Freenet statistics

Welcome to my Freenet statistics page!

Fetch statistics

Description

These results are based on fetching and inserting 32 single blocks of data. Every day 32 single blocks are inserted into Freenet. At various delays we try to fetch these blocks again. More specifically, each day we try to retrieve the 32 blocks inserted (2delta)-1 days ago for delta = 1,2, … 8. All the fetch results which have been obtained with a specific version of freenet are part of the same sample. From these tests we get (1) results about the retrievability of data, (2) how long it takes for requests to complete and (3) how long inserts take.

The following graphs plot the average fetch performance for a time period. Keep in mind that we cannot clearly distinguish between fetch related improvements and insertion related improvements, i.e. they are co-dependent in Freenet. In addition to these pretty graphs we'd also like to know whether a Freenet version is substantially different from another one. Freenet has two types of versions': regular versions and mandatory versions. Mandatory versions are assumed to substantially alter the network behaviour of Freenet, whereas regular builds do not. The statistics gathered for Freenet are grouped by mandatory version, because we assume that regular versions have no network impact. For each mandatory version we test whether all the results obtained using that version are significantly different from those obtained with another version. This test is performed using a Kruskal-Wallis test.

The following R output may be a bit difficult to grasp, so hang on. Here we run a Kruskal-Wallis test on a bunch of samples. Each sample is based on one of the last 10 mandatory builds. Each build is thus treated as a (ordered) factor. Obviously, we would like to see a build that improves on past performance significantly with a P < 0.05. For each delta (time past since the last attempt to fetch it) the output of the pairwise comparisions of a post hoc test for the Kruskal-Wallis test is performed. A Jonckheere-Terpstra test is also performed to test whether there is a trend over time, but note that some assumptions of this test are violated. More specifically, the sizes of the groups that are compared are not guaranteed to be equal. Note that, this table with post hor tests, is only shown if the Kruskal-Wallis test itself has a p-value < 0.05. This all results in eight plots and possibly tables, one for each delta.

Now, onwards to the actual results!

Fetch graphs

Plot and linear regression model for data inserted 1 days ago and last fetched 1 days ago

plot of chunk fetchplots plot of chunk fetchplots

Multiple comparison test after Kruskal-Wallis 
p.value: 0.05 
Comparisons
          obs.dif critical.dif difference
1352-1354  53.177       544.55      FALSE
1352-1443 127.333       577.02      FALSE
1352-1448  32.417       661.06      FALSE
1352-1449  89.333       632.09      FALSE
1352-1450 113.420       531.30      FALSE
1352-1451 273.760       502.16      FALSE
1352-1464  44.833       999.43      FALSE
1352-1465  73.010       506.02      FALSE
1352-1466 168.040       501.44      FALSE
1354-1443 180.510       360.64      FALSE
1354-1448  85.594       483.84      FALSE
1354-1449  36.156       443.45      FALSE
1354-1450  60.243       281.77      FALSE
1354-1451 220.583       221.97      FALSE
1354-1464   8.344       892.17      FALSE
1354-1465 126.187       230.59      FALSE
1354-1466 221.217       220.33       TRUE
1443-1448  94.917       520.12      FALSE
1443-1449 216.667       482.77      FALSE
1443-1450 240.754       340.31      FALSE
1443-1451 401.093       292.72       TRUE
1443-1464 172.167       912.35      FALSE
1443-1465  54.323       299.31      FALSE
1443-1466  40.707       291.49      FALSE
1448-1449 121.750       580.61      FALSE
1448-1450 145.837       468.89      FALSE
1448-1451 306.176       435.58      FALSE
1448-1464  77.250       967.69      FALSE
1448-1465  40.593       440.04      FALSE
1448-1466 135.623       434.75      FALSE
1449-1450  24.087       427.08      FALSE
1449-1451 184.426       390.23      FALSE
1449-1464  44.500       948.14      FALSE
1449-1465 162.343       395.19      FALSE
1449-1466 257.373       389.30      FALSE
1450-1451 160.340       187.13      FALSE
1450-1464  68.587       884.14      FALSE
1450-1465 186.430       197.28      FALSE
1450-1466 281.460       185.20       TRUE
1451-1464 228.926       866.94      FALSE
1451-1465 346.770        93.79       TRUE
1451-1466 441.800        64.61       TRUE
1464-1465 117.843       869.19      FALSE
1464-1466 212.873       866.52      FALSE
1465-1466  95.030        89.86       TRUE

Is there a positive or negative trend over time according to the Jonckheere-Terpstra test? YES

Plot and linear regression model for data inserted 3 days ago and last fetched 2 days ago

plot of chunk fetchplots plot of chunk fetchplots
Multiple comparison test after Kruskal-Wallis 
p.value: 0.05 
Comparisons
          obs.dif critical.dif difference
1352-1354  90.031       535.67      FALSE
1352-1443  49.071       587.53      FALSE
1352-1448   4.375       650.27      FALSE
1352-1449   9.800       621.78      FALSE
1352-1450  15.609       522.64      FALSE
1352-1451 154.257       493.97      FALSE
1352-1464  57.500       983.12      FALSE
1352-1465 200.823       497.88      FALSE
1352-1466 299.292       493.30      FALSE
1354-1443  40.960       385.83      FALSE
1354-1448  85.656       475.95      FALSE
1354-1449  99.831       436.22      FALSE
1354-1450  74.423       277.17      FALSE
1354-1451 244.289       218.36       TRUE
1354-1464  32.531       877.61      FALSE
1354-1465 110.792       227.06      FALSE
1354-1466 209.261       216.83      FALSE
1443-1448  44.696       533.65      FALSE
1443-1449  58.871       498.53      FALSE
1443-1450  33.463       367.52      FALSE
1443-1451 203.329       325.47      FALSE
1443-1464   8.429       910.19      FALSE
1443-1465 151.752       331.37      FALSE
1443-1466 250.221       324.45      FALSE
1448-1449  14.175       571.14      FALSE
1448-1450  11.234       461.24      FALSE
1448-1451 158.632       428.49      FALSE
1448-1464  53.125       951.90      FALSE
1448-1465 196.448       432.98      FALSE
1448-1466 294.917       427.71      FALSE
1449-1450  25.409       420.11      FALSE
1449-1451 144.457       383.87      FALSE
1449-1464  67.300       932.67      FALSE
1449-1465 210.623       388.88      FALSE
1449-1466 309.092       383.00      FALSE
1450-1451 169.866       184.10      FALSE
1450-1464  41.891       869.72      FALSE
1450-1465 185.215       194.34      FALSE
1450-1466 283.684       182.28       TRUE
1451-1464 211.757       852.80      FALSE
1451-1465 355.081        92.88       TRUE
1451-1466 453.550        63.92       TRUE
1464-1465 143.323       855.07      FALSE
1464-1466 241.792       852.41      FALSE
1465-1466  98.469        89.21       TRUE

Is there a positive or negative trend over time according to the Jonckheere-Terpstra test? YES

Plot and linear regression model for data inserted 7 days ago and last fetched 4 days ago

plot of chunk fetchplots plot of chunk fetchplots
Multiple comparison test after Kruskal-Wallis 
p.value: 0.05 
Comparisons
          obs.dif critical.dif difference
1352-1354  166.07       522.64      FALSE
1352-1443   10.83       678.26      FALSE
1352-1448  211.67       634.46      FALSE
1352-1449  182.47       606.66      FALSE
1352-1450  238.85       508.70      FALSE
1352-1451  381.28       481.98      FALSE
1352-1464  349.17       959.21      FALSE
1352-1465   23.23       486.10      FALSE
1352-1466   56.93       481.34      FALSE
1354-1443  155.24       522.64      FALSE
1354-1448   45.59       464.38      FALSE
1354-1449   16.39       425.61      FALSE
1354-1450   72.78       268.11      FALSE
1354-1451  215.21       213.11       TRUE
1354-1464  183.09       856.27      FALSE
1354-1465  142.84       222.27      FALSE
1354-1466  223.00       211.65       TRUE
1443-1448  200.83       634.46      FALSE
1443-1449  171.63       606.66      FALSE
1443-1450  228.02       508.70      FALSE
1443-1451  370.45       481.98      FALSE
1443-1464  338.33       959.21      FALSE
1443-1465   12.40       486.10      FALSE
1443-1466   67.76       481.34      FALSE
1448-1449   29.20       557.25      FALSE
1448-1450   27.19       448.63      FALSE
1448-1451  169.62       418.09      FALSE
1448-1464  137.50       928.75      FALSE
1448-1465  188.44       422.83      FALSE
1448-1466  268.60       417.35      FALSE
1449-1450   56.39       408.37      FALSE
1449-1451  198.82       374.56      FALSE
1449-1464  166.70       909.99      FALSE
1449-1465  159.24       379.85      FALSE
1449-1466  239.40       373.74      FALSE
1450-1451  142.43       176.17      FALSE
1450-1464  110.31       847.83      FALSE
1450-1465  215.62       187.15       TRUE
1450-1466  295.79       174.41       TRUE
1451-1464   32.12       832.07      FALSE
1451-1465  358.05        92.51       TRUE
1451-1466  438.22        62.86       TRUE
1464-1465  325.94       834.47      FALSE
1464-1466  406.10       831.70      FALSE
1465-1466   80.16        89.11      FALSE

Is there a positive or negative trend over time according to the Jonckheere-Terpstra test? YES

Plot and linear regression model for data inserted 15 days ago and last fetched 8 days ago

plot of chunk fetchplots plot of chunk fetchplots
Multiple comparison test after Kruskal-Wallis 
p.value: 0.05 
Comparisons
          obs.dif critical.dif difference
1352-1354  203.00       506.15      FALSE
1352-1443     NaN          Inf         NA
1352-1448     NaN          Inf         NA
1352-1449  262.50       603.54      FALSE
1352-1450  379.79       483.91      FALSE
1352-1451  451.80       458.51      FALSE
1352-1464     NaN          Inf         NA
1352-1465   81.23       463.54      FALSE
1352-1466   42.13       457.93      FALSE
1354-1443     NaN          Inf         NA
1354-1448     NaN          Inf         NA
1354-1449   59.50       451.83      FALSE
1354-1450  176.79       272.13      FALSE
1354-1451  248.80       223.87       TRUE
1354-1464     NaN          Inf         NA
1354-1465  121.77       233.99      FALSE
1354-1466  160.87       222.68      FALSE
1443-1448     NaN          Inf         NA
1443-1449     NaN          Inf         NA
1443-1450     NaN          Inf         NA
1443-1451     NaN          Inf         NA
1443-1464     NaN          Inf         NA
1443-1465     NaN          Inf         NA
1443-1466     NaN          Inf         NA
1448-1449     NaN          Inf         NA
1448-1450     NaN          Inf         NA
1448-1451     NaN          Inf         NA
1448-1464     NaN          Inf         NA
1448-1465     NaN          Inf         NA
1448-1466     NaN          Inf         NA
1449-1450  117.29       426.77      FALSE
1449-1451  189.30       397.74      FALSE
1449-1464     NaN          Inf         NA
1449-1465  181.27       403.52      FALSE
1449-1466  220.37       397.07      FALSE
1450-1451   72.01       167.63      FALSE
1450-1464     NaN          Inf         NA
1450-1465  298.56       180.92       TRUE
1450-1466  337.66       166.05       TRUE
1451-1464     NaN          Inf         NA
1451-1465  370.57        93.79       TRUE
1451-1466  409.67        60.29       TRUE
1464-1465     NaN          Inf         NA
1464-1466     NaN          Inf         NA
1465-1466   39.10        90.93      FALSE

Is there a positive or negative trend over time according to the Jonckheere-Terpstra test? YES

Plot and linear regression model for data inserted 31 days ago and last fetched 16 days ago

plot of chunk fetchplots plot of chunk fetchplots
Multiple comparison test after Kruskal-Wallis 
p.value: 0.05 
Comparisons
          obs.dif critical.dif difference
1352-1354   5.625       488.88      FALSE
1352-1443     NaN          Inf         NA
1352-1448     NaN          Inf         NA
1352-1449     NaN          Inf         NA
1352-1450 229.625       501.58      FALSE
1352-1451 272.507       450.86      FALSE
1352-1464 296.875       593.48      FALSE
1352-1465  58.272       457.07      FALSE
1352-1466 122.814       450.26      FALSE
1354-1443     NaN          Inf         NA
1354-1448     NaN          Inf         NA
1354-1449     NaN          Inf         NA
1354-1450 235.250       296.74      FALSE
1354-1451 278.132       199.37       TRUE
1354-1464 302.500       434.38      FALSE
1354-1465  52.647       213.03      FALSE
1354-1466 117.189       198.01      FALSE
1443-1448     NaN          Inf         NA
1443-1449     NaN          Inf         NA
1443-1450     NaN          Inf         NA
1443-1451     NaN          Inf         NA
1443-1464     NaN          Inf         NA
1443-1465     NaN          Inf         NA
1443-1466     NaN          Inf         NA
1448-1449     NaN          Inf         NA
1448-1450     NaN          Inf         NA
1448-1451     NaN          Inf         NA
1448-1464     NaN          Inf         NA
1448-1465     NaN          Inf         NA
1448-1466     NaN          Inf         NA
1449-1450     NaN          Inf         NA
1449-1451     NaN          Inf         NA
1449-1464     NaN          Inf         NA
1449-1465     NaN          Inf         NA
1449-1466     NaN          Inf         NA
1450-1451  42.882       228.76      FALSE
1450-1464  67.250       448.63      FALSE
1450-1465 287.897       240.75       TRUE
1450-1466 352.439       227.56       TRUE
1451-1464  24.368       391.10      FALSE
1451-1465 330.779        98.26       TRUE
1451-1466 395.321        59.01       TRUE
1464-1465 355.147       398.24      FALSE
1464-1466 419.689       390.41       TRUE
1465-1466  64.542        95.46      FALSE

Is there a positive or negative trend over time according to the Jonckheere-Terpstra test? YES

Plot and linear regression model for data inserted 63 days ago and last fetched 32 days ago

plot of chunk fetchplots plot of chunk fetchplots
Multiple comparison test after Kruskal-Wallis 
p.value: 0.05 
Comparisons
          obs.dif critical.dif difference
1352-1354     NaN          Inf         NA
1352-1443     NaN          Inf         NA
1352-1448     NaN          Inf         NA
1352-1449     NaN          Inf         NA
1352-1450     NaN          Inf         NA
1352-1451     NaN          Inf         NA
1352-1464     NaN          Inf         NA
1352-1465     NaN          Inf         NA
1352-1466     NaN          Inf         NA
1354-1443     NaN          Inf         NA
1354-1448     NaN          Inf         NA
1354-1449     NaN          Inf         NA
1354-1450     NaN          Inf         NA
1354-1451     NaN          Inf         NA
1354-1464     NaN          Inf         NA
1354-1465     NaN          Inf         NA
1354-1466     NaN          Inf         NA
1443-1448     NaN          Inf         NA
1443-1449     NaN          Inf         NA
1443-1450     NaN          Inf         NA
1443-1451     NaN          Inf         NA
1443-1464     NaN          Inf         NA
1443-1465     NaN          Inf         NA
1443-1466     NaN          Inf         NA
1448-1449     NaN          Inf         NA
1448-1450     NaN          Inf         NA
1448-1451     NaN          Inf         NA
1448-1464     NaN          Inf         NA
1448-1465     NaN          Inf         NA
1448-1466     NaN          Inf         NA
1449-1450     NaN          Inf         NA
1449-1451     NaN          Inf         NA
1449-1464     NaN          Inf         NA
1449-1465     NaN          Inf         NA
1449-1466     NaN          Inf         NA
1450-1451     NaN          Inf         NA
1450-1464     NaN          Inf         NA
1450-1465     NaN          Inf         NA
1450-1466     NaN          Inf         NA
1451-1464   109.0       413.80      FALSE
1451-1465   202.6       101.13       TRUE
1451-1466   361.8        54.88       TRUE
1464-1465   311.5       421.84      FALSE
1464-1466   470.7       413.20       TRUE
1465-1466   159.2        98.65       TRUE

Is there a positive or negative trend over time according to the Jonckheere-Terpstra test? YES

Plot and linear regression model for data inserted 127 days ago and last fetched 64 days ago

plot of chunk fetchplots plot of chunk fetchplots
Multiple comparison test after Kruskal-Wallis 
p.value: 0.05 
Comparisons
          obs.dif critical.dif difference
1352-1354     NaN          Inf         NA
1352-1443     NaN          Inf         NA
1352-1448     NaN          Inf         NA
1352-1449     NaN          Inf         NA
1352-1450     NaN          Inf         NA
1352-1451     NaN          Inf         NA
1352-1464     NaN          Inf         NA
1352-1465     NaN          Inf         NA
1352-1466     NaN          Inf         NA
1354-1443     NaN          Inf         NA
1354-1448     NaN          Inf         NA
1354-1449     NaN          Inf         NA
1354-1450     NaN          Inf         NA
1354-1451     NaN          Inf         NA
1354-1464     NaN          Inf         NA
1354-1465     NaN          Inf         NA
1354-1466     NaN          Inf         NA
1443-1448     NaN          Inf         NA
1443-1449     NaN          Inf         NA
1443-1450     NaN          Inf         NA
1443-1451     NaN          Inf         NA
1443-1464     NaN          Inf         NA
1443-1465     NaN          Inf         NA
1443-1466     NaN          Inf         NA
1448-1449     NaN          Inf         NA
1448-1450     NaN          Inf         NA
1448-1451     NaN          Inf         NA
1448-1464     NaN          Inf         NA
1448-1465     NaN          Inf         NA
1448-1466     NaN          Inf         NA
1449-1450     NaN          Inf         NA
1449-1451     NaN          Inf         NA
1449-1464     NaN          Inf         NA
1449-1465     NaN          Inf         NA
1449-1466     NaN          Inf         NA
1450-1451     NaN          Inf         NA
1450-1464     NaN          Inf         NA
1450-1465     NaN          Inf         NA
1450-1466     NaN          Inf         NA
1451-1464   6.009       302.15      FALSE
1451-1465  27.395        92.70      FALSE
1451-1466 292.936        50.79       TRUE
1464-1465  33.404       311.10      FALSE
1464-1466 298.945       301.28      FALSE
1465-1466 265.541        89.80       TRUE

Is there a positive or negative trend over time according to the Jonckheere-Terpstra test? YES

Plot and linear regression model for data inserted 255 days ago and last fetched 128 days ago

plot of chunk fetchplots plot of chunk fetchplots
Multiple comparison test after Kruskal-Wallis 
p.value: 0.05 
Comparisons
          obs.dif critical.dif difference
1352-1354     NaN          Inf         NA
1352-1443     NaN          Inf         NA
1352-1448     NaN          Inf         NA
1352-1449     NaN          Inf         NA
1352-1450     NaN          Inf         NA
1352-1451     NaN          Inf         NA
1352-1464     NaN          Inf         NA
1352-1465     NaN          Inf         NA
1352-1466     NaN          Inf         NA
1354-1443     NaN          Inf         NA
1354-1448     NaN          Inf         NA
1354-1449     NaN          Inf         NA
1354-1450     NaN          Inf         NA
1354-1451     NaN          Inf         NA
1354-1464     NaN          Inf         NA
1354-1465     NaN          Inf         NA
1354-1466     NaN          Inf         NA
1443-1448     NaN          Inf         NA
1443-1449     NaN          Inf         NA
1443-1450     NaN          Inf         NA
1443-1451     NaN          Inf         NA
1443-1464     NaN          Inf         NA
1443-1465     NaN          Inf         NA
1443-1466     NaN          Inf         NA
1448-1449     NaN          Inf         NA
1448-1450     NaN          Inf         NA
1448-1451     NaN          Inf         NA
1448-1464     NaN          Inf         NA
1448-1465     NaN          Inf         NA
1448-1466     NaN          Inf         NA
1449-1450     NaN          Inf         NA
1449-1451     NaN          Inf         NA
1449-1464     NaN          Inf         NA
1449-1465     NaN          Inf         NA
1449-1466     NaN          Inf         NA
1450-1451     NaN          Inf         NA
1450-1464     NaN          Inf         NA
1450-1465     NaN          Inf         NA
1450-1466     NaN          Inf         NA
1451-1464  136.26       203.94      FALSE
1451-1465   55.63        68.05      FALSE
1451-1466  164.73        45.80       TRUE
1464-1465   80.63       207.97      FALSE
1464-1466   28.47       201.78      FALSE
1465-1466  109.10        61.30       TRUE

Is there a positive or negative trend over time according to the Jonckheere-Terpstra test? YES

Data lifetime

plot of chunk lifetime

How long does it take for a request to complete?

The following graph plots how long it takes for an average request for a single block of data to complete.

plot of chunk request_time

Error: argument is of length zero

How long does it take to insert stuff?

plot of chunk insert_time

Multiple comparison test after Kruskal-Wallis 
p.value: 0.05 
Comparisons
          obs.dif critical.dif difference
1352-1354  62.250       596.07      FALSE
1352-1443  68.091       617.09      FALSE
1352-1448 184.250       723.60      FALSE
1352-1449 242.200       691.90      FALSE
1352-1450 365.458       580.17      FALSE
1352-1451 410.107       549.58      FALSE
1352-1464 575.000       723.60      FALSE
1352-1465 300.314       552.95      FALSE
1352-1466 335.398       548.68      FALSE
1354-1443   5.841       371.08      FALSE
1354-1448 122.000       529.62      FALSE
1354-1449 179.950       485.41      FALSE
1354-1450 303.208       305.78      FALSE
1354-1451 347.857       242.76       TRUE
1354-1464 512.750       529.62      FALSE
1354-1465 238.064       250.30      FALSE
1354-1466 273.148       240.73       TRUE
1443-1448 116.159       553.17      FALSE
1443-1449 174.109       511.00      FALSE
1443-1450 297.367       344.96      FALSE
1443-1451 342.016       290.57       TRUE
1443-1464 506.909       553.17      FALSE
1443-1465 232.223       296.90      FALSE
1443-1466 267.307       288.88      FALSE
1448-1449  57.950       635.55      FALSE
1448-1450 181.208       511.67      FALSE
1448-1451 225.857       476.69      FALSE
1448-1464 390.750       669.93      FALSE
1448-1465 116.064       480.58      FALSE
1448-1466 151.148       475.66      FALSE
1449-1450 123.258       465.75      FALSE
1449-1451 167.907       427.03      FALSE
1449-1464 332.800       635.55      FALSE
1449-1465  58.114       431.36      FALSE
1449-1466  93.198       425.88      FALSE
1450-1451  44.649       200.58      FALSE
1450-1464 209.542       511.67      FALSE
1450-1465  65.144       209.65      FALSE
1450-1466  30.060       198.12      FALSE
1451-1464 164.893       476.69      FALSE
1451-1465 109.793        96.87       TRUE
1451-1466  74.709        68.43       TRUE
1464-1465 274.686       480.58      FALSE
1464-1466 239.602       475.66      FALSE
1465-1466  35.084        91.67      FALSE

Is there a positive or negative trend over time according to the Jonckheere-Terpstra test?: No.

Bootstrapping

The following graph plots how long it takes for a node to bootstrap. That means, the time in seconds required to get 6 opennet peers or more via seed nodes. This data is, at present, pulled from the main Freenet developer (toad) who tries to bootstrap a node every day and checks how long this takes.

Error: Aesthetics must either be length one, or the same length as the
dataProblems:date, time

plot of chunk bootstrapping

Note: at about 07.06.2013 I started my own bootstrapping tests in addition to those of toad. Any improvement in the bootstrapping performance can be due to that fact.

Seed node statistics

The following graphs are based on data gathered from my own seednode and may not be representative for the network at large.

Opennet freenet versions connected to my seednode

plot of chunk versions

Number of Announcements and related messages sent

plot of chunk announcements

Peer status of connected opennet peers

plot of chunk peer_status

Author: digger3 - USK@zALLY9pbzMNicVn280HYqS2UkK0ZfX5LiTcln-cLrMU,GoLpCcShPzp3lbQSVClSzY7CH9c9HTw0qRLifBYqywY,AQACAAE/WebOfTrust/1895 .

I tend to hang around on Sone and freenode IRC