From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail.taht.net (mail.taht.net [IPv6:2a01:7e00:e000:2d4:f00f:f00f:b33b:b33b]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by lists.bufferbloat.net (Postfix) with ESMTPS id 3EE793B2A4 for ; Tue, 8 Jun 2021 19:52:09 -0400 (EDT) Received: from smtpclient.apple (unknown [IPv6:2600:380:453b:c8c3:846f:5a25:98ac:25b6]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.taht.net (Postfix) with ESMTPSA id 9C79E22C32; Tue, 8 Jun 2021 23:52:06 +0000 (UTC) From: Dave Taht Message-Id: Content-Type: multipart/alternative; boundary="Apple-Mail=_E608B36D-DD4F-4DF7-9B63-B69EAE6BB222" Mime-Version: 1.0 (Mac OS X Mail 14.0 \(3654.80.0.2.43\)) Date: Tue, 8 Jun 2021 16:37:15 -0700 In-Reply-To: Cc: Nils Andreas Svee , starlink@lists.bufferbloat.net, Hauke Mehrtens To: Nathan Owens References: <98e5e35f-dcca-0b1c-65c1-3da193779b5f@phrozen.org> <95CC2CF3-6EBB-4F8F-AD2C-58E791459E8D@teklibre.net> <701613ff-7a53-267f-77d0-a4e015b8410c@hauke-m.de> <6A2BF4D0-EF5E-4256-AE95-B61F12200BD8@teklibre.net> <82e37c76-c116-4fd7-9ec1-4963b1cef843@www.fastmail.com> X-Mailer: Apple Mail (2.3654.80.0.2.43) Subject: Re: [Starlink] (udm pro) Was: Re: Does the dish run OpenWrt? X-BeenThere: starlink@lists.bufferbloat.net X-Mailman-Version: 2.1.20 Precedence: list List-Id: "Starlink has bufferbloat. Bad." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 08 Jun 2021 23:52:09 -0000 --Apple-Mail=_E608B36D-DD4F-4DF7-9B63-B69EAE6BB222 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=utf-8 It always had fq_codel (=E2=80=9Csmart queues=E2=80=9D) but they buried = it in the =E2=80=9Cold gui=E2=80=9D. Their default bandwidth test = doesn=E2=80=99t set it right and they missed scaling the quantum = properly so it scaled to a gbit. toss the sqm-scripts onto it=E2=80=A6 lastly, sch_cake builds out of tree for linux versions as old as linux = 3.10. > On Jun 8, 2021, at 3:31 PM, Nathan Owens wrote: >=20 > Yep, I run the latest code: > Linux Nathan-UDMP 4.19.152-al-linux-v10.2.0-v1.10.0-12.3672-bace201 #1 = SMP Fri May 28 18:00:20 UTC 2021 aarch64 GNU/Linux >=20 > At least it has fq_codel now...=20 >=20 > On Tue, Jun 8, 2021 at 3:29 PM Nils Andreas Svee > wrote: > Nice find! >=20 > I guess someone finally managed to coax the kernel sources out of = them. Still nothing on the download page, so I guess you got to reach = out to them yourself if you'd like a copy. >=20 > Are you on the beta firmware that got upgraded to kernel 4.19 Nathan? = Wouldn't have high expectations of CAKE being enabled, but at least you = won't have to backport. >=20 > Either way the software on those things are a disappointment, so if I = ever get one it'll be to run OpenWrt on (or just plain Alpine/Debian or = similar). >=20 > Best Regards > Nils >=20 --Apple-Mail=_E608B36D-DD4F-4DF7-9B63-B69EAE6BB222 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=utf-8 It = always had fq_codel (=E2=80=9Csmart queues=E2=80=9D) but they buried it = in the =E2=80=9Cold gui=E2=80=9D. Their default bandwidth test doesn=E2=80= =99t set it right and they missed scaling the quantum properly so it = scaled to a gbit.

toss= the sqm-scripts onto it=E2=80=A6

lastly, sch_cake builds out of tree for = linux versions as old as linux 3.10.

On Jun = 8, 2021, at 3:31 PM, Nathan Owens <nathan@nathan.io> wrote:

Yep, I run the latest code:
Linux Nathan-UDMP = 4.19.152-al-linux-v10.2.0-v1.10.0-12.3672-bace201 #1 SMP Fri May 28 = 18:00:20 UTC 2021 aarch64 GNU/Linux

At least it has fq_codel now...

On Tue, Jun 8, 2021 at 3:29 PM Nils = Andreas Svee <me@lochnair.net> wrote:
Nice find!

I guess someone finally = managed to coax the kernel sources out of them. Still nothing on the = download page, so I guess you got to reach out to them yourself if you'd = like a copy.

Are you on the beta firmware that got upgraded to kernel = 4.19 Nathan? Wouldn't have high expectations of CAKE being enabled, but = at least you won't have to backport.

Either way the software = on those things are a disappointment, so if I ever get one it'll be to = run OpenWrt on (or just plain Alpine/Debian or similar).

Best Regards
Nils


= --Apple-Mail=_E608B36D-DD4F-4DF7-9B63-B69EAE6BB222--