CakeQOS CakeQOS-Merlin

[Treadler]

As long as you know what needs to be run to "start cake" again, you can manually add lines similar to these as a test:

#!/bin/sh

if [ "$1" = "restart" ] && [ "$2" = "wrs" ]; then { sh /jffs/addons/cake-qos/cake-qos.sh checkrun & } ; fi
if [ "$1" = "start" ] && [ "$2" = "sig_check" ]; then { sh /jffs/addons/cake-qos/cake-qos.sh checkrun & } ; fi

Test at your own risk since I can't test this myself. But this is the same checkrun command that runs at the top of every hour in the cake-qos script.

I also wouldn't encourage everyone to do this yet, because if it ever gets added to the main script, it will be a pain to deal with these manual entries and any entries the script may want to add in the future, and avoid duplication.

Many thanks!

Did a manual Trend Micro check, Cake died, at top of the hour your solution restarted Cake.

Worst case scenario, Cake will be missing in action for <1 hour.

 

[dave14305]

Many thanks!

Did a manual Trend Micro check, Cake died, at top of the hour your solution restarted Cake.

Worst case scenario, Cake will be missing in action for <1 hour.

The top of the hour check is already built into the script. My snippet should have been immediately run, but maybe there's a timing issue. May as well uninstall Cake and install FlexQoS (trojan horse warning!)

 

[Treadler]

The top of the hour check is already built into the script. My snippet should have been immediately run, but maybe there's a timing issue. May as well uninstall Cake and install FlexQoS (trojan horse warning!)

Chuckle!

Already included you say?! Maybe I should have tested that.

Back to the drawing board.......

 

[dave14305]

Chuckle!

Already included you say?! Maybe I should have tested that.

Back to the drawing board.......

Make sure you put that code in /jffs/scripts/service-event-end and the script is executable. Check syslog for the service-event-end entries.

 

[Odkrys]

Thanks for the files. On my AX58U running:

4.1.52 #1 SMP PREEMPT Fri Aug 14 15:32:27 EDT 2020 armv7l ASUSWRT-Merlin

When I try to insmod the sch_cake.ko file I get "invalid module format".

Sorry, I uploaded file again.

 

[Treadler]

Make sure you put that code in /jffs/scripts/service-event-end and the script is executable. Check syslog for the service-event-end entries.

Cake does indeed fix itself at the top of the hour.
Who knew?!

 

[sfx2000]

VPN is for security and will interfere with QOS. Chose one or the other

Further explanation here - Cake SQM traffic shaping is connection based - and OpenVPN being application layer, all traffic goes thru that one connection flow.

 

[sfx2000]

download: unlimited
upload: unlimited

For Cake SQM - you have to define the limits - no limits means it's disabled

 

[heysoundude]

Isn't the larger problem the fact that this still only works with two of the Merlin-supported router models?

Also, back on my favourite topic of DSCP tagging. Has anyone had a look at seeing if tagging can be added to this script such as is done in OpenWRT? https://forum.openwrt.org/t/using-dscp-for-qos/14252/9
My own exploration only served to confirm how much I have yet to learn .

Yes, that is a problem...but when older models cease to be supported (and I suspect a rather significant cull in a short bit), it will cease to be an issue. sorry (not sorry) to say it, but that's the facts of the matter.

Having no idea what you meant by DSCP tagging, I took to wiki: https://en.wikipedia.org/wiki/Differentiated_services
Yikes...those IETF people have a lot on their plates!! Nice to see my chosen QoS method is ahead of their curve.

 

[Morris]

For Cake SQM - you have to define the limits - no limits means it's disabled

using zero as the limit enables "Automatic" and in my testing it's not as good as setting the limits.

 

[Morris]

Yes, that is a problem...but when older models cease to be supported (and I suspect a rather significant cull in a short bit), it will cease to be an issue. sorry (not sorry) to say it, but that's the facts of the matter.

Having no idea what you meant by DSCP tagging, I took to wiki: https://en.wikipedia.org/wiki/Differentiated_services
Yikes...those IETF people have a lot on their plates!! Nice to see my chosen QoS method is ahead of their curve.

The classification world is not a fun place to live. There is always changes and that mess up the classification and new apps you need to address. If you want to live in that world, use Adaptive QOS and add Flex-QOS.

 

[crkpot]

Do these files provided by @Odkrys mean ax56u and ax58u are going to be supported by cake-merlin soon??

 

[Wade Coxon]

Yes, that is a problem...but when older models cease to be supported (and I suspect a rather significant cull in a short bit), it will cease to be an issue. sorry (not sorry) to say it, but that's the facts of the matter.

It's not my policy Just that amtm will not adopt scripts that do not work on the majority of Merlin-supported models.
I'd personally be happy to see it included, and let any future Merlin supported models gradually be added to the cake list .

Having no idea what you meant by DSCP tagging, I took to wiki: https://en.wikipedia.org/wiki/Differentiated_services
Yikes...those IETF people have a lot on their plates!! Nice to see my chosen QoS method is ahead of their curve.

Your chosen QoS method can make use of DSCP to better manage the traffic flows. That's what the Diffserv4, 8 etc models mean. Traffic gets classified, and then binned, and then the bins can be treated preferentially on top of the normal Cake queue management system. It's a Cake enhancement.
As Cake it right now, Diffserv modes are practically useless unless you are somehow receiving pre-classified traffic on your Asus router. That's why Besteffort is the current best choice.

 

[sfx2000]

The classification world is not a fun place to live.

Keep it simple, and just works...

Spoiler: low end stuff with cakeSQM

@thumper:~# tc -d qdisc
qdisc noqueue 0: dev lo root refcnt 2
qdisc fq_codel 0: dev eth0 root refcnt 2 limit 10240p flows 1024 quantum 1514 target 5ms interval 100ms memory_limit 4Mb ecn drop_batch 64
qdisc cake 8009: dev eth1 root refcnt 2 bandwidth 90Mbit besteffort triple-isolate nonat nowash no-ack-filter split-gso rtt 100ms noatm overhead 44
qdisc ingress ffff: dev eth1 parent ffff:fff1 ----------------
qdisc noqueue 0: dev br-lan root refcnt 2
qdisc noqueue 0: dev wg0 root refcnt 2
qdisc noqueue 0: dev wlan0 root refcnt 2
qdisc cake 800a: dev ifb4eth1 root refcnt 2 bandwidth 90Mbit besteffort triple-isolate nonat wash no-ack-filter split-gso rtt 100ms noatm overhead 44

@thumper:~# tc -s qdisc
qdisc noqueue 0: dev lo root refcnt 2
Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
qdisc fq_codel 0: dev eth0 root refcnt 2 limit 10240p flows 1024 quantum 1514 target 5ms interval 100ms memory_limit 4Mb ecn drop_batch 64
Sent 227107635 bytes 185514 pkt (dropped 0, overlimits 0 requeues 2)
backlog 0b 0p requeues 2
maxpacket 1514 drop_overlimit 0 new_flow_count 1 ecn_mark 0
new_flows_len 0 old_flows_len 0
qdisc cake 8009: dev eth1 root refcnt 2 bandwidth 90Mbit besteffort triple-isolate nonat nowash no-ack-filter split-gso rtt 100ms noatm overhead 44
Sent 360807553 bytes 482490 pkt (dropped 17, overlimits 206314 requeues 46)
backlog 0b 0p requeues 46
memory used: 444928b of 4500000b
capacity estimate: 90Mbit
min/max network layer size: 28 / 1500
min/max overhead-adjusted size: 72 / 1544
average network hdr offset: 14

Tin 0
thresh 90Mbit
target 5ms
interval 100ms
pk_delay 53us
av_delay 32us
sp_delay 19us
backlog 0b
pkts 482507
bytes 360826123
way_inds 5797
way_miss 2120
way_cols 0
drops 17
marks 0
ack_drop 0
sp_flows 1
bk_flows 1
un_flows 0
max_len 60560
quantum 1514

qdisc ingress ffff: dev eth1 parent ffff:fff1 ----------------
Sent 957709433 bytes 811305 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
qdisc noqueue 0: dev br-lan root refcnt 2
Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
qdisc noqueue 0: dev wg0 root refcnt 2
Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
qdisc noqueue 0: dev wlan0 root refcnt 2
Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
qdisc cake 800a: dev ifb4eth1 root refcnt 2 bandwidth 90Mbit besteffort triple-isolate nonat wash no-ack-filter split-gso rtt 100ms noatm overhead 44
Sent 968068503 bytes 810645 pkt (dropped 660, overlimits 732094 requeues 0)
backlog 0b 0p requeues 0
memory used: 434Kb of 4500000b
capacity estimate: 90Mbit
min/max network layer size: 46 / 1500
min/max overhead-adjusted size: 90 / 1544
average network hdr offset: 14

Tin 0
thresh 90Mbit
target 5ms
interval 100ms
pk_delay 134us
av_delay 31us
sp_delay 4us
backlog 0b
pkts 811305
bytes 969067703
way_inds 2858
way_miss 2235
way_cols 0
drops 660
marks 0
ack_drop 0
sp_flows 1
bk_flows 1
un_flows 0
max_len 1514
quantum 1514

 

[Treadler]

Thanks for v 1.0.3 @Adamm

 

[XIII]

Thanks for v 1.0.3 @Adamm

How to update?

Install over current? Or uninstall first?

 

[Treadler]

How to update?

Install over current? Or uninstall first?

Cake menu option # 5.
Updates the current version.

 

[XIII]

Cake menu option # 5.
Updates the current version.

Ah, yes. This is the one time I should not have been reading the documentation first it seems...

 

[heysoundude]

Thanks for v 1.0.3 @Adamm

changelog? or just a quick blast of what got tweaked and why?

 

[XIII]

changelog? or just a quick blast of what got tweaked and why?

Updated Cake binaries:

v1.0.3 · ttgapers/cakeqos-merlin@a142d16

Update cake binaries

github.com