CakeQOS CakeQOS-Merlin

[ugandy]

oh. i wonder if it could be because HW acceleration is turned off for cake?

that affects LAN<->WAN but likely LAN<->LAN too, no?

if this is the case, it is a big limitation, that the wifi to wifi transfer speeds and available wifi to wifi bandwidth are capped when using cake.

i don't have a lot of client-to-client transfers, but if someone does (NAS, etc) this is a concern.
should be clarified

found this: https://www.snbforums.com/threads/broadcoms-hardware-acceleration.18144/page-2#post-599962
from this answer, HW accel is only for WAN<->LAN, so cake should not affect LAN<->LAN speed

 

[ttgapers]

• Without cake-qos I score bufferbloat B; with cake-qos (up/down either 90% or 95%) I still score B...
• With cake-qos enabled I loose about 16% up/down Mbps

Does this mean cake-qos has only disadvantages in my situation and I should not be using it?

Or do I misunderstand why so many people are enthusiastic about it?

These are my current settings for my 250/25 connection:

[1]  --> Download Speed             | [225 Mbit]
[2]  --> Upload Speed               | [22 Mbit]
[3]  --> Queue Priority             | [besteffort]
[4]  --> Extra Download Options     | [docsis ack-filter dual-dsthost]
[5]  --> Extra Upload Options       | [docsis ack-filter dual-srchost]

Note that it's the Ookla speedtest CLI on the router (AC86U) itself that reports the 250/25 that are in my ISP's advertising; DSLReports gives lower values with/without cake-qos...

What are the results if you selectively remove dual-dsthost and dual-srchost?

 

[rgnldo]

Cake-QOS requires technical knowledge from the user. Hence the need for presets. But it only works with the unlimited profile and DL / UP rates with “0”.

 

[Toink]

Yeah been in the industry toooooo long to keep redundant data all over the place. The link for Advanced Users (Tweaking) is handily posted on Post #1 for all....so feel free to yell RTFM

Just kidding, loving the community support on these forums from since 2014 as my handle says. Happy to be able to semi-contribute back!

Ouch! Having been with the community longer than you, that really stings! But as I've posted in my 3 posts prior to @ugandy's, I read the entire topic from page 1 and that I missed that part in the link coz I skipped to the lower part "Making cake sing and dance..." Most of the time, making a post more simple and concise make's it more understandable for the common good.

I'm glad, @ugandy pointed that out and made it worthwhile for me. Indeed that's one of the good things in this community.

Thank you @ttgapers for this new tool!

Going back to the topic, I settled for 0 down/ 0 up. Ethernet for both extra up/down options.

 

[ttgapers]

Ouch! Having been with the community longer than you, that really stings! But as I've posted in my 3 posts prior to @ugandy's, I read the entire topic from page 1 and that I missed that part in the link coz I skipped to the lower part "Making cake sing and dance..." Most of the time, making a post more simple and concise make's it more understandable for the common good.

I'm glad, @ugandy pointed that out and made it worthwhile for me. Indeed that's one of the good things in this community.

Thank you @ttgapers for this new tool!

Going back to the topic, I settled for 0 down/ 0 up. Ethernet for both extra up/down options.

View attachment 24517

No worries @Toink - was more in good jest that meant to sting

I am happy you are enjoying Cake, and love all the feedback!

Cheers!

 

[XIII]

i would install pingplotter and observe the latency during the spdMerlin test, and compare with/without cake

I’m not familiar with that tool.

Do you mean this?

https://www.pingplotter.com/

 

[ugandy]

I’m not familiar with that tool.

Do you mean this?

https://www.pingplotter.com/

yes. has trial period but can be used for free after.
i ended up buying it because i used it often (thanks comcast )

you can pick a couple of targets (8.8.8.8 , 1.1.1.1, 192.168.1.1, etc) and monitor the latency.
check it while testing with speedtest.net, fast.com, dslreports.com, etc
or during the day if your ISP has issues.

it's like connmon on steroids for your clients.

or, you can just run "ping -t 8.8.8.8" on your windows client, for the same results in a text format (no graphs)

 

[heysoundude]

Good old NBN.

I was reading up on the infrastructure for another purpose and the Government really messed that up. Instead of spending that 51 billion on running Fiber to every residence, they gave you a mix of HFC,FTTP,FTTN,FTTC,FTTB.

With FTTP = FTTH the rest are DSL connections and one Coaxial one.

https://www.nbnco.com.au/learn/network-technology/fibre-to-the-curb-explained-fttc

Yup, some earlier roll out areas (Including me - woo hoo!) were fortunate enough to get the originally intended fibre.
Once the politicians got involved it became an ever increasing mismatch of obsolete tech & cobbled together compromises.
Very much a case of “we didn’t need the Internet in 1950, why do we need it now?”

So if the CRTC or Industry Canada mentions Australian consultants/experts in conjunction with a similar initiative here, I should scream bloody murder?

 

[heysoundude]

You are being watched. Private and state-sponsored organizations are monitoring and recording your online activities. https://www.privacytools.io/

Open source software
CakeQOS+Suricata

You forgot unbound

 

[rgnldo]

You forgot unbound

As we are in a AiProtection-QOS post.
Also, FW Merlin includes other open source software.

 

[Makaveli]

So if the CRTC or Industry Canada mentions Australian consultants/experts in conjunction with a similar initiative here, I should scream bloody murder?

lol yes buy a loud horn and scream as loud as you can if it ain't FTTH we don't want it.

 

[XIII]

Could betterspeedtest.sh be a good test to verify my buffer bloat?

I'm in doubt because of the 0 MB download/upload in the values below...

Without CakeQOS:

2020-07-04 21:57:54 Testing against netperf.bufferbloat.net (ipv4) with 5 simultaneous sessions while pinging gstatic.com (60 seconds in each direction)
...............
 Download: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 9.299
    10pct: 9.299
   Median: 10.918
      Avg: 11.365
    90pct: 12.722
      Max: 12.795
...............
   Upload: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 8.783
    10pct: 8.783
   Median: 10.862
      Avg: 10.800
    90pct: 11.265
      Max: 12.122

With CakeQOS (225 down, 25 up, best effort, docsis ack-filter, docsis ack-filter):

      2020-07-04 21:58:46 Testing against netperf.bufferbloat.net (ipv4) with 5 simultaneous sessions while pinging gstatic.com (60 seconds in each direction)
...............
 Download: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 10.603
    10pct: 10.603
   Median: 12.580
      Avg: 12.271
    90pct: 13.470
      Max: 14.832
...............
   Upload: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 10.640
    10pct: 10.640
   Median: 10.847
      Avg: 11.448
    90pct: 12.827
      Max: 13.482

 

[ugandy]

Could betterspeedtest.sh be a good test to verify my buffer bloat?

I'm in doubt because of the 0 MB download/upload in the values below...

Without CakeQOS:

2020-07-04 21:57:54 Testing against netperf.bufferbloat.net (ipv4) with 5 simultaneous sessions while pinging gstatic.com (60 seconds in each direction)
...............
 Download: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 9.299
    10pct: 9.299
   Median: 10.918
      Avg: 11.365
    90pct: 12.722
      Max: 12.795
...............
   Upload: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 8.783
    10pct: 8.783
   Median: 10.862
      Avg: 10.800
    90pct: 11.265
      Max: 12.122

With CakeQOS (225 down, 25 up, best effort, docsis ack-filter, docsis ack-filter):

      2020-07-04 21:58:46 Testing against netperf.bufferbloat.net (ipv4) with 5 simultaneous sessions while pinging gstatic.com (60 seconds in each direction)
...............
 Download: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 10.603
    10pct: 10.603
   Median: 12.580
      Avg: 12.271
    90pct: 13.470
      Max: 14.832
...............
   Upload: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 10.640
    10pct: 10.640
   Median: 10.847
      Avg: 11.448
    90pct: 12.827
      Max: 13.482

interesting script!
i downloaded it but when i triy to run it on my ax88, i get an error saying my router is missing "mktemp"
how did you install this?

edit: ah! i had to do "opkg install coreutils-mktemp", plus a couple other utilities.

nice script! it's like spdMerlin and connmon bundled in one!

my results with cake on:

cromo@RT-AX88U-8158:/tmp/home/root# /jffs/addons/util/betterspeedtest.sh -t 60 -H netperf-west.bufferbloat.net
2020-07-04 14:22:48 Testing against netperf-west.bufferbloat.net (ipv4) with 5 simultaneous sessions while pinging gstatic.com (60 seconds in each direction)
............................................................
 Download: 258.99 Mbps
  Latency: (in msec, 60 pings, 0.00% packet loss)
      Min: 5.033
    10pct: 7.650
   Median: 10.857
      Avg: 11.084
    90pct: 14.314
      Max: 18.417
.............................................................
   Upload: 17.69 Mbps
  Latency: (in msec, 61 pings, 0.00% packet loss)
      Min: 8.744
    10pct: 10.360
   Median: 13.540
      Avg: 13.618
    90pct: 15.778
      Max: 23.522

very nice!

@ttgapers why not offer option to bundle this script with cake-qos install?
then cake-qos could have option to do quick test and check speeds/latency results on the spot

how I installed it:

26 mkdir -p /jffs/addons/util
27 /usr/sbin/curl -s "https://raw.githubusercontent.com/richb-hanover/OpenWrtScripts/master/betterspeedtest.sh" -o "/jffs/addons/util/betterspeedtest.sh"
29 chmod +x /jffs/addons/util/betterspeedtest.sh
31 opkg install coreutils-mktemp
33 opkg install coreutils-seq
36 opkg install netperf
42 opkg install procps-ng-pgrep
45 /jffs/addons/util/betterspeedtest.sh -t 60 -H netperf-west.bufferbloat.net

 

[rgnldo]

Could betterspeedtest.sh be a good test to verify my buffer bloat?

I'm in doubt because of the 0 MB download/upload in the values below...

Without CakeQOS:

2020-07-04 21:57:54 Testing against netperf.bufferbloat.net (ipv4) with 5 simultaneous sessions while pinging gstatic.com (60 seconds in each direction)
...............
 Download: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 9.299
    10pct: 9.299
   Median: 10.918
      Avg: 11.365
    90pct: 12.722
      Max: 12.795
...............
   Upload: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 8.783
    10pct: 8.783
   Median: 10.862
      Avg: 10.800
    90pct: 11.265
      Max: 12.122

With CakeQOS (225 down, 25 up, best effort, docsis ack-filter, docsis ack-filter):

      2020-07-04 21:58:46 Testing against netperf.bufferbloat.net (ipv4) with 5 simultaneous sessions while pinging gstatic.com (60 seconds in each direction)
...............
 Download: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 10.603
    10pct: 10.603
   Median: 12.580
      Avg: 12.271
    90pct: 13.470
      Max: 14.832
...............
   Upload: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 10.640
    10pct: 10.640
   Median: 10.847
      Avg: 11.448
    90pct: 12.827
      Max: 13.482

Nice!

 

[ugandy]

      2020-07-04 21:58:46 Testing against netperf.bufferbloat.net (ipv4) with 5 simultaneous sessions while pinging gstatic.com (60 seconds in each direction)
...............
 Download: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 10.603
    10pct: 10.603
   Median: 12.580
      Avg: 12.271
    90pct: 13.470
      Max: 14.832
...............
   Upload: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 10.640
    10pct: 10.640
   Median: 10.847
      Avg: 11.448
    90pct: 12.827
      Max: 13.482

@XIII note that if the script says 0.00Mbps, it wasn't actually transferring data, it was just doing "ping".

i had to run it with: "betterspeedtest.sh -t 60 -H netperf-west.bufferbloat.net"
to actually get transfers going...

 

[ttgapers]

Could betterspeedtest.sh be a good test to verify my buffer bloat?

I'm in doubt because of the 0 MB download/upload in the values below...

Without CakeQOS:

2020-07-04 21:57:54 Testing against netperf.bufferbloat.net (ipv4) with 5 simultaneous sessions while pinging gstatic.com (60 seconds in each direction)
...............
 Download: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 9.299
    10pct: 9.299
   Median: 10.918
      Avg: 11.365
    90pct: 12.722
      Max: 12.795
...............
   Upload: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 8.783
    10pct: 8.783
   Median: 10.862
      Avg: 10.800
    90pct: 11.265
      Max: 12.122

With CakeQOS (225 down, 25 up, best effort, docsis ack-filter, docsis ack-filter):

      2020-07-04 21:58:46 Testing against netperf.bufferbloat.net (ipv4) with 5 simultaneous sessions while pinging gstatic.com (60 seconds in each direction)
...............
 Download: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 10.603
    10pct: 10.603
   Median: 12.580
      Avg: 12.271
    90pct: 13.470
      Max: 14.832
...............
   Upload: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 10.640
    10pct: 10.640
   Median: 10.847
      Avg: 11.448
    90pct: 12.827
      Max: 13.482

I am bit opposed to including the tests into the script for a couple reasons:

1. The Cake support team (native), doesn't bundle it, and includes as steps for users to test (and has simple, medium and best tests).
2. Those tests are QoS script agnostic, therefore better to keep them unbundled so to speak as it's helpful for all QoS implementations here.
3. Personally I run those scripts on another ubuntu box I have internal and runs well when required. I prefer not running the scripts on the router itself, and having to include other dependencies as noted.
4. Additional complexity and support required, I don't think makes it worth it at this point.

Feel free to add as an issue/feature request, but do be weary that it may possibly be quashed/delayed/postponed due to the above.

Good questions!

 

[Pak Kriss]

Many thanks to everyone who made this possible!

 

[XIII]

@XIII note that if the script says 0.00Mbps, it wasn't actually transferring data, it was just doing "ping".

i had to run it with: "betterspeedtest.sh -t 60 -H netperf-west.bufferbloat.net"
to actually get transfers going...

I must have been low on coffee yesterday, because I was trying netperf-eu instead of the FQDN netperf-eu.bufferbloat.net...

I do get non-zero downloads/uploads now

Time for some more tests later today! (If anyone can also tell me which stupid mistake I’m making in the Suricata topic that would be great)

PS: I’m also not in favor of including the test scripts in CakeQOS (it’s not that hard to install them manually)

 

[maxbraketorque]

Could betterspeedtest.sh be a good test to verify my buffer bloat?

I'm in doubt because of the 0 MB download/upload in the values below...

Without CakeQOS:

2020-07-04 21:57:54 Testing against netperf.bufferbloat.net (ipv4) with 5 simultaneous sessions while pinging gstatic.com (60 seconds in each direction)
...............
 Download: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 9.299
    10pct: 9.299
   Median: 10.918
      Avg: 11.365
    90pct: 12.722
      Max: 12.795
...............
   Upload: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 8.783
    10pct: 8.783
   Median: 10.862
      Avg: 10.800
    90pct: 11.265
      Max: 12.122

With CakeQOS (225 down, 25 up, best effort, docsis ack-filter, docsis ack-filter):

      2020-07-04 21:58:46 Testing against netperf.bufferbloat.net (ipv4) with 5 simultaneous sessions while pinging gstatic.com (60 seconds in each direction)
...............
 Download: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 10.603
    10pct: 10.603
   Median: 12.580
      Avg: 12.271
    90pct: 13.470
      Max: 14.832
...............
   Upload: 0.00 Mbps
  Latency: (in msec, 15 pings, 0.00% packet loss)
      Min: 10.640
    10pct: 10.640
   Median: 10.847
      Avg: 11.448
    90pct: 12.827
      Max: 13.482

You're ping times are lower without Cake.

 

[ugandy]

You're ping times are lower without Cake.

those numbers were bogus. there was no concurrent upload/download (see 0.00 Mbps).
ping difference was within noise range