[Bloat] Re: LibreQoS CLI - in OpenWrt?

[Bloat] Re: LibreQoS CLI - in OpenWrt?

[Rich Brown]

> On Jul 27, 2026, at 02:03, bloat-request@lists.bufferbloat.net wrote:
> 
> It was asked for by QoO guys for their IETF 126 Vienna Hackathon, but it
> will be useful for everyone, so it was added to the speedtest website now:
> 
> https://test.libreqos.com/cli/

This is cool, indeed. I wonder if it would be a suitable replacement for my `betterspeedtest.sh` script in OpenWrtScript (https://github.com/richb-hanover/OpenWrtScripts#betterspeedtestsh).

Has anyone tested this on OpenWrt? Or considered adding this as a package? (This would be a fun project, but I’m up to my elbows in stuff right now...)

Rich

[Bloat] Re: LibreQoS CLI - in OpenWrt?

[jf]

Agree, it is very cool; it’s on my to-do list to replace the speedtestcpp package with this. I run a test every two hours on my x86 router. That test hits the Ookla infrastructure, but the target server selection seems simplistic, and results are quite variable (or my ISP has +/- 500Mbps variances across a week).

One question I have is: Does this test account for *all* traffic flowing through the router during the tests, or solely on its own payloads? 

Ideally, as a CLI running on the router itself, it should capture 100% of the bidirectional traffic so its reports are accurate. If one or more nodes in the network are streaming video, that could be a significant delta.

The IQtests we ran on the IQrouter took all traffic into consideration, and it made a difference.

Cheers,

Jonathan

> On Jul 29, 2026, at 12:30 PM, Rich Brown <richb.hanover@gmail.com> wrote:
> 
> 
>> On Jul 27, 2026, at 02:03, bloat-request@lists.bufferbloat.net wrote:
>> 
>> It was asked for by QoO guys for their IETF 126 Vienna Hackathon, but it
>> will be useful for everyone, so it was added to the speedtest website now:
>> 
>> https://test.libreqos.com/cli/
> 
> This is cool, indeed. I wonder if it would be a suitable replacement for my `betterspeedtest.sh` script in OpenWrtScript (https://github.com/richb-hanover/OpenWrtScripts#betterspeedtestsh).
> 
> Has anyone tested this on OpenWrt? Or considered adding this as a package? (This would be a fun project, but I’m up to my elbows in stuff right now...)
> 
> Rich
> _______________________________________________
> Bloat mailing list -- bloat@lists.bufferbloat.net
> To unsubscribe send an email to bloat-leave@lists.bufferbloat.net

[Bloat] Re: LibreQoS CLI - in OpenWrt?

[Matt Taggart]

On 7/29/26 9:30 AM, Rich Brown wrote:
> 
>> On Jul 27, 2026, at 02:03, bloat-request@lists.bufferbloat.net wrote:
>>
>> It was asked for by QoO guys for their IETF 126 Vienna Hackathon, but it
>> will be useful for everyone, so it was added to the speedtest website now:
>>
>> https://test.libreqos.com/cli/

The direct URL the installer uses for linux/x86_64 is

https://download.libreqos.com/cli/latest/libreqos-test-linux-x86_64

it's a dynamically linked executable that only uses base system 
libraries, no additional dependencies.
How is it licensed and is the source somewhere?

I did find this

https://github.com/LibreQoE/LibreQoS

It's not clear if the cli stuff is in there yet.
I look forward to this being packages in distros so I can install it 
everywhere!

> This is cool, indeed. I wonder if it would be a suitable replacement for my `betterspeedtest.sh` script in OpenWrtScript (https://github.com/richb-hanover/OpenWrtScripts#betterspeedtestsh).

I love betterspeedtest.sh and appreciate it existing and will continue 
to use it even if there are alternatives.

But I know a big part of making it available is actually making the 
backend servers it uses for testing available. I guess this new libreqos 
cli tool uses the same infra as their web based tester, is that 
documented somewhere?

Matt

[Bloat] Re: LibreQoS CLI - in OpenWrt?

[Sebastian Moeller]

Hi there...

On July 29, 2026 6:46:19 PM GMT+02:00, jf--- via Bloat <bloat@lists.bufferbloat.net> wrote:
>Agree, it is very cool; it’s on my to-do list to replace the speedtestcpp package with this. I run a test every two hours on my x86 router. That test hits the Ookla infrastructure, but the target server selection seems simplistic, and results are quite variable (or my ISP has +/- 500Mbps variances across a week).

Just to keep in mind (I am certain you are aware, but others on the list might not) someone typically has to pay for such servers and their capacity/traffic, so I would ask the LibreQoS guys first whether they consider automated repeated tests fair use (even thougb every 12 hours is rather tame).


>
>One question I have is: Does this test account for *all* traffic flowing through the router during the tests, or solely on its own payloads? 


This test can only account for its own traffic... but you can combine it it monitoring your wan interfaces traffic counters, no?

>
>Ideally, as a CLI running on the router itself, it should capture 100% of the bidirectional traffic so its reports are accurate. If one or more nodes in the network are streaming video, that could be a significant delta.

Sure, but that is a different use case to this bufferbloat test... IMHO the libreqos test is primarily interested in measuring bufferbloat, not capacity, it mostly uses the data transfers to saturate the link to allow meaningfull loaded latency measurements.

>
>The IQtests we ran on the IQrouter took all traffic into consideration, and it made a difference.

Sure, different usecase, I guess.

>
>Cheers,
>
>Jonathan
>
>> On Jul 29, 2026, at 12:30 PM, Rich Brown <richb.hanover@gmail.com> wrote:
>> 
>> 
>>> On Jul 27, 2026, at 02:03, bloat-request@lists.bufferbloat.net wrote:
>>> 
>>> It was asked for by QoO guys for their IETF 126 Vienna Hackathon, but it
>>> will be useful for everyone, so it was added to the speedtest website now:
>>> 
>>> https://test.libreqos.com/cli/
>> 
>> This is cool, indeed. I wonder if it would be a suitable replacement for my `betterspeedtest.sh` script in OpenWrtScript (https://github.com/richb-hanover/OpenWrtScripts#betterspeedtestsh).
>> 
>> Has anyone tested this on OpenWrt? Or considered adding this as a package? (This would be a fun project, but I’m up to my elbows in stuff right now...)
>> 
>> Rich
>> _______________________________________________
>> Bloat mailing list -- bloat@lists.bufferbloat.net
>> To unsubscribe send an email to bloat-leave@lists.bufferbloat.net
>
>_______________________________________________
>Bloat mailing list -- bloat@lists.bufferbloat.net
>To unsubscribe send an email to bloat-leave@lists.bufferbloat.net

-- 
Sent from my Android device with K-9 Mail. Please excuse my brevity.