From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-la0-f43.google.com (mail-la0-f43.google.com [209.85.215.43]) (using TLSv1 with cipher RC4-SHA (128/128 bits)) (Client CN "smtp.gmail.com", Issuer "Google Internet Authority" (verified OK)) by huchra.bufferbloat.net (Postfix) with ESMTPS id 77A5F21F0FB for ; Sat, 3 Nov 2012 17:24:58 -0700 (PDT) Received: by mail-la0-f43.google.com with SMTP id i5so4241831lah.16 for ; Sat, 03 Nov 2012 17:24:56 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type:x-gm-message-state; bh=SXwYg26ErkRQBRHS/5zrIW/3moMRs4bFGSD6sc2hTxE=; b=CFqdpeCMjgaVjQb/AL+PkXhikWu6tMuoy8js4Est4KqQbmdYXO3mTX76ehbQ+dxjbE mZMwVe8uqnXaeBlbQA14soJCsdhcorD7AIAEajI+TE0rN5Smuw/QDiEE0SaO0viq1F3f 9T5pS5Zo6vBdwAewYL5WAoBWutOltt7n9lrYXtWPP75SPKoP5dag1Ijfn5LvZ034N3T/ 3Rl+vAkdwnSpWcrGvZQ/J37kPhW8XnT7EO2duwIYM43fp6xCf9s4VPF+Szex9qdE/+yJ zWlpGUF+4hE3p1fIg+r8MSj/VqXlEZ6/P5YgeoCe9YDywwNWPnCLl9u6arRU1khoAWgb dlMg== Received: by 10.112.83.196 with SMTP id s4mr2331887lby.29.1351988696105; Sat, 03 Nov 2012 17:24:56 -0700 (PDT) MIME-Version: 1.0 Received: by 10.112.127.35 with HTTP; Sat, 3 Nov 2012 17:24:35 -0700 (PDT) In-Reply-To: References: From: Maxim Kharlamov Date: Sun, 4 Nov 2012 13:24:35 +1300 Message-ID: To: Dave Taht Content-Type: multipart/alternative; boundary=f46d040167f3f8c42904cda06536 X-Gm-Message-State: ALoCoQlWsyFJDa2RNaEbTdhmncl064kae7YrXW3HsUAXyaA8V3HzF5EmP79JFP87T6oHw9H9qnnj Cc: cerowrt-devel@lists.bufferbloat.net Subject: Re: [Cerowrt-devel] nov 3 cerowrt status X-BeenThere: cerowrt-devel@lists.bufferbloat.net X-Mailman-Version: 2.1.13 Precedence: list List-Id: Development issues regarding the cerowrt test router project List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 04 Nov 2012 00:24:59 -0000 --f46d040167f3f8c42904cda06536 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Hi Dave, Sugarland works great indeed. I've been using it on my main router since the time it was released. I don't experiment with *_codel though. My ISP advertises /48 IPv6 network prefix over DHCPv6-PD - not sure what I'm doing wrong, but I've never hit bug 413, even though I've been running IPv6 network for almost a year now. There were some minor glitches with previous versions of cerowrt, but generally it just works. Thanks, Max On 4 November 2012 12:18, Dave Taht wrote: > I'm kind of assuming that with the list as quiet as it is, that > sugarland is working well for everyone, with the exception of the > reported bugs? You've got it deployed as your main router, life is > good, things like simple_qos.sh are working....? Is anyone mucking > with efq_codel, nfq_codel? > > I've been sitting on the cerowrt-3.3 tree, waiting to rebase on the > openwrt AA final. Don't know when that will happen. > > While doing some testing of dhcpv6 on a real cable network with ipv6, > I hit bug 413. There are a few other bugs in there, and all the same > bugs we had last month. > http://www.bufferbloat.net/projects/cerowrt/issues/413 > > While we wait for openwrt AA to stablize, I have been tracking openwrt > head. Linux 3.6 support for the ar71xx architecture has arrived! > > Rather than muck with cero, it has been simpler to just do builds of > the ubnt code (picostation 2HP) as it evolves towards being > functional. There has been a lot of movement and re-organization of > the next-gen openwrt trees going on, upgrades to fundamental things > like toolchains, etc. Lot of breakage. > > I have HIGH hopes that with the Linux 3.6 TCP small queues I can > finally get router based measurements to where I trust them, and also > be able to run heavier workloads (as TSQ uses less memory) > > Moving forward: > > I've been working on a specification for some very difficult string of > network tests, that no router (including cerowrt) can pass today. If > you are interested in commenting on a draft, please email me > privately. > > I'm giving a talk at ELC this week in barcelona, about what the heck > we can do to fix wifi. I'll be all through europe from now til > December 15th, expecting to make stops in london, berlin, and paris. I > enjoyed meeting up with jesper and toke from the main bloat list, > accidentally, last week,in copenhagen, as well as running into > old^H^H^H aging gracefully embedded linux revolutionaries like matt > locke and mark ovek, now of Linaro at the same conference. > > There are some interesting homenet related meetings at ietf in Atlanta > this coming week, they are broadcast live and available via jabber. > > Mark Townsley of homenet intends to demo 8 boxes running some ospf > based prefix distribution code, I'm hoping a few cerowrt based routers > get plugged into the testbed too. Jim gettys is at ietf as well of > simon kelly of dnsmasq fame, as well as the minstrel andrew mcgregor - > good gang of debloaters there - go! go drinking! The mdnsext working > group is looking promising, too. > > Cerowrt donations continue to pour in, and I'm about 2 weeks behind on > updating the credits page. Apologies. I have a plan for the donations > - second up is a second SSD for snapon! first up was obtaining a new > server at linode to host some new stuff, which I based in london, and > have now forgotten how to access. > > I would like to get a release of cerowrt out on the 20th, with a > gui-enabled simple_qos script (needs a better name - Ceroshaper? > suggestions?), fixes for the ipv6 problem and closing the dnsmasq port > hole (bug 411) > > would love a refresh of quagga-re, ccn, and other tools, too, and to > get mdlna fixed, but the above - notably the gui to simple_qos - is > going to take all my available braincells. > > I would like to find a way, going forward, to spread around more of > the work and find more ways to speed the processes, particularly that > of CI and getting regular builds done. I've had a couple people > volunteer stuff privately, but not a lot of time to think about it. I > don't want to be the bottleneck... and I really want to get started on > 3.6 as soon as possible. > > -- > Dave T=C3=A4ht > > Fixing bufferbloat with cerowrt: > http://www.teklibre.com/cerowrt/subscribe.html > _______________________________________________ > Cerowrt-devel mailing list > Cerowrt-devel@lists.bufferbloat.net > https://lists.bufferbloat.net/listinfo/cerowrt-devel > --f46d040167f3f8c42904cda06536 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Hi Dave,

Sugarland works great indeed. I've been usi= ng it on my main router since the time it was released.
I don'= ;t experiment with *_codel though.

My ISP advertis= es /48 IPv6 network prefix over DHCPv6-PD - not sure what I'm doing wro= ng, but I've never hit bug 413, even though I've been running IPv6 = network for almost a year now.
There were some minor glitches with previous versions of cerowrt, but = generally it just works.


Thanks,Max



On 4 November 2012 12:18, Dave Taht <= dave.taht@gmail.com> wrote:
I'm kind of assuming that with the list as quiet as it is, that
sugarland is working well for everyone, with the exception of the
reported bugs? You've got it deployed as your main router, life is
good, things like simple_qos.sh are working....? Is anyone mucking
with efq_codel, nfq_codel?

I've been sitting on the cerowrt-3.3 tree, waiting to rebase on the
openwrt AA final. Don't know when that will happen.

While doing some testing of dhcpv6 on a real cable network with ipv6,
I hit bug 413. There are a few other bugs in there, and all the same
bugs we had last month.
http://www.bufferbloat.net/projects/cerowrt/issues/413

While we wait for openwrt AA to stablize, I have been tracking openwrt
head. Linux 3.6 support for the ar71xx architecture has arrived!

Rather than muck with cero, it has been simpler to just do builds of
the ubnt code (picostation 2HP) as it evolves towards being
functional. There has been a lot of movement and re-organization of
the next-gen openwrt trees going on, upgrades to fundamental things
like toolchains, etc. Lot of breakage.

I have HIGH hopes that with the Linux 3.6 TCP small queues I can
finally get router based measurements to where I trust them, and also
be able to run heavier workloads (as TSQ uses less memory)

Moving forward:

I've been working on a specification for some very difficult string of<= br> network tests, that no router (including cerowrt) can pass today. If
you are interested in commenting on a draft, please email me
privately.

I'm giving a talk at ELC this week in barcelona, about what the heck we can do to fix wifi. I'll be all through europe from now til
December 15th, expecting to make stops in london, berlin, and paris. I
enjoyed meeting up with jesper and toke from the main bloat list,
accidentally, last week,in copenhagen, as well as running into
old^H^H^H aging gracefully embedded linux revolutionaries like matt
locke and mark ovek, now of Linaro at the same conference.

There are some interesting homenet related meetings at ietf in Atlanta
this coming week, they are broadcast live and available via jabber.

Mark Townsley of homenet intends to demo 8 boxes running some ospf
based prefix distribution code, I'm hoping a few cerowrt based routers<= br> get plugged into the testbed too. Jim gettys is at ietf as well of
simon kelly of dnsmasq fame, as well as the minstrel andrew mcgregor -
good gang of debloaters there - go! go drinking! The mdnsext working
group is looking promising, too.

Cerowrt donations continue to pour in, and I'm about 2 weeks behind on<= br> updating the credits page. Apologies. I have a plan for the donations
- second up is a second SSD for snapon! first up was obtaining a new
server at linode to host some new stuff, which I based in london, and
have now forgotten how to access.

I would like to get a release of cerowrt out on the 20th, with a
gui-enabled simple_qos script (needs a better name - Ceroshaper?
suggestions?), fixes for the ipv6 problem and closing the dnsmasq port
hole (bug 411)

would love a refresh of quagga-re, ccn, and other tools, too, and to
get mdlna fixed, but the above - notably the gui to simple_qos - is
going to take all my available braincells.

I would like to find a way, going forward, to spread around more of
the work and find more ways to speed the processes, particularly that
of CI and getting regular builds done. I've had a couple people
volunteer stuff privately, but not a lot of time to think about it. I
don't want to be the bottleneck... and I really want to get started on<= br> 3.6 as soon as possible.

--
Dave T=C3=A4ht

Fixing bufferbloat with cerowrt: http://www.teklibre.com/cerowrt/subscrib= e.html
_______________________________________________
Cerowrt-devel mailing list
Cerowrt-devel@lists.= bufferbloat.net
https://lists.bufferbloat.net/listinfo/cerowrt-devel

--f46d040167f3f8c42904cda06536--