CakeQOS CakeQoS-Merlin v2.0.0 alpha

[rlj2]

I've seen alot of CAKE talk and I wanted to know if I install the latest 386.2 beta 2, is the built in CAKE the same as CAKE 2.0.0 alpha or they're separately? I haven't kept up with all the info going around. I apologize if this has been asked already

Cake 2.0 is basically just a script/gui to control Beta 2s built in cake . Right now you wont see much difference with cake2.0 over just hitting the cake button in merlin. From what it sounds theres alot planned for cakegui in the future.

 

[Kingp1n]

Cake 2.0 is basically just a script/gui to control Beta 2s built in cake

Thanks for that, for those trying out the CAKE 2.0 alpha and getting great results, have you also tried the default settings from built in CAKE and compared those results by any chance?

Also, will the goal be to get CAKE 2.0 into the stable 386.2 firmware (built-in CAKE) ?

 

[rlj2]

Thanks for that, for those trying out the CAKE 2.0 alpha and getting great results, have you also tried the default settings from built in CAKE and compared those results by any chance?

Edited my comment above. you can tweak settings with the gui, but you can also do the same from cli with built-in. Again, they default pretty much identical

 

[Kingp1n]

I know each environment will be different but I wonder how much of a gain/benefit you can get from the additional options in the CAKE alpha 2.0 vs the default settings within the built in CAKE settings.

It will take alot of trial & error and some experimenting to get the right setup.

I'm happy we'll be getting the additional options to play with but I hope it won't become a headache (at least for me) when I'm constantly changing my settings daily haha

I have to remember to keep it simple with CAKE-QoS. Thanks for this to all who made CAKE happen in our Asus routers.

@dave14305 the QoS wizard!!! And @RMerlin for allowing CAKE into his firmware!

 

[RMerlin]

Also, will the goal be to get CAKE 2.0 into the stable 386.2 firmware (built-in CAKE) ?

No. My Cake implementation will remain as basic as possible so it works for the majority of users with no further configuration beyond choosing an appropriate overhead preset. This script is what people who wants an easy way to completely customize their setup will use.

 

[dave14305]

that's cool - what changes for IPv6?

Goodput = Bandwidth * ( ( MTU - 40 - 20 ) / ( MTU + Overhead ) )

An extra 20 bytes lost.

 

[dave14305]

here you go. FYI I have updated to beta 2 prior to getting three logs.

Mar 21 15:56:41 kernel: Init chrdev /dev/idp with major 190
Mar 21 15:56:41 kernel: tdts: tcp_conn_max = 8000
Mar 21 15:56:41 kernel: tdts: tcp_conn_timeout = 300 sec
Mar 21 15:56:45 kernel: SHN Release Version: 2.0.1 c03f6c5
Mar 21 15:56:45 kernel: UDB Core Version: 0.2.20
Mar 21 15:56:45 kernel: Init chrdev /dev/idpfw with major 191
Mar 21 15:56:45 kernel: IDPfw: flush fc
Mar 21 15:56:45 kernel: IDPfw: IDPfw is ready
Mar 21 15:56:45 kernel: sizeof forward pkt param = 192
Mar 21 15:56:45 BWDPI: fun bitmap = 3
Mar 21 15:56:45 rc_service: httpd 1253:notify_rc restart_qos;restart_firewall
Mar 21 15:56:45 custom_script: Running /jffs/scripts/service-event (args: restart qos)
Mar 21 15:56:45 CakeQOS-Merlin: Configuring Cake options...
Mar 21 15:56:46 BWDPI: force to flush flowcache entries
Mar 21 15:56:46 kernel: IDPfw: Exit IDPfw
Mar 21 15:56:46 kernel: mod epilog takes 0 jiffies
Mar 21 15:56:46 kernel: IDPfw: Exit IDPfw
Mar 21 15:56:46 kernel: Exit chrdev /dev/idpfw with major 191
Mar 21 15:56:46 kernel: Exit chrdev /dev/idp with major 190
Mar 21 15:56:46 kernel: Archer TCP Pure ACK Enabled
Mar 21 15:56:46 kernel: ^[[0;33;41m[ERROR archer] sysport_tm_command,1224: Feature Unavailable^[[0m
Mar 21 15:56:46 custom_config: Appending content of /jffs/configs/cake-qos.conf.add.
Mar 21 15:56:46 custom_script: Running /jffs/scripts/service-event-end (args: restart qos)
Mar 21 15:56:46 custom_script: Running /jffs/scripts/service-event (args: restart firewall)
Mar 21 15:56:46 custom_script: Running /jffs/scripts/nat-start
Mar 21 15:56:47 custom_script: Running /jffs/scripts/service-event-end (args: restart firewall)
Mar 21 15:56:48 kernel: Init chrdev /dev/idp with major 190
Mar 21 15:56:48 kernel: tdts: tcp_conn_max = 8000
Mar 21 15:56:48 kernel: tdts: tcp_conn_timeout = 300 sec
Mar 21 15:56:52 kernel: SHN Release Version: 2.0.1 c03f6c5
Mar 21 15:56:52 kernel: UDB Core Version: 0.2.20
Mar 21 15:56:52 kernel: Init chrdev /dev/idpfw with major 191
Mar 21 15:56:52 kernel: IDPfw: flush fc
Mar 21 15:56:52 kernel: IDPfw: IDPfw is ready
Mar 21 15:56:52 kernel: sizeof forward pkt param = 192
Mar 21 15:56:52 BWDPI: fun bitmap = 3
Mar 21 15:56:58 BWDPI: force to flush flowcache entries
Mar 21 15:56:58 kernel: IDPfw: Exit IDPfw
Mar 21 15:56:58 kernel: mod epilog takes 0 jiffies
Mar 21 15:56:58 kernel: IDPfw: Exit IDPfw
Mar 21 15:56:58 kernel: Exit chrdev /dev/idpfw with major 191
Mar 21 15:56:58 kernel: Exit chrdev /dev/idp with major 190
Mar 21 15:56:58 custom_config: Appending content of /jffs/configs/cake-qos.conf.add.
Mar 21 15:56:59 custom_script: Running /jffs/scripts/nat-start

It looks like Adaptive QoS is still active or trying to be active, maybe if you're opening the Bandwidth Monitor page.

 

[heysoundude]

Goodput = Bandwidth * ( ( MTU - 40 - 20 ) / ( MTU + Overhead ) )

An extra 20 bytes lost.

insignificant if you're able to jumbo-frame, as v6 I believe was intended to...

 

[RMerlin]

Thanks for the great work on this script @dave14305 and @ttgapers. I've done a bit of testing and all working fine.

still getting the archer error when applying any changes but it does not seem to cause any issues...

Mar 21 11:26:25 CakeQOS-Merlin: Configuring Cake options...
Mar 21 11:26:26 kernel: Archer TCP Pure ACK Enabled
Mar 21 11:26:26 kernel: ^[[0;33;41m[ERROR archer] sysport_tm_command,1224: Feature Unavailable^[[0m

Ignore that. sysport_tm is only supported by the RT-AX56U, however the firmware still runs the command to disable it regardless of the exact model.

 

[jata]

Ignore that. sysport_tm is only supported by the RT-AX56U, however the firmware still runs the command to disable it regardless of the exact model.

Thanks for the clarification. Haven't seen the error until recently but I did have logging turned down as 386.1 was so stable.

 

[JGrana]

It's been 24 hours on 386.2_beta2 on an AX88U. Enabled Cake with pretty much defaults (I filter duplicate Acks on upload). My ISP (Spectrum) has me at 200mb down and 10mb up.

I have been getting A/A+ since then. I had run a few tests yesterday and now this morning.

Nice job guys!

 

[Wade Coxon]

Another minor update to the webui to show a tooltip with what kind of throughput (goodput) you might expect based on your bandwidth, MTU and overhead. I've seen this calculation used by Sebastian Moeller on the Cake mail list and OpenWRT forums many times, and I've added it when you hover over the upload/download fields.

View attachment 32242

Goodput = Bandwidth * ( ( MTU - 20 - 20 ) / ( MTU + Overhead ) )

If your test results are reasonably close to these expectations, then you are doing well. This is for IPv4 Ethernet.

I just applied a hotfix update, and this tooltip appears to have stopped working.

 

[Mutzli]

It's working beautifully:

 

[rlj2]

I just applied a hotfix update, and this tooltip appears to have stopped working.

It was reverted for now.

 

[dave14305]

I just applied a hotfix update, and this tooltip appears to have stopped working.

It didn’t work for people with pppoe connections, so I have to come up with a more involved method for retrieving the right MTU value from the router.

 

[skeal]

What's up with the low latency gaming, I've never heard of the expression?

Bufferbloat and Internet Speed Test - Waveform

View the full results, and test your own bufferbloat

www.waveform.com

 

[DanNuke86]

What's up with the low latency gaming, I've never heard of the expression?

Bufferbloat and Internet Speed Test - Waveform

View the full results, and test your own bufferbloat

www.waveform.com

Gaming doesn't use a lot of bandwidth but latency is crucial for online multiplayer games, particular FPS where every ms counts. The lower and more stable the latency, the better the gaming experience! I'd be interested to know exactly how the test conducts and scores on this count though..

 

[skeal]

Gaming doesn't use a lot of bandwidth but latency is crucial for online multiplayer games, particular FPS where every ms counts. The lower and more stable the latency, the better the gaming experience! I'd be interested to know exactly how the test conducts and scores on this count though..

Thanks. I play Counter-Strike and have experienced no change in my game flow.

 

[iTyPsIDg]

Gaming doesn't use a lot of bandwidth but latency is crucial for online multiplayer games, particular FPS where every ms counts. The lower and more stable the latency, the better the gaming experience! I'd be interested to know exactly how the test conducts and scores on this count though..

They explain at the bottom of the page.

Spoiler: How do you determine which services will work well on my connection?

 

[rlj2]

It didn’t work for people with pppoe connections, so I have to come up with a more involved method for retrieving the right MTU value from the router.

Dave, just out of curiosity, in my case. The asus router is connected to a ethernet port on the dsl/pppoe modem. Does the asus even know its pppoe or care?