From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-yw1-x1136.google.com (mail-yw1-x1136.google.com [IPv6:2607:f8b0:4864:20::1136]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by lists.bufferbloat.net (Postfix) with ESMTPS id 5427E3CB37 for ; Wed, 18 Oct 2023 23:38:18 -0400 (EDT) Received: by mail-yw1-x1136.google.com with SMTP id 00721157ae682-5a8628e54d4so3229947b3.0 for ; Wed, 18 Oct 2023 20:38:18 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20230601; t=1697686697; x=1698291497; darn=lists.bufferbloat.net; h=cc:to:subject:message-id:date:from:in-reply-to:references :mime-version:from:to:cc:subject:date:message-id:reply-to; bh=N/qsHxjEoSwtgS5k5ZoBS8Q4KLRGSjrWb5fAwAj4zyk=; b=RmRQ/4l8Ci3BqKcWlkUp9fUw6nxyCmYxFVCWoFdnI5oPmObkwXQjLuyeruNdKCkX8z jLDtWpIZMAB0QSCg2yuA6r5tPX7MPIKpQzOeCb8ThUCpWY1x2m/I/DOcjjrTUUczR1LL jM8hvyOp91z1YDkDC96csqJ8A9XyeecZuKxTFbHdwD7alUVROFbEunl+c8htttDgZ1bM OjbPJTy2FcEy0lVSpO2Mej1ThYh15BaACAgsUSHHHAtpVZ3zPMbXxmbm1lpYOkkAgPLm r7hBLDFva4ecGlmgBqycBB8YJodP62nabLWTnzR34qNiZIr6E9sw74+Y3SxrXlwQhLey 0LXg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1697686697; x=1698291497; h=cc:to:subject:message-id:date:from:in-reply-to:references :mime-version:x-gm-message-state:from:to:cc:subject:date:message-id :reply-to; bh=N/qsHxjEoSwtgS5k5ZoBS8Q4KLRGSjrWb5fAwAj4zyk=; b=QCxkqsRq+5Wz1BgiboE6w/z5v3bRaa960/aXXUkGtESqq2rLb0BSYGSSKZp9WQc7yb EyumzTU2TYDy9k90nYYfBTlRnYrL0qXplYx1F2AbwQDg4Ex0ANRjYj6zfhC0cv7ft9Xn Ovd9fCPrpN2m5nj/zCsEOsgBgpO5S/AqQs8DuiyQy8iKWt/n36TM0FOOrbIZhkfnh+8h rDH7NNQMaaFOCCkkv0mPBh5f6WfNPEX8MHUlm5g3P178z9NUUXtPSqRgcEvwNK8yQQ6+ mE2uBzzu0d6weIzOhb+8ZcVLTVDXx7DgZKo/qeLmIVJ74OoJUZGUEO1xNF/hETC4xkr5 TH2w== X-Gm-Message-State: AOJu0YyFA81vlq0AEpc50XE0V4sTMjOqs5sejRYD8t/a7+1BhZjqUZqG R9Uq1G+UDuE+iKQmYN4nO25pcfLTG6mBuQpLzVo= X-Google-Smtp-Source: AGHT+IGudDcHkOtq3YTys+36SkZil72sV5FoQSRKFOclKRJ77ZeTmLxSk6y8XXHERaPnxRwLrlyQPCx377lDImpi7o4= X-Received: by 2002:a05:690c:f8f:b0:5a8:1767:1fef with SMTP id df15-20020a05690c0f8f00b005a817671fefmr525641ywb.9.1697686697477; Wed, 18 Oct 2023 20:38:17 -0700 (PDT) MIME-Version: 1.0 References: In-Reply-To: From: dan Date: Wed, 18 Oct 2023 21:38:06 -0600 Message-ID: To: David Lang Cc: thejoff@mail.com, libreqos , Dave Taht Content-Type: multipart/alternative; boundary="0000000000008b0a340608097cd7" Subject: Re: [LibreQoS] [NNagain] shared wireless access X-BeenThere: libreqos@lists.bufferbloat.net X-Mailman-Version: 2.1.20 Precedence: list List-Id: Many ISPs need the kinds of quality shaping cake can do List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 19 Oct 2023 03:38:18 -0000 --0000000000008b0a340608097cd7 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Oh man, I love a good mesh network chat! The term mesh has been mis-applied often, most consumer mesh devices are not mesh, they are self-organizing trees but the branches are not actively connected. I describe this as a monkey in a tree. The monkey/user is only ever on one branch at a time. A true mesh has multiple connected paths and is represented by a spider web. With that out of the way... I run a 5Ghz mesh in production delivering 50Mbps services. This network currently lives behind libreqos. It's a hybrid with a 5Ghz mesh medium and 60Ghz backhaul links (mikrotik wap 60 or cube 60s) piping fresh hot bits deeper into the mesh. It's single radio, WDS dynamic based, limited by the hardware (mikrotik) and the 'ac radios. I'm absolutely dying for a dual radio WiFi6 product with beamforming antennas and either WDS or 802.11s mode that I can get openwrt on so I can run batman-adv. Even higher on my list is a pure mmwave mesh product. real mesh, not the self-organizing tree model like unifi 'mesh' uses. This could be single radio and rely on the complete lack of nlos to keep noise down but dual radio even better. I have some mikrotik script solutions for bridging individual VLANs down from 802.11'ad links into a device that had those VLANs all pre-configured in batman-adv but I lack the 360 degree AP for it... and I have too many projects going on already... On Wed, Oct 18, 2023 at 6:48=E2=80=AFPM David Lang via LibreQoS < libreqos@lists.bufferbloat.net> wrote: > A note on mesh networks. > > a true mesh where any node can pop up and self organizes to join the > network > runs into the hidden transmitter problem (you have stations A, B and C, A > and C > can hear B but can't hear each other so as they try to communicate, they > step on > each other so that B can't understand either of them) > > relaying on the same band (let alone on the same channel) doesn't scale. > > pure wifi connected nodes where you use 2.4GHz channels for user access > and 5GHz > channels for infrastructure links can work very well, but you need to pla= n > your > channel allocations properly, have nodes with more than one 5GHz radio (o= r > you > run into the hidden transmitter problem again on the infrastructure side)= , > and > use wires to connect the infrastructure nodes when you can. > > David Lang > > On Sat, 14 Oct 2023, Dave Taht via Nnagain wrote: > > > Date: Sat, 14 Oct 2023 20:54:14 -0700 > > From: Dave Taht via Nnagain > > To: thejoff@mail.com, libreqos > > Cc: Dave Taht > > Subject: [NNagain] shared wireless access > > > > Sorry about that. this list is intended to be more political in scope. > > Moving the nnagain list to the bcc. > > > > The libreqos list and chatroom has some smart wisps on it, but your > > conditions are a bit vague? Could you describe your scenario more > > fully? > > > > There are many mesh network types out there. I am most familiar with > > openwrt-derived solutions on pre-802.11ax. It is possible using older > > wifi tech (that supports adhoc mode) to build a highly redundant mesh > > of connecting everything to everything. Examples of that include > > guifi, wlan-slovinia, nyc-mesh, freifunk, etc. But people tend to > > rebuild/reflash their own routers to support it fully. > > > > If you are talking about having more than one exit node from a mesh, > > the babel protocol is sometimes used for that. (not clear what you are > > asking) > > > > Other examples today might include ubnt's mesh stuff, or tarana's NLOS > > wan, or tachyon... but there are many products in the market, and help > > can be found in a couple different chatrooms. > > > > > > On Sat, Oct 14, 2023 at 8:41=E2=80=AFPM le berger des photons via Nnaga= in > > wrote: > >> > >> as interesting as this all is, this wasn't the discussion I'm looking > for. Perhaps you know of somewhere I can go to find what I'm looking for= . > I'm looking to figure out how to share two different accesses among the > same group of clients depending on varying conditions of the main wifi > links which serve them all. Thanks for any direction. > >> > >> On Sun, Oct 15, 2023 at 2:25=E2=80=AFAM Dave Cohen via Nnagain < > nnagain@lists.bufferbloat.net> wrote: > >>> > >>> I=E2=80=99m a couple years removed from dealing with this on the prov= ider side > but the focus has shifted rapidly to adding core capacity and large > capacity ports to the extent that smaller capacity ports like 1 Gbps aren= =E2=80=99t > going to see much more price compression. Cost per bit will come down at > higher tiers but there simply isn=E2=80=99t enough focus at lower levels = at the > hardware providers to afford carriers more price compression at 1 Gbps, > even 10 Gbps. I would expect further price compression in access costs bu= t > not really in transit costs below 10 Gbps. > >>> > >>> In general I agree that IXs continue to proliferate relative to > quantity, throughput and geographic reach, almost to the degree that > mainland Europe has been covered for years. In my home market of Atlanta, > I=E2=80=99m aware of at least four IXs that have been established here or= entered > the market in the last three years - there were only two major ones prior > to that. This is a net positive for a wide variety of reasons but I don= =E2=80=99t > think it=E2=80=99s created much of an impact in terms of pulling down tra= nsit > prices. There are a few reasons for this, but primarily because that grow= th > hasn=E2=80=99t really displaced transit demand (at least in my view) and = has really > been more about a relatively stable set of IX participants creating more > resiliency and driving other performance improvements in that leg of the > peering ecosystem. > >>> > >>> Dave Cohen > >>> craetdave@gmail.com > >>> > >>> > On Oct 14, 2023, at 7:02 PM, Dave Taht via Nnagain < > nnagain@lists.bufferbloat.net> wrote: > >>> > > >>> > =EF=BB=BFThis set of trendlines was very interesting. Unfortunately= the data > >>> > stops in 2015. Does anyone have more recent data? > >>> > > >>> > > https://drpeering.net/white-papers/Internet-Transit-Pricing-Historical-An= d-Projected.php > >>> > > >>> > I believe a gbit circuit that an ISP can resell still runs at about > >>> > $900 - $1.4k (?) in the usa? How about elsewhere? > >>> > > >>> > ... > >>> > > >>> > I am under the impression that many IXPs remain very successful, > >>> > states without them suffer, and I also find the concept of doing > micro > >>> > IXPs at the city level, appealing, and now achievable with cheap > gear. > >>> > Finer grained cross connects between telco and ISP and IXP would > lower > >>> > latencies across town quite hugely... > >>> > > >>> > PS I hear ARIN is planning on dropping the price for, and bundling = 3 > >>> > BGP AS numbers at a time, as of the end of this year, also. > >>> > > >>> > > >>> > > >>> > -- > >>> > Oct 30: > https://netdevconf.info/0x17/news/the-maestro-and-the-music-bof.html > >>> > Dave T=C3=A4ht CSO, LibreQos > >>> > _______________________________________________ > >>> > Nnagain mailing list > >>> > Nnagain@lists.bufferbloat.net > >>> > https://lists.bufferbloat.net/listinfo/nnagain > >>> _______________________________________________ > >>> Nnagain mailing list > >>> Nnagain@lists.bufferbloat.net > >>> https://lists.bufferbloat.net/listinfo/nnagain > >> > >> _______________________________________________ > >> Nnagain mailing list > >> Nnagain@lists.bufferbloat.net > >> https://lists.bufferbloat.net/listinfo/nnagain > > > > > > > > -- > > Oct 30: > https://netdevconf.info/0x17/news/the-maestro-and-the-music-bof.html > > Dave T=C3=A4ht CSO, LibreQos > > _______________________________________________ > > Nnagain mailing list > > Nnagain@lists.bufferbloat.net > > https://lists.bufferbloat.net/listinfo/nnagain > _______________________________________________ > LibreQoS mailing list > LibreQoS@lists.bufferbloat.net > https://lists.bufferbloat.net/listinfo/libreqos > --0000000000008b0a340608097cd7 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable
Oh man, I love a good mesh network=C2=A0chat!=C2=A0 The te= rm mesh has been mis-applied often, most consumer mesh devices are not mesh= , they are self-organizing trees but the branches are not actively connecte= d.=C2=A0 I describe this as a monkey in a tree.=C2=A0 The monkey/user is on= ly ever on one branch at a time.=C2=A0 A true mesh has multiple connected p= aths and is represented by a spider web.=C2=A0 With that out of the way...<= div>
I run a 5Ghz mesh in production delivering 50Mbps servic= es.=C2=A0 This network currently lives behind libreqos.=C2=A0 It's a hy= brid with a 5Ghz mesh medium and 60Ghz backhaul links (mikrotik wap 60 or c= ube 60s) piping fresh hot bits deeper into the mesh.=C2=A0 It's single = radio, WDS dynamic based, limited by the hardware (mikrotik) and the 'a= c radios.=C2=A0

I'm absolutely dying for a dual radio WiFi6 prod= uct with beamforming antennas and either WDS or 802.11s mode that I can get= openwrt on so I can run batman-adv.

Even higher on my list is a pur= e mmwave mesh product.=C2=A0 real mesh, not the self-organizing tree model = like unifi 'mesh' uses.=C2=A0 This could be single radio and rely o= n the complete lack of nlos to keep noise down but dual radio even better.= =C2=A0 =C2=A0I have some mikrotik script solutions for bridging individual = VLANs down from 802.11'ad links into a device that had those VLANs all = pre-configured in batman-adv but I lack the 360 degree AP for it... and I h= ave too many projects going on already...=C2=A0

On Wed, Oct 18, 2023= at 6:48=E2=80=AFPM David Lang via LibreQoS <libreqos@lists.bufferbloat.net> wrote:
A note on mesh network= s.

a true mesh where any node can pop up and self organizes to join the networ= k
runs into the hidden transmitter problem (you have stations A, B and C, A a= nd C
can hear B but can't hear each other so as they try to communicate, the= y step on
each other so that B can't understand either of them)

relaying on the same band (let alone on the same channel) doesn't scale= .

pure wifi connected nodes where you use 2.4GHz channels for user access and= 5GHz
channels for infrastructure links can work very well, but you need to plan = your
channel allocations properly, have nodes with more than one 5GHz radio (or = you
run into the hidden transmitter problem again on the infrastructure side), = and
use wires to connect the infrastructure nodes when you can.

David Lang

On Sat, 14 Oct 2023, Dave Taht via Nnagain wrote:

> Date: Sat, 14 Oct 2023 20:54:14 -0700
> From: Dave Taht via Nnagain <nnagain@lists.bufferbloat.net>
> To: thejoff@mail= .com, libreqos <libreqos@lists.bufferbloat.net>
> Cc: Dave Taht <dave.taht@gmail.com>
> Subject: [NNagain] shared wireless access
>
> Sorry about that. this list is intended to be more political in scope.=
> Moving the nnagain list to the bcc.
>
> The libreqos list and chatroom has some smart wisps on it, but your > conditions are a bit vague? Could you describe your scenario more
> fully?
>
> There are many mesh network types out there. I am most familiar with > openwrt-derived solutions on pre-802.11ax. It is possible using older<= br> > wifi tech (that supports adhoc mode) to build a highly redundant mesh<= br> > of connecting everything to everything. Examples of that include
> guifi, wlan-slovinia, nyc-mesh, freifunk, etc. But people tend to
> rebuild/reflash their own routers to support it fully.
>
> If you are talking about having more than one exit node from a mesh, > the babel protocol is sometimes used for that. (not clear what you are=
> asking)
>
> Other examples today might include ubnt's mesh stuff, or tarana= 9;s NLOS
> wan, or tachyon... but there are many products in the market, and help=
> can be found in a couple different chatrooms.
>
>
> On Sat, Oct 14, 2023 at 8:41=E2=80=AFPM le berger des photons via Nnag= ain
> <nnagain@lists.bufferbloat.net> wrote:
>>
>> as interesting as this all is,=C2=A0 this wasn't the discussio= n I'm looking for.=C2=A0 Perhaps you know of somewhere I can go to find= what I'm looking for.=C2=A0 I'm looking to figure out how to share= two different accesses among the same group of clients depending on varyin= g conditions of the main wifi links which serve them all.=C2=A0 Thanks for = any direction.
>>
>> On Sun, Oct 15, 2023 at 2:25=E2=80=AFAM Dave Cohen via Nnagain <= ;nnagain= @lists.bufferbloat.net> wrote:
>>>
>>> I=E2=80=99m a couple years removed from dealing with this on t= he provider side but the focus has shifted rapidly to adding core capacity = and large capacity ports to the extent that smaller capacity ports like 1 G= bps aren=E2=80=99t going to see much more price compression. Cost per bit w= ill come down at higher tiers but there simply isn=E2=80=99t enough focus a= t lower levels at the hardware providers to afford carriers more price comp= ression at 1 Gbps, even 10 Gbps. I would expect further price compression i= n access costs but not really in transit costs below 10 Gbps.
>>>
>>> In general I agree that IXs continue to proliferate relative t= o quantity, throughput and geographic reach, almost to the degree that main= land Europe has been covered for years. In my home market of Atlanta, I=E2= =80=99m aware of at least four IXs that have been established here or enter= ed the market in the last three years - there were only two major ones prio= r to that. This is a net positive for a wide variety of reasons but I don= =E2=80=99t think it=E2=80=99s created much of an impact in terms of pulling= down transit prices. There are a few reasons for this, but primarily becau= se that growth hasn=E2=80=99t really displaced transit demand (at least in = my view) and has really been more about a relatively stable set of IX parti= cipants creating more resiliency and driving other performance improvements= in that leg of the peering ecosystem.
>>>
>>> Dave Cohen
>>> craet= dave@gmail.com
>>>
>>> > On Oct 14, 2023, at 7:02 PM, Dave Taht via Nnagain <nnagain@li= sts.bufferbloat.net> wrote:
>>> >
>>> > =EF=BB=BFThis set of trendlines was very interesting. Unf= ortunately the data
>>> > stops in 2015. Does anyone have more recent data?
>>> >
>>> > https://drpeering.net/white-papers/Internet-Transit-Pricing-Historica= l-And-Projected.php
>>> >
>>> > I believe a gbit circuit that an ISP can resell still run= s at about
>>> > $900 - $1.4k (?) in the usa? How about elsewhere?
>>> >
>>> > ...
>>> >
>>> > I am under the impression that many IXPs remain very succ= essful,
>>> > states without them suffer, and I also find the concept o= f doing micro
>>> > IXPs at the city level, appealing, and now achievable wit= h cheap gear.
>>> > Finer grained cross connects between telco and ISP and IX= P would lower
>>> > latencies across town quite hugely...
>>> >
>>> > PS I hear ARIN is planning on dropping the price for, and= bundling 3
>>> > BGP AS numbers at a time, as of the end of this year, als= o.
>>> >
>>> >
>>> >
>>> > --
>>> > Oct 30: https:= //netdevconf.info/0x17/news/the-maestro-and-the-music-bof.html
>>> > Dave T=C3=A4ht CSO, LibreQos
>>> > _______________________________________________
>>> > Nnagain mailing list
>>> > Nnagain@lists.bufferbloat.net
>>> > https://lists.bufferbloat.net/listin= fo/nnagain
>>> _______________________________________________
>>> Nnagain mailing list
>>> Nnagain@lists.bufferbloat.net
>>> https://lists.bufferbloat.net/listinfo/nn= again
>>
>> _______________________________________________
>> Nnagain mailing list
>> Nnagain@lists.bufferbloat.net
>> https://lists.bufferbloat.net/listinfo/nnagai= n
>
>
>
> --
> Oct 30: https://netdevconf.= info/0x17/news/the-maestro-and-the-music-bof.html
> Dave T=C3=A4ht CSO, LibreQos
> _______________________________________________
> Nnagain mailing list
> Nna= gain@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/nnagain
_______________________________________________
LibreQoS mailing list
LibreQo= S@lists.bufferbloat.net
https://lists.bufferbloat.net/listinfo/libreqos
--0000000000008b0a340608097cd7--