Yes I have been using it successfully for the last couple of years and
it works well. However the vaapi encoder path needed for non nvidia
encoding has some bugs with how bandwidth calculation works for FEC
insertion. So YMMV for anything above around 30mbit encodes (which is
plenty for 4K streamin
Whilst I understand it's not designed for High Bandwidth uplinks - how does
it scale to 1 to 10Gbit symmetrical (or near symmetrical) uplinks ?
One of the problems i've had with Cake is that it becomes CPU bound beyond
around 3 or 4 gigabit.
On Sat, 16 Apr 2022 at 07:41, Andrew Somerville
wrote:
Those are really high split rates. We (as in UFB in NZ) looked at 32:1
splits but it's rare - in practice it's often half that. Splits end up
based on contention of regulated L2 plans which are sold to RSP's to
on-sell to customers. Based on available backhaul bandwidth rather than any
factor. i.e
Completely aside I have never got Cake SQM to work with connection's beyond
about a gigabit biderectional ; without loosing gigabits of throughput even
when running on beefy hardware. Has been a problem here for some time now.
On Fri, 17 Dec 2021, 9:36 pm Joel Wirāmu Pauling, wrote:
>
now...The whole SFP+ adapter concept has
> seemed to me to be a "tweener" in hardware design space. Too many failure
> points. That said, I like fiber's properties as a medium for distances.
> > >
> > >
> > > On Thursday, December 16, 2021 2:31pm, "Jo
t;
> On Thu, 16 Dec 2021, David P. Reed wrote:
>
> > Thanks, That's good to know...The whole SFP+ adapter concept has seemed
> to me to be a "tweener" in hardware design space. Too many failure points.
> That said, I like fiber's properties as a medium for distan
Heat issues you mention with UTP are gone; with the 803.bz stuff (i.e
Base-N).
It was mostly due to the 10G-Base-T spec being old and out of line with the
SFP+ spec ; which led to higher power consumption than SFP+ cages were
rated to draw and aforementioned heat problems; this is not a problem wi
I was hoping the Lawsuit with Cambria would result in something.
tl;dr Ubiquity took cambria to court because the claimed they copied stuff,
cambria responded with their own litigation by saying Ubiquity wasn't
complying with GPL.
But they settled under NDA and we all lost out.
On Thu, Mar 11, 2
I read this as basically Huawei and China led Consumers of Huawei kit.
I guess potentially if they can solve v6 fragmentation problems then I
would be for it, Segment routing, also should have been a thing for a long
time.
On Sun, 5 Jul 2020 at 09:31, Valdis Klētnieks
wrote:
> On Thu, 02 Jul 20
Good riddance; most of the current platforms are using Lantiq Wireless
which they got from some acquizition/cross-licence deal and are poorly
documented at worst and completely innoperable.
I doubt this will impact any of their Industrial IO and Edge stuff which is
actually good and anyone sensibl
Another neat thing about 400 and 800GE is that you can get MPO optics that
allow splitting a single 4x100 or 8x100 into individual 100G feeds. Good
for port density and/or adding capacity to processing/Edge/Appliances
Now there are decent ER optics for 100G you can now do 40-70KM runs of each
100G
4 Apr 2020 at 12:32, Valdis Klētnieks
wrote:
> On Sat, 04 Apr 2020 12:08:48 +1300, Joel Wirāmu Pauling said:
>
> > The Xiaomi Dafang Cameras are good value and there is a completely open
> > tool chain including uboot for them:
>
> Hopefully that security cam, having an open t
Correct.
On Sat, 4 Apr 2020 at 12:10, Jonathan Morton wrote:
> > On 4 Apr, 2020, at 2:08 am, Joel Wirāmu Pauling
> wrote:
> >
> > 128G of Ram
>
> That's somewhat more than I have in my desktop PCs. Did you me
The Xiaomi Dafang Cameras are good value and there is a completely open
tool chain including uboot for them:
Get one with 128G of Ram if you want to do 1080P RTSP streams.
https://github.com/EliasKotlyar/Xiaomi-Dafang-Hacks
On Sat, 4 Apr 2020 at 11:53, David P. Reed wrote:
> The ESP32-CAM devi
Kinda sad there isn't an 802.3bz config.
-Joel
On Mon., 4 Feb. 2019, 17:46 Dave Taht, wrote:
>
> https://www.nxp.com/products/processors-and-microcontrollers/arm-based-processors-and-mcus/qoriq-layerscape-arm-processors/qoriq-lx2160a-reference-design-board:LX2160A-RDB
>
>
>
> --
>
> Dave Täht
>
Must have some seriously terrible speeds on the main connection to
make this worth while; i.e sub 100mbit.
The hassles with port forwarding/per connection tracking and having to
then optimise latency sensitive traffic using rules to prefer wired vs
wifi uplink etc really make this pretty niche use
Needs 802.3bz capable Copper ports IMNSHO.
I don't plan on buying new dev boards from until there is something with
these that isn't terrible.
On 4 August 2018 at 03:30, dpr...@deepplum.com wrote:
> https://www.cnx-software.com/2018/08/03/clearfog-gt-8k-
> high-end-networking-sbc-marvell-armada
Just a note - a lot of this mess is due to China's rapid dev cycles and
race to the bottom on cost vs supply.
Generally a fab house who is in turn contracted by an OEM in China will
have 1 maybe 2 engineers who will do the initial low level C bits required
for a product/board. They get it working
bruary 2018 at 02:21, Mikael Abrahamsson wrote:
> On Thu, 15 Feb 2018, Joel Wirāmu Pauling wrote:
>
>> Again it's not the speed, it's the throughput. TB3 delivers near to what
>> my local x86 can do in terms of throughput. Also network should never be
>> slower than d
it's late /throughput/latency/%s
On 15 February 2018 at 00:45, Joel Wirāmu Pauling wrote:
> Aquantia 10GBase-T TB3 self powered, adapters are now available. They
> support 803.11bz.
>
> Again it's not the speed, it's the throughput. TB3 delivers near to
> what m
15 February 2018 at 00:33, Mikael Abrahamsson wrote:
> On Thu, 25 Jan 2018, Joel Wirāmu Pauling wrote:
>
>> Kia Ora (Hi in Māori).
>>
>> Today I delivered my talk on 10Gbit(+) in the home at Linuxconf
>> Australasia. Some specific shout outs to those on the list who
Kia Ora (Hi in Māori).
Today I delivered my talk on 10Gbit(+) in the home at Linuxconf
Australasia. Some specific shout outs to those on the list who helped form
some of the content and especially for the continued efforts with FLENT
which I have been making extensive use of both professionally an
As I am writing up my slide-pack for LCA2018 this reminded me to test out
irtt sleep bench against my running system.
Seems either the Skylake Parts are much better in Combination with current
kernels at this than what you were running on - what is the kernel of the
x86 result?
---
aenertia@kiorew
I am planing on running thunderbolt3 network flent RRUL test pre and post
KPTI patch set.
On 12 January 2018 at 15:09, Dave Taht wrote:
> I have rebooted all 7 of the bufferbloat.net servers on linode into
> their post-meltdown-fix 4.14.12 kernel. I am pretty sure the
> underlying substrates are
I was too busy skateboarding holding on to the bumper of my Limo to take
notice obviously.
On 5 January 2018 at 11:26, Jonathan Morton wrote:
> > On 5 Jan, 2018, at 12:09 am, dpr...@deepplum.com wrote:
> >
> > I should point out here that I was one of the researchers that helped
> develop the or
principle of least privilege", which
> comes from the 1970's work on secure operating systems.
>
>
>
> I should point out here that I was one of the researchers that helped
> develop the original multi-level security systems then. Those "colored
> books" come
ems. It's a hardware bug that lets any userspace
> process access anything the kernel can address.
>
>
>
> -Original Message-----
> From: "Joel Wirāmu Pauling"
> Sent: Thursday, January 4, 2018 4:52pm
> To: "Dave Taht"
> Cc: "Jonathan Mort
saying "be afraid, be very afraid" are saying a lot of nonsense
> about Meltdown and Sceptre. It seems to be an echo chamber effect - the
> papers were released yesterday afternoon, but in a rush to get "quoted",
> all the wannabe-quoted people are saying things th
10:54, Dave Taht wrote:
> On Thu, Jan 4, 2018 at 1:52 PM, Joel Wirāmu Pauling
> wrote:
> > Well as I've argued before Lede ideally should be using to Kernel
> Namespaces
> > (poor mans containers) for at a minimum the firewall and per-interface
> > routing instances
e platform. Also I don't want
to be admining centos/rhel servers at home.
On 5 January 2018 at 10:47, Dave Taht wrote:
> On Thu, Jan 4, 2018 at 1:44 PM, Joel Wirāmu Pauling
> wrote:
> >
> >
> > On 5 January 2018 at 01:09, Jonathan Morton
> wrote:
> >>
>
On 5 January 2018 at 01:09, Jonathan Morton wrote:
>
>
> I don't think we need to worry about it too much in a router context.
> Virtual server folks, OTOH...
>
> - Jonathan Morton
>
> Disagree - The Router is pretty much synonymous with NFV
; I run my lede instances at home on hypervisors -
Here in New Zealand ; any provider operating a 'Network of National
Significance' must hand over any encryption keys on demand to what is
our local equivalent of the NSA. This makes a very high disincentive
for anyone to provide end to end encryption by design in their access
and transport networks
Radio/AP co-ordination is something I sorely wish LEDE/Openwrt had.
Run a VM on your server as controller and setup everything else as
Drone AP's.
On 5 December 2017 at 12:40, Matt Taggart wrote:
> On 12/03/2017 09:44 AM, Dave Taht wrote:
> [snip]
>>
>> Another goal was a largely fruitless quest
On 5 December 2017 at 06:00, Dave Taht wrote:
>>> The route table lookup also really expensive on the main cpu.
>
> To clarify the context here, I was asking specifically if the X5 mellonox card
> did routing table offlload or only switching.
>
To clarify what I know the X5 using it's smart offlo
set vendors, I think we
> will see CPE with 10G SFP+ and 802.11ax Q3/Q4 this year.
> Price is of course a bit steeper than the 15USD USB DSL modem :P, but
> probably fits nicely for the SMB segment.
>
> Pedro
>
> On Mon, Dec 4, 2017 at 11:47 AM, Joel Wirāmu Pauling
> wrote:
>
at 1G does not
> necessarily need to be generated by only one host on the LAN.
>
>
>
> Pedro
>
> On Mon, Dec 4, 2017 at 11:27 AM, Joel Wirāmu Pauling
> wrote:
>>
>> How to deliver a switch, when the wiring and port standard isn't
>> actually workable?
&
C network standard quite badly - which looked so
promising, so for the moment Tb3 it is for me at least.
On 4 December 2017 at 23:18, Mikael Abrahamsson wrote:
> On Mon, 4 Dec 2017, Joel Wirāmu Pauling wrote:
>
>> I'm not going to pretend that 1Gig isn't enough for most
I'm not going to pretend that 1Gig isn't enough for most people. But I
refuse to believe it's the networks equivalent of a 10A power (20A
depending on where you live in the world) AC residential phase
distribution circuit.
This isn't a question about what people need, it's more about what the
mark
On 4 December 2017 at 13:11, Michael Richardson wrote:
>
> Dave Taht wrote:
.
>
> > On the really high end the 48 core arm boxes from cavium look
> interesting.
>
> I'm told that there is some special sauce to get them to go at the speeds the
> specs say. Basically you run Linux on one core
Just from a Telco/Industry perspective slant.
Everything in DC has moved to SFP28 interfaces at 25Gbit as the server
port of interconnect. Everything TOR wise is now QSFP28 - 100Gbit.
Mellanox X5 cards are the current hotness, and their offload
enhancements (ASAP2 - which is sorta like DPDK on ste
I quite liked the rk3399 board ; it has enough lanes for good duplex
ethernet support, but I haven't seen anyone bundling with multiple
Gigabit ports on a board so you are stuck with miniPciE add-in or USB3
dongles for more than one port.
You can get hold of them relatively easily via aliexpress a
I've installed several 3800's into Fibre Install van's attached to
Inverters and LTE Dongles to provide in-van wifi. One has been going for
the last 5 years without issue so spikes seem to be well tolerated.
On 6 June 2017 at 09:51, Michael Richardson wrote:
> Such an interesting thread.
> But,
Joel
On 10 March 2017 at 10:34, Dave Taht wrote:
> On Thu, Mar 9, 2017 at 1:21 PM, Joel Wirāmu Pauling
> wrote:
> > 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.
>
> I have been using
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
My biggest bug bear is that reliance on netperf/netserver with -DEMO mode
compilation time flag breaks compilation on recent RHEL and Fedora boxes
due to recent GCC incompatibilities.
Also the range of tests requiring fairly complex setup of external tools
which rely on java (DTG-ITG) etc also cau
I spend a lot of Time on site at big carriers. If it was tasteful - vendor
free and passable in a business casual sense ; it would be something I wear
to promote visibility.
Hoodies - of decent quality are IME a good way to go.
-Joel
On 26 September 2016 at 20:16, Dave Taht wrote:
> Usually in
I am using a TP-Link ac1750 off a Shaw CMT 150/10mbit without issues using
fq_codel and minimal tweaking. I had a wndr3800 before that and concur it's
too underpowered.
I drive AC wifi to the internet at the wan rate so can't complain.
I know tp-link have recently come under fire, but if you can
I have half a dozen c7 v2 bought over the last 6 months - they are still
atheros. I have not touched the cheaper c5's
Good to know. Although frustrating however.
On 3 February 2016 at 18:09, Dave Taht wrote:
> So I've been getting setup again for a run against make-wifi-fast.
> Today I went to
On Jun 25, 2015 10:35 AM, "Dave Taht" wrote:
>
> I have been abusing it on a picostation and nanostation now for 48
> hours. The archer c7v2 (as a source specific gateway) for a week. A
> couple wndr3800s. No crashes. Can still trigger the dreaded wifi TX
> DMA bug, but it seems harder now. DID re
This is great news. Currently all the OVS sdn nodes I have seen, seem to
be using mq scheduler on nics and hoping for the best with kernel defaults.
On Mar 10, 2015 2:18 PM, "Dave Taht" wrote:
> Looks like openvswitch is about to gain sfq, codel, and fq_codel. See:
>
> https://www.mail-archive.c
Do those patches include Broadcom wireless AC support for mainline or
wireless-next or is it still a Wireless free 'open' wireless router?
On Jan 7, 2015 8:13 PM, "Aaron Wood" wrote:
> So it appears that Marvell pushed a bunch of patches to OpenWRT on
> Christmas, and as a result, trunk OpenWRT (
IIRC only some of the ath9k radio's are able to do 5mhz channels -
namely the enterprise chip variants.
-Joel
On 9 October 2014 11:14, Dave Taht wrote:
> https://www.usenix.org/system/files/conference/nsdi12/nsdi12-final142.pdf
>
> I've had 5mhz channels working in the ath9k at various points in
Most of the time, they just want to access a remote VPS/Torrent
Seedbox or $service from their local network.
On 3 October 2014 15:38, Joel Wirāmu Pauling wrote:
> Yup - Routing is going to be better on performance. But is definately
> the more advanced/less common use case from my expe
; On Thu, Oct 2, 2014 at 7:24 PM, Joel Wirāmu Pauling wrote:
>> I.e Your topology looks like this :
>>
>> [(Remote LAN) - VPN Client]---[INTERNET]---(Local LAN)[WAN][LAN][REMOTE-LAN])
>>
>> Your Local LAN knows nothing about Remote LAN and Vice versa. There is
>>
Somewhat related question. Is anyone successfully using VxLANs in
Toronto release?
On 3 October 2014 15:24, Joel Wirāmu Pauling wrote:
> I.e Your topology looks like this :
>
> [(Remote LAN) - VPN Client]---[INTERNET]---(Local LAN)[WAN][LAN][REMOTE-LAN])
>
> Your Local LAN know
I.e Your topology looks like this :
[(Remote LAN) - VPN Client]---[INTERNET]---(Local LAN)[WAN][LAN][REMOTE-LAN])
Your Local LAN knows nothing about Remote LAN and Vice versa. There is
just a single Inteface/Client member that is a member of REMOTE-LAN.
So to get traffic from Local LAN to Remote
In Cerowrt the various net devices have been relabeled; as per here :
http://www.bufferbloat.net/projects/cerowrt/wiki/Device_naming_scheme
I usually add a new device via Luci (call it somethingvpn) and select
custom device (tap0 or tun0).
Than add a new Firewall zone (VPN)
I tend to edit the
Put as many pairs as you can fit into the conduit to leave quite a lot of
slack (2-5metres)
We bury our splitters with ofdm break outs in waterproof boxes every 500m -
2km or so for the GPON roll out and blow the Fibre to the premise from the
split. Burying splinter boxes prevents vandalism/floodi
I've found that OpenVPN on the ar71xx boards with tls-client security
and UDP based tunnel encap max hit a cpu bound upper transfer limit of
about 10mbit.
Just FYI.
-Joel
On 22 September 2014 17:21, Dave Taht wrote:
> Eric:
>
> Most of the cerowrt folk are on cerowrt-devel.
>
> http://wiki.open
:
>
> On 12 Sep, 2014, at 4:49 am, Joel Wirāmu Pauling wrote:
>
>> So if ookla implemented a udp based test, changed it's statical weighting
>> and data mining methods overnight. At least in NZ that might help.
>
> Isn
Changing to better test suites isn't the answer, I know I've tried a lot to
educate and push for bespoke test sites for our customers .
This is all about happy illusions on the user side.
This is a branding and markets problem. In NZ, 'the' site for reference
that users go to is speedtest. Becaus
I have been heavily involved with the UFB (Ultrafast Broadband) PON
deployment here in New Zealand.
I am not sure how the regulated environment is playing out in Canada
(I am moving there in a month so I guess I will find out). But here
the GPON architecture is METH based and Layer2 only. Provider
runs is a lot cheaper than Cat6a/7 runs needed for 10gbit copper.
On 9 September 2014 16:03, Joel Wirāmu Pauling wrote:
> Just a head's up I have had issues with the X5xx intel SFP+ optics
> interoperating with other vendors. This may not be an issue for your
> deployment.
>
&
Just a head's up I have had issues with the X5xx intel SFP+ optics
interoperating with other vendors. This may not be an issue for your
deployment.
If you want good interop then Broadcom based optics seem to be the
best bet at the moment.
On 9 September 2014 12:09, Wes Felter wrote:
> The Intel
https://wikidevi.com/wiki/Securifi_Almond%2B
On 2 September 2014 22:05, Joel Wirāmu Pauling wrote:
> On a somewhat related note - I've just received my NZ/AU Region
> Almond+ which is an arm9 Dual core router based on the Cortina CSC SoC
> :
>
> https://www.cortina-systems
On a somewhat related note - I've just received my NZ/AU Region
Almond+ which is an arm9 Dual core router based on the Cortina CSC SoC
:
https://www.cortina-systems.com/product/digital-home-processors/16-products/996-cs7542-cs7522
More details :
On 2 September 2014 21:27, Jonathan Morton wrote:
Hi all,
I just spent an hour scratching my head.
I have a fairly easy setup, a vpn a 6in4 tun devices and seperated
networks for 2.4 5 and wired networks.
I just flashed the toronto released. And everything seemed to be
working, but I kept getting connection refused for non http(s) and DNS
.
Af
67 matches
Mail list logo