search

lemoer / freifunk-ideas

0 stars 0 forks source link

Gluon: MCS-Statistics in Respondd #4

Open lemoer opened 3 years ago

lemoer commented 3 years ago

ath9k

(Maybe this applies to other chipsets using minstrel in kernel as well).

RX Statistics:

# cat /sys/kernel/debug/ieee80211/phy0/netdev:client0/stations/4c\:dd\:31\:74\:08\:04/node_recv
                   HT20      HT40       SGI       LGI
     MCS  0 :          1         0         0         1
     MCS  1 :          0         0         0         0
     MCS  2 :         18        11        10        19
     MCS  3 :         39        10        16        33
     MCS  4 :        202        39       100       141
     MCS  5 :          0         0         0         0
     MCS  6 :          0         0         0         0
     MCS  7 :       1622        88      1066       644
     MCS  8 :          0         0         0         0
     MCS  9 :          0         0         0         0
     MCS 10 :          0         0         0         0
     MCS 11 :          0         0         0         0
     MCS 12 :          0         0         0         0
     MCS 13 :          0         0         0         0
     MCS 14 :          0         0         0         0
     MCS 15 :          0         0         0         0
     MCS 16 :          0         0         0         0
     MCS 17 :          0         0         0         0
     MCS 18 :          0         0         0         0
     MCS 19 :          0         0         0         0
     MCS 20 :          0         0         0         0
     MCS 21 :          0         0         0         0
     MCS 22 :          0         0         0         0
     MCS 23 :          0         0         0         0

  CCK-1M/LP :         99
  CCK-2M/LP :          0
CCK-5.5M/LP :          0
 CCK-11M/LP :          0
  CCK-2M/SP :          0
CCK-5.5M/SP :          0
 CCK-11M/SP :          0
    OFDM-6M :        121
    OFDM-9M :          0
   OFDM-12M :          0
   OFDM-18M :          0
   OFDM-24M :       1740
   OFDM-36M :          0
   OFDM-48M :          0
   OFDM-54M :          0

I checked, that the sum of the "MCS" values correlates to the rx packets on client0.

TX Statistics:

# cat /sys/kernel/debug/ieee80211/phy0/netdev:client0/stations/4c\:dd\:31\:74\:08\:04/rc_stats
              best   ____________rate__________    ________statistics________    _____last____    ______sum-of________
mode guard #  rate  [name   idx airtime  max_tp]  [avg(tp) avg(prob) sd(prob)]  [retry|suc|att]  [#success | #attempts]
HT20  LGI  1         MCS0     0    1477     4.8       2.4      52.4      0.0       3     0 0             2   4        
HT20  LGI  1         MCS1     1     738     9.7       7.3      70.3      0.0       4     0 0            23   35       
HT20  LGI  1         MCS2     2     492    14.6       2.4      27.5      0.0       5     0 0            40   62       
HT20  LGI  1    C    MCS3     3     369    17.0       9.7      47.7      0.0       4     0 0            85   310      
HT20  LGI  1         MCS4     4     246    24.4       7.3      27.8      0.0       5     0 0           220   349      
HT20  LGI  1         MCS5     5     185    29.2       9.7      28.9      0.0       5     0 0           167   581      
HT20  LGI  1         MCS6     6     164    31.7       4.8      19.6      0.0       5     0 0           440   980      
HT20  LGI  1  A      MCS7     7     148    34.1      14.6      37.9      0.0       5     0 0           201   552      
HT20  SGI  1         MCS0    30    1329     4.8       4.8     100.0      0.0       0     0 0             1   1        
HT20  SGI  1         MCS1    31     665     9.7       4.8      49.3      0.0       2     0 0            11   23       
HT20  SGI  1   B  P  MCS2    32     443    14.6       9.7      59.7      0.0       5     1 5            83   136      
HT20  SGI  1         MCS3    33     332    19.5       7.3      42.7      0.0       5     0 0           189   438      
HT20  SGI  1     D   MCS4    34     222    26.8       9.7      33.1      0.0       4     0 0           293   612      
HT20  SGI  1         MCS5    35     166    31.7       7.3      22.2      0.0       5     0 0           546   1281     
HT20  SGI  1         MCS6    36     148    34.1       9.7      26.5      0.0       5     0 0           218   500      
HT20  SGI  1         MCS7    37     133    36.6       7.3      22.9      0.0       6     0 0           193   451

I checked, that the sum of the "sum-of #success" values correlates to the tx packets on client0. The "sum-of #attempts" column has even higher values, which might be interesting as well. They show, how many packets were sent, but failed.

The rc_stats can also be obtained as csv:

# cat /sys/kernel/debug/ieee80211/phy0/netdev:client0/stations/4c\:dd\:31\:74\:08\:04/rc_stats_csv
HT20,LGI,1,,MCS0 ,0,1477,4.8,2.4,52.4,0.0,3,0,0,2,4,2615,117,1.2
HT20,LGI,1,,MCS1 ,1,738,9.7,7.3,70.3,0.0,4,0,0,23,35,2615,117,1.2
HT20,LGI,1,,MCS2 ,2,492,14.6,2.4,27.5,0.0,5,0,0,40,62,2615,117,1.2
HT20,LGI,1,D,MCS3 ,3,369,17.0,9.7,47.7,0.0,4,0,0,85,310,2615,117,1.2
HT20,LGI,1,,MCS4 ,4,246,24.4,7.3,27.8,0.0,5,0,0,220,349,2615,117,1.2
HT20,LGI,1,,MCS5 ,5,185,29.2,9.7,28.9,0.0,5,0,0,167,581,2615,117,1.2
HT20,LGI,1,,MCS6 ,6,164,31.7,4.8,19.6,0.0,5,0,0,440,980,2615,117,1.2
HT20,LGI,1,AP,MCS7 ,7,148,34.1,24.4,66.4,0.0,5,1,1,218,581,2615,117,1.2
HT20,SGI,1,,MCS0 ,30,1329,4.8,4.8,100.0,0.0,0,0,0,1,1,2615,117,1.2
HT20,SGI,1,,MCS1 ,31,665,9.7,4.8,49.3,0.0,2,0,0,11,23,2615,117,1.2
HT20,SGI,1,C,MCS2 ,32,443,14.6,9.7,58.0,0.0,5,0,0,84,137,2615,117,1.2
HT20,SGI,1,B,MCS3 ,33,332,19.5,12.2,63.7,0.0,5,0,0,190,439,2615,117,1.2
HT20,SGI,1,,MCS4 ,34,222,26.8,9.7,33.1,0.0,4,0,0,293,612,2615,117,1.2
HT20,SGI,1,,MCS5 ,35,166,31.7,7.3,22.2,0.0,5,0,0,546,1281,2615,117,1.2
HT20,SGI,1,,MCS6 ,36,148,34.1,9.7,26.5,0.0,5,0,0,218,500,2615,117,1.2
HT20,SGI,1,,MCS7 ,37,133,36.6,7.3,22.9,0.0,6,0,0,193,451,2615,117,1.2

ath10k

The above mentioned debugfs files do not work in ath10k. Ath10k does not use the mac80211 minstrel algorithm in the kernel. The wifi chips offload the MCS selection. Therefore ath10k has other mechanisms.

Theoretically there is a debugfs file called tx_stats:

lemoer commented 3 years ago

CC: @CodeFetch, @s-2

lemoer commented 3 years ago

The next step would be to think about aggregating the statistics in a meaningful way.

s-2 commented 3 years ago

Awesome!

Since these are just counters (reset upon (re-)association I would guess?) it seems we have to poll and diff these periodically to build a histogram etc. for the current time interval...

So it's probably not quite real-time (however it would also be cool if the user could see at least their current connection quality from the router stats (nextnode) page)...

But for a beginning, we might just accumulate the node_recv stuff (looks simpler at least ;) ) for fixed intervals, e.g. 15 minutes, for all clients within that interval, so we could further accumulate these in larger steps (30 min / 1h / 2h / ... / 1d / ...) when needed.

I must admit I'm absolutely not familiar with plotting stuff in Grafana, but this Heatmap Graph looks quite cool :) https://grafana.com/docs/grafana/latest/getting-started/intro-histograms/#heatmaps

The colouring might be confusing though, when the user would expect red to mean "weak link quality" rather than "most packets had this link quality" :thinking:

lemoer commented 3 years ago

Both TX and RX stats are both also available for mesh partners.

Here is an example for rc_stats:

root@NDS-Alte-Kaserne:~# cat /sys/kernel/debug/ieee80211/phy0/netdev\:mesh0/stations/7a\:89\:39\:44\:16\:b1/rc_stats

              best   ____________rate__________    ________statistics________    _____last____    ______sum-of________
mode guard #  rate  [name   idx airtime  max_tp]  [avg(tp) avg(prob) sd(prob)]  [retry|suc|att]  [#success | #attempts]
CCK    LP  1          1.0M  120   10548     0.0       0.0      79.2      0.0       0     0 0          4488   5353     
CCK    LP  1          2.0M  121    5476     0.0       0.0       0.0      0.0       0     0 0             0   0        
CCK    LP  1          5.5M  122    2411     2.4       0.0       0.0      0.0       0     0 0             0   0        
CCK    LP  1         11.0M  123    1535     4.8       0.0       0.0      0.0       0     0 0             0   0        
CCK    SP  1          2.0M  125    5380     0.0       0.0       0.0      0.0       0     0 0             0   0        
CCK    SP  1          5.5M  126    2315     2.4       0.0       0.0      0.0       0     0 0             0   0        
CCK    SP  1         11.0M  127    1439     4.8       0.0       0.0      0.0       0     0 0             0   0        
HT20  LGI  1         MCS0     0    1477     4.8       4.8     100.0      0.0       3     0 0             2   2        
HT20  LGI  1         MCS1     1     738     9.7       9.7     100.0      0.0       0     0 0             1   1        
HT20  LGI  1         MCS2     2     492    14.6      14.6     100.0      0.0       5     0 0             5   5        
HT20  LGI  1         MCS3     3     369    17.0      17.0      97.6      0.0       5     0 0            25   37       
HT20  LGI  1         MCS4     4     246    24.4      24.4      97.9      0.0       5     0 0            82   120      
HT20  LGI  1         MCS5     5     185    29.2      29.2     100.0      0.0       5     0 0           377   533      
HT20  LGI  1         MCS6     6     164    31.7      31.7      99.7      0.0       5     0 0           353   496      
HT20  LGI  1         MCS7     7     148    34.1      34.1      98.9      0.0       5     0 0          1104   1503     
HT20  LGI  2         MCS8    10     738     9.7       9.7     100.0      0.0       0     0 0             1   1        
HT20  LGI  2         MCS9    11     369    17.0      17.0      98.5      0.0       5     0 0            19   26       
HT20  LGI  2         MCS10   12     246    24.4      24.4      95.7      0.0       5     0 0           164   225      
HT20  LGI  2         MCS11   13     185    29.2      29.2     100.0      0.0       5     0 0           577   764      
HT20  LGI  2         MCS12   14     123    36.6      36.6      85.4      0.0       6     0 0          4323   5308     
HT20  LGI  2         MCS13   15      92    43.9      31.7      65.9      0.0       6     0 0         10665   13179    
HT20  LGI  2         MCS14   16      82    46.3      36.6      70.6      0.0       6     0 0         14147   17882    
HT20  LGI  2         MCS15   17      74    48.8      36.6      68.3      0.0       6     0 0          9942   13298    
HT20  LGI  3         MCS16   20     492    14.6      14.6     100.0      0.0       0     0 0             1   1        
HT20  LGI  3         MCS17   21     246    24.4      24.4      99.1      0.0       5     0 0           148   205      
HT20  LGI  3         MCS18   22     164    31.7      31.7      99.8      0.0       5     0 0           687   883      
HT20  LGI  3         MCS19   23     123    36.6      21.9      56.0      0.0       6     0 0          4329   5329     
HT20  LGI  3  A      MCS20   24      82    46.3      46.3      96.8      0.0       6     1 1         23384   28284    
HT20  LGI  3     D   MCS21   25      62    51.2      39.0      67.2      0.0       6     0 0         18645   24098    
HT20  LGI  3         MCS22   26      55    53.7      24.4      43.3      0.0       6     0 0         14976   20647    
HT20  LGI  3         MCS23   27      49    56.1      24.4      41.7      0.0       6     0 0          9304   14782    
HT20  SGI  1         MCS0    30    1329     4.8       0.0       0.0      0.0       0     0 0             0   0        
HT20  SGI  1         MCS1    31     665     9.7       9.7     100.0      0.0       0     0 0             1   1        
HT20  SGI  1         MCS2    32     443    14.6      14.6     100.0      0.0       5     0 0             1   1        
HT20  SGI  1         MCS3    33     332    19.5      19.5     100.0      0.0       5     0 0            41   56       
HT20  SGI  1         MCS4    34     222    26.8      26.8     100.0      0.0       5     0 0           223   287      
HT20  SGI  1         MCS5    35     166    31.7      31.7     100.0      0.0       5     0 0          1313   1767     
HT20  SGI  1         MCS6    36     148    34.1      34.1      95.8      0.0       5     0 0          1569   2025     
HT20  SGI  1         MCS7    37     133    36.6      36.6      88.4      0.0       6     0 0          2457   3105     
HT20  SGI  2         MCS8    40     665     9.7       9.7      96.4      0.0       0     0 0             3   4        
HT20  SGI  2         MCS9    41     332    19.5      19.5     100.0      0.0       5     0 0            97   120      
HT20  SGI  2         MCS10   42     222    26.8      26.8      97.2      0.0       5     0 0           223   301      
HT20  SGI  2         MCS11   43     166    31.7      31.7      97.9      0.0       5     0 0          1142   1482     
HT20  SGI  2         MCS12   44     111    39.0      36.6      84.6      0.0       6     0 0          6835   8471     
HT20  SGI  2         MCS13   45      83    46.3      34.1      68.3      0.0       6     0 0         14976   18451    
HT20  SGI  2   B  P  MCS14   46      74    48.8      41.5      76.5      0.0       6     0 0         16095   20303    
HT20  SGI  2         MCS15   47      67    51.2      29.2      52.9      0.0       6     0 0         13128   17761    
HT20  SGI  3         MCS16   50     443    14.6      14.6     100.0      0.0       5     0 0             1   1        
HT20  SGI  3         MCS17   51     222    26.8      26.8      99.2      0.0       5     0 0           267   374      
HT20  SGI  3         MCS18   52     148    34.1      34.1      96.9      0.0       5     0 0          2680   3398     
HT20  SGI  3         MCS19   53     111    39.0      34.1      78.5      0.0       6     0 0          7863   9733     
HT20  SGI  3    C    MCS20   54      74    48.8      39.0      71.7      0.0       6     0 0         26683   32582    
HT20  SGI  3         MCS21   55      56    53.7      34.1      59.8      0.0       6     0 0         23674   30577    
HT20  SGI  3         MCS22   56      49    56.1      36.6      59.2      0.0       6     0 0         16415   22942    
HT20  SGI  3         MCS23   57      44    58.5      24.4      39.8      0.0       6     0 0          7557   12566

(here we can see, that the higher MCS values have multiple streams).

And here for node_recv:

root@NDS-Alte-Kaserne:~# cat /sys/kernel/debug/ieee80211/phy0/netdev\:mesh0/stations/7a\:89\:39\:44\:16\:b1/node_recv 
                    HT20      HT40       SGI       LGI
     MCS  0 :          9         0         3         6
     MCS  1 :        391        35       175       251
     MCS  2 :       2537       377      1585      1329
     MCS  3 :       4623       945      3087      2481
     MCS  4 :      38304      5467     24625     19146
     MCS  5 :     116512     13150     73946     55716
     MCS  6 :     158094     18126     97982     78238
     MCS  7 :     178237     20951    108592     90596
     MCS  8 :        257        31       153       135
     MCS  9 :       5159      1102      3697      2564
     MCS 10 :      37635      5046     25581     17100
     MCS 11 :     144683     14679     89303     70059
     MCS 12 :     447138     38304    266353    219089
     MCS 13 :     391359     48389    223706    216042
     MCS 14 :     399590     47698    225058    222230
     MCS 15 :     289290     35597    161316    163571
     MCS 16 :       1471       182       836       817
     MCS 17 :      36854      4124     24215     16763
     MCS 18 :     224978     20133    133976    111135
     MCS 19 :     545994     44823    320632    270185
     MCS 20 :     815074     78400    465582    427892
     MCS 21 :     307777     41356    161931    187202
     MCS 22 :     213725     28706    112729    129702
     MCS 23 :     108628     14542     53693     69477

  CCK-1M/LP :          0
  CCK-2M/LP :          0
CCK-5.5M/LP :          0
 CCK-11M/LP :          1
  CCK-2M/SP :          0
CCK-5.5M/SP :          0
 CCK-11M/SP :          0
    OFDM-6M :          0
    OFDM-9M :          0
   OFDM-12M :   45392736
   OFDM-18M :          0
   OFDM-24M :          0
   OFDM-36M :          0
   OFDM-48M :          0
   OFDM-54M :          0
lemoer commented 3 years ago

The counters actually reset, if a client is reassociated. Therefore if we want to aggregate all clients into one counter, we need to keep counters for "gone" clients, such that the overall counter keeps increasing monotonically even if clients are disassociated.

CodeFetch commented 3 years ago

Is it needed to have monotonically increasing counters?

What I would be interested in were floating average, best client, worst client, median, upper and lower quartile. From this one can already get an idea about the distribution. Of course a histogram would be nice, but do we really want to send so much data? I think it is hard for the user to understand. Still from another perspective I would say in the future we might want to train an analytics model which automatically detects outliers. So I don't know...