From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-ot0-x22d.google.com (mail-ot0-x22d.google.com [IPv6:2607:f8b0:4003:c0f::22d]) (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 414FC3B2A4 for ; Thu, 9 Mar 2017 16:21:43 -0500 (EST) Received: by mail-ot0-x22d.google.com with SMTP id o24so64905705otb.1 for ; Thu, 09 Mar 2017 13:21:43 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=aenertia.net; s=dkimaenertianet; h=mime-version:sender:in-reply-to:references:from:date:message-id :subject:to:cc; bh=KqRYRJOcHAW/mVztFFj0JvkIOF3feMrRv07jG5l6VmQ=; b=T6ZsiCYmzVzJf2OLqYuveV1b64PNJANvfz5qH6CJKpt6c3u+s8jGm/012grX3pvNBN hbJmSCx6HwwN38oHBi/gNbEClXfXiqkIP3X4t3fVzWWH9E7436BLjWP1Wdq19axWyPPy SsorDKNdDnZmUAOgul4qwCpD9ZlERt7Xrthkk= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:sender:in-reply-to:references:from :date:message-id:subject:to:cc; bh=KqRYRJOcHAW/mVztFFj0JvkIOF3feMrRv07jG5l6VmQ=; b=tQiuEV5PVdJt7fCRidb8bc6wpRIDNjPo3Yj5cnpRsSjLh2Eoi6pwUwXIYrr9Kl90hd vKzV6xSrv3Ak52ih9KDkwtrbc5oTTCpkx7ond1dIimax/H53NtaeRVC0hO5/in7MOXjm ALACpMa69zL+jttvoclcNPBdsGBRE4gmMbWaSW+KkvO/xLhxC5HlwRMfodIcLCbzgakD mXoHJjpFdKo8s7osSSE2xowwLKSSsvNq98OcpJ/bi7mqLtA5wI5NEjk86+Pw4lPvphlZ +QL+Q9INdWOIgu5KRsfzmDJv1dBrmGlOyEygNwScOVi+aFv+K3J34aosoaqAgyKB0qn9 rPgg== X-Gm-Message-State: AFeK/H3c7jeI1FGt8Xh2ex2n4EuyiAHKPtcO3cGAhf3pcALApmu5UsyPFCHyPNx9Jy3eqk3j3Bm9odE3EixOgA== X-Received: by 10.157.17.195 with SMTP id y3mr7632713oty.150.1489094502613; Thu, 09 Mar 2017 13:21:42 -0800 (PST) MIME-Version: 1.0 Sender: aenertia@aenertia.net Received: by 10.74.6.68 with HTTP; Thu, 9 Mar 2017 13:21:22 -0800 (PST) In-Reply-To: References: From: =?UTF-8?Q?Joel_Wir=C4=81mu_Pauling?= Date: Fri, 10 Mar 2017 10:21:22 +1300 X-Google-Sender-Auth: pNPlm45Z9VxRJyYxGfH_DPFKJu8 Message-ID: To: Dave Taht Cc: "cerowrt-devel@lists.bufferbloat.net" Content-Type: multipart/alternative; boundary=94eb2c18fd0add9f27054a52d46d Subject: Re: [Cerowrt-devel] edgerouter Xes and lede X-BeenThere: cerowrt-devel@lists.bufferbloat.net X-Mailman-Version: 2.1.20 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: Thu, 09 Mar 2017 21:21:43 -0000 --94eb2c18fd0add9f27054a52d46d Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable I've just bought a couple of n3160 Braswell boards as was getting sick of poor performance out of the MIP's stuff I have. If you don't care about the AES-NI instruction set's you can pickup j1900 based 4 port routers off aliexpress for under $100 (with intel 200 series NIC's no less). The n3160's you can get for around 100$ but the cheap one's only have 2 gbit ports (rtl's). The n3160's are only 6w at maximum draw and so far I've had much better results out of them than any of the MIP's or ARM stuff i've been using for the last few years. The UBN stuff has been recommended in the past but when you can shift the l3 forwarding to a much more featureful and standard x86_64 platform and just hang your Wireless as dumb switches off of that I can't really think why bothering with the esoteric stuff requiring ttl converters to get up and running is going to warrant more investment at this point. No one has yet to produce a proper x86 AC router - you can kludge one yourself with compex cards... I for one would totally buy one if such a thing were available. On 10 March 2017 at 05:45, Dave Taht wrote: > For a long time now I've had several edgerouters in production using > their default (fq_codel-enabled) vyatta based OS. There are also > backports of cake available for it. > > For 50-80 bucks, they are a *really nice quad core mips box* - capable > of forwarding at a gbit (and for that matter, can serve netperf at a > gbit) > > But dealing with their 3.10 ancient kernel and foreign configuration > interface has always been a pain for me, so yesterday I reflashed my x > and x-sfp with lede's final 17.01 release, and: > > they are *marvelous* as lede boxes. Oodles of flash. The quad core > works well. I'm mad at myself for not reflashing them long ago... > > It's a PITA to reflash them - you have to open one up, put on a 3.3v > ttl serial converter, tell uboot to load a fresh kernel via tftp, then > copy a lede image over that and sysupgrade -n to rewrite the flash... > > but after you do all that, you never have to do it again. > > (I should probably write a howto) > > I haven't gone so far as to try deploying one - I'm still benchmarking > - but as I've spent a lot of time trying to get little hackerboards to > push a gbit (only the odroid c2 can) - and for that matter, I can't > get the linksys ac1200 past 700Mbits - being able to do it on a box > that comes with a case, that runs lede well, that has a bunch of ports > on it - is a win. > > One thing that wasn't apparent to me from the marketing and my own > usage, which presented 5 separate distinct ethX interfaces in the > vyatta OS, is that, on lede, it has only a single ethernet interface > onboard, and does it's magic for forwarding with vlan tagging (it may > well be there are actual separate ethernet chips on it or the original > binary firmware emulated those, but....) > > I tested SQM to 200Mbits, that worked fine. > > -- > Dave T=C3=A4ht > Let's go make home routers and wifi faster! With better software! > http://blog.cerowrt.org > _______________________________________________ > Cerowrt-devel mailing list > Cerowrt-devel@lists.bufferbloat.net > https://lists.bufferbloat.net/listinfo/cerowrt-devel > --94eb2c18fd0add9f27054a52d46d Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
I've just bought a couple of n3160 Braswell boards as was g= etting sick of poor performance out of the MIP's stuff I have.

If = you don't care about the AES-NI instruction set's you can pickup j1= 900 based 4 port routers off aliexpress for under $100 (with intel 200 seri= es NIC's no less).

The n3160's you can get for around 100$ but= the cheap one's only have 2 gbit ports (rtl's).

The n3160'= ;s are only 6w at maximum draw and so far I've had much better results = out of them than any of the MIP's or ARM stuff i've been using for = the last few years.

The UBN stuff has been recommended in the past but= when you can shift the l3 forwarding to a much more featureful and standar= d x86_64 platform and just hang your Wireless as dumb switches off of that = I can't really think why bothering with the esoteric stuff requiring tt= l converters to get up and running is going to warrant more investment at t= his point.=C2=A0

No one has yet to produce a proper x86 =C2=A0AC route= r - you can kludge one yourself with compex cards... I for one would totall= y buy one if such a thing were available.

On 10 March 2017 at 05:45, Dave Taht <dave.taht@gmail.com> wrote:
For a long time now I've had several edgerouters in production using=
their default (fq_codel-enabled) vyatta based OS. There are also
backports of cake available for it.

For 50-80 bucks, they are a *really nice quad core mips box* - capable
of forwarding at a gbit (and for that matter, can serve netperf at a
gbit)

But dealing with their 3.10 ancient kernel and foreign configuration
interface has always been a pain for me, so yesterday I reflashed my x
and x-sfp with lede's final 17.01 release, and:

they are *marvelous* as lede boxes. Oodles of flash. The quad core
works well. I'm mad at myself for not reflashing them long ago...

It's a PITA to reflash them - you have to open one up, put on a 3.3v ttl serial converter, tell uboot to load a fresh kernel via tftp, then
copy a lede image over that and sysupgrade -n to rewrite the flash...

but after you do all that, you never have to do it again.

(I should probably write a howto)

I haven't gone so far as to try deploying one - I'm still benchmark= ing
- but as I've spent a lot of time trying to get little hackerboards to<= br> push a gbit (only the odroid c2 can) - and for that matter, I can't
get the linksys ac1200 past 700Mbits - being able to do it on a box
that comes with a case, that runs lede well, that has a bunch of ports
on it - is a win.

One thing that wasn't apparent to me from the marketing and my own
usage, which presented 5 separate distinct ethX interfaces in the
vyatta OS, is that, on lede, it has only a single ethernet interface
onboard, and does it's magic for forwarding with vlan tagging (it may well be there are actual separate ethernet chips on it or the original
binary firmware emulated those, but....)

I tested SQM to 200Mbits, that worked fine.

--
Dave T=C3=A4ht
Let's go make home routers and wifi faster! With better software!
ht= tp://blog.cerowrt.org
_______________________________________________
Cerowrt-devel mailing list
Cerowrt-devel@lists.= bufferbloat.net
https://lists.bufferbloat.net/listinfo/cero= wrt-devel

--94eb2c18fd0add9f27054a52d46d--