>>>>> 
>>>>> .
>>>>> _______________________________________________
>>>>> Cake mailing list -- cake@lists.bufferbloat.net
>>>>> To unsubscribe send an email to cake-leave@lists.bufferbloat.net
>>>>> 
>>>> 
>> 
>> ------------------------------
>> 
>> Subject: Digest Footer
>> 
>> _______________________________________________
>> Cake mailing list -- cake@lists.bufferbloat.net
>> To unsubscribe send an email to cake-leave@lists.bufferbloat.net
>> 
>> 
>> ------------------------------
>> 
>> End of Cake Digest, Vol 131, Issue 6
>> ************************************
>> 
> _______________________________________________
> Cake mailing list -- cake@lists.bufferbloat.net
> To unsubscribe send an email to cake-leave@lists.bufferbloat.net