any statistics available

[dtaht]

I was pleased to learn of this project today. Do your users have any visibility into any of the statistics cake keeps? I am primarily interested in backlog, drops, ack drops, and marks, but there are a variety of others.

By which I meant graphs, over time.

[underd0se]

@dtaht I'm a user of this project and yes, we have access to stats in the router GUI as well as to configuration and status information.

Edit: We also have access above from the command line. Below's the output of "cake-qos status download"

#########################################################
##               _                                     ##
##              | |                                    ##
##    ___  __ _ | | __ ___          __ _   ___   ___   ##
##   / __|/ _` || |/ // _ \ ______ / _` | / _ \ / __|  ##
##  | (__ |(_| ||   <|  __/|______| (_| || (_) |\__ \  ##
##   \___|\__,_||_|\_\\___|        \__, | \___/ |___/  ##
##                                    | |              ##
##                                    |_|              ##
##                                                     ##
##      https://github.com/ttgapers/cakeqos-merlin     ##
##                        v2.0.0                       ##
##                                                     ##
#########################################################

#########################################################

qdisc cake 8002: root refcnt 2 bandwidth 97280Kbit diffserv4 dual-dsthost nat wash ingress no-ack-filter split-gso rtt 100ms noatm overhead 18 mpu 64
 Sent 96197833801 bytes 109545275 pkt (dropped 202859, overlimits 51820240 requeues 0)
 backlog 82308b 55p requeues 0
 memory used: 944704b of 4750Kb
 capacity estimate: 97280Kbit
 min/max network layer size:           46 /    1500
 min/max overhead-adjusted size:       64 /    1518
 average network hdr offset:           14

                   Bulk  Best Effort        Video        Voice
  thresh       6080Kbit    97280Kbit    48640Kbit    24320Kbit
  target            5ms          5ms          5ms          5ms
  interval        100ms        100ms        100ms        100ms
  pk_delay         19us        7.1ms          0us          9us
  av_delay          1us       5.47ms          0us          4us
  sp_delay          1us       3.78ms          0us          3us
  backlog            0b       82308b           0b           0b
  pkts               38    109744394            0         3757
  bytes            4502  96504024291            0       225600
  way_inds            0      6013476            0            0
  way_miss           19     11395065            0            2
  way_cols            0          790            0            0
  drops               0       202859            0            0
  marks               0           53            0            0
  ack_drop            0            0            0            0
  sp_flows            1            4            0            1
  bk_flows            0            1            0            0
  un_flows            0            0            0            0
  max_len           146         1514            0           70
  quantum           300         1514         1484          742

#########################################################`

[ttgapers]

@dtaht thanks for your comments/question. Appreciate it.

On the graphing etc., I believe this is something @dave14305 has been toying with for the future. @dave14305 would be better placed to comment on if/when/how that might be happening.

[dtaht]

My overall campaign is to try and inform more users as to te value of dropping packets intelligently, and to do some a/b testing to pass the good word along.

in your example above you might want to turn on the ack-filter on the upload.

It would be good if the cake community could all agree on the right rules for re-marking videoconferencing traffic.

I am also always curious as to what applications are using ecn, also.

and seeing live stats, with a backlog, and happy users always is nice. thx for doing this port!

[dave14305]

@dtaht thanks for your comments/question. Appreciate it.

On the graphing etc., I believe this is something @dave14305 has been toying with for the future. @dave14305 would be better placed to comment on if/when/how that might be happening.

I’ve been using OpenWrt and digging into the multiple QoS/SQM efforts there (e.g. qosify, adapting QoS rules to nftables, etc.).

The big question was always, “What’s interesting in the CAKE stats output?” We have some feedback now from @dtaht to answer that question.

However I’m not sure if I’ll be switching back to Merlin firmware in the near future to work on this Addon.

[dtaht]

Dream:

packets, ack-drops, drops, marks, on a log scale on one graph backlog in bytes for each tin, linear scale on another graph peak delay for each tin (also one graph), linear

there has been a variety of attempts to get there out there...

[dave14305]

packets, ack-drops, drops, marks, on a log scale on one graph

This is where it could be sloppy if we're dealing with 4 metrics for 1 to 8 tins (hopefully no one really uses diffserv8).