cake-autorate: adding support for Asus Merlin

[Lynx]

I recommend placing opkg install bash iputils-ping fping as a prerequisite step after ensuring Entware is installed.

Thanks - done:

Improve Asus Merlin installation instructions · lynxthecat/cake-autorate@91cefa5

See #299

github.com

I've also changed the post-install guidance to:

Alter Asus Merlin post-install guidance · lynxthecat/cake-autorate@4119801

See #299

github.com

As we know that this works and there is an unknown issue with the conventional route - see:

Add remaining Asus Merlin fixes to installer by lynxthecat · Pull Request #299 · lynxthecat/cake-autorate

@LiveWire1968 any chance you could test the latest code on this asus-merlin branch and we'll add any necessary remaining fixes? I think(?) the installation instructions should be something like...

github.com

 

[underdose]

Excellent news! What does total CPU usage look like for cake-autorate on your device? You can check with ‘atop 2’, then tap ‘p’ to collate the processes, and observe the total CPU percentage for 'bash'.

Checked it with htop and it's 5.6-5.7%.

 

[JGrana]

Question. I also have Starlink. Using spdMerlin and looking at the last 7 days, here are some up/dl number:

Max UL - 39Mbps
Avg UL - 20 Mbps
Min UL - 1.8 Mbps

Max DL - 442 Mbps
Avg DL - 232 Mbps
Min DL - 19 Mbps

Avg Jitter - 4.5 msec
Max Jitter - 45 msec (almost always in the evening during streaming…)

Pretty typical for Starlink!

In my house, it’s just my wife and I now and our usage is - 30% both on the web - usually not streaming
20% Streaming in the evening (usually when Starlink is getting a bit congested)
40% Network fairly quiet, some uploads, sw updates in the background etc.
10% late at night running remote backups

Given the speeds above and usage model - is it worth installing cake/cake-autorate?

 

[Lynx]

Question. I also have Starlink. Using spdMerlin and looking at the last 7 days, here are some up/dl number:

Max UL - 39Mbps
Avg UL - 20 Mbps
Min UL - 1.8 Mbps

Max DL - 442 Mbps
Avg DL - 232 Mbps
Min DL - 19 Mbps

Avg Jitter - 4.5 msec
Max Jitter - 45 msec (almost always in the evening during streaming…)

Pretty typical for Starlink!

In my house, it’s just my wife and I now and our usage is - 30% both on the web - usually not streaming
20% Streaming in the evening (usually when Starlink is getting a bit congested)
40% Network fairly quiet, some uploads, sw updates in the background etc.
10% late at night running remote backups

Given the speeds above and usage model - is it worth installing cake/cake-autorate?

Actually Starlink has come on leaps and bounds such that I’m not even sure bufferbloat is an issue for Starlink connections these days. Try saturating the connection and seeing how much latency increases under a ping test. Probably by only a small amount now. If that’s indeed the case, the only benefit of cake with cake-autorate is ensuring fair usage of the connection and potentially smoothing over or highlighting any irregularities in the connection. From the recent Starlink data I’ve seen, I’d consider foregoing QOS altogether.

For 4G/5G connections, cake with cake-autorate is still essential I think.

 

[JGrana]

Actually Starlink has come on leaps and bounds such that I’m not even sure bufferbloat is an issue for Starlink connections these days. Try saturating the connection and seeing how much latency increases under a ping test. Probably by only a small amount now. If that’s indeed the case, the only benefit of cake with cake-autorate is ensuring fair usage of the connection and potentially smoothing over or highlighting any irregularities in the connection. From the recent Starlink data I’ve seen, I’d consider foregoing QOS altogether.

For 4G/5G connections, cake with cake-autorate is still essential I think.

Thanks. I agree that over the past 6 months it’s been steadily getting better.
I will take your advice and leave QOS out of the equation.
If my kids were still living at home and streaming away, I would probably give it a shot.

 

[Lynx]

I’d encourage you to verify with simple testing. Even just running from a Windows command prompt ‘ping -n 10000 1.1.1.1’ and saturating the connection with a download and looking at the latency increase. Or you can try the Waveform bufferbloat test.

 

[rlj2]

Actually Starlink has come on leaps and bounds such that I’m not even sure bufferbloat is an issue for Starlink connections these days. Try saturating the connection and seeing how much latency increases under a ping test. Probably by only a small amount now. If that’s indeed the case, the only benefit of cake with cake-autorate is ensuring fair usage of the connection and potentially smoothing over or highlighting any irregularities in the connection. From the recent Starlink data I’ve seen, I’d consider foregoing QOS altogether.

For 4G/5G connections, cake with cake-autorate is still essential I think.

I dont really run it now with my starlink, other then to help for the install/port process.

 

[SomeWhereOverTheRainBow]

I’d encourage you to verify with simple testing. Even just running from a Windows command prompt ‘ping -n 10000 1.1.1.1’ and saturating the connection with a download and looking at the latency increase. Or you can try the Waveform bufferbloat test.

I am doing it captain!!! I don't think she can take anymore! Shes about to blow!