watchd/metric_test.py at master · Feverup/watchd · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
#!/usr/bin/python

import watchd.metrics
import sys

stream1 = "1465835356 9.220013e+01 8.465013e+01 9.305006e+01 8.128337e+01 9.184971e+01 8.361654e+01 8.761658e+01 7.796592e+01 8.441707e+01 7.660061e+01"
stream2 = "1465835356 8.923330e+01 8.089996e+01 8.853328e+01 7.746370e+01 8.991515e+01 8.106831e+01 8.345048e+01 7.833369e+01 8.696770e+01 8.078445e+01"
stream3 = "1465835416 8.719994e+01 8.958326e+01 8.764982e+01 8.616683e+01 8.005028e+01"
stream4 = "1465835416 8.374998e+01 8.623329e+01 8.074983e+01 8.426661e+01 7.838331e+01"

class test_metric ( watchd.metrics.aggregated_metric ) :

    @classmethod
    def from_datastream ( cls , name , config , *streams ) :
        obj = cls(name, config)
        for stream in streams :
            tstamp , values = stream.split(None,1)
            tstamp = int(tstamp)
            for val in values.split() :
                obj[tstamp] = val
                tstamp += 60
        return obj

wstream1 = "1466008406 4 13.0 15.0 16.0 17.0"
wstream2 = "1466008466 2 28.0 32.0"

class test_weighted ( watchd.metrics.weighted_metric ) :

    @classmethod
    def from_datastream ( cls , name , config , *streams ) :
        obj = cls(name, config)
        for stream in streams :
            tstamp , healthy , values = stream.split(None,2)
            tstamp = int(tstamp)
            weight = float(healthy)
            for val in values.split() :
                obj[tstamp] = val , weight
        return obj


tol = 1e-4

if __name__ == "__main__" :
    config = { 'standard_test': { 'metric_list':'' , 'alarms': [ { 'alarm':'test' ,  'statistics':'' , 'action':'autoscale:' } ] },
               'weight_test':   { 'metric_list':'' , 'alarms': [ { 'alarm':'test' , 'statistics':'' , 'action':'autoscale:' } ] }
               }

    metric = test_metric.from_datastream ( 'standard_test' , config , stream1 , stream2 , stream3 , stream4 )
    t_predict = 1465835837 + 300
    if len(sys.argv) > 1 :
        print 'y <- matrix(nrow=0, ncol=2)'
        for date in metric.keys() :
            for val in metric[date] :
                print 'y <- rbind( y , c( %s , %s ) )' % ( date , val )
        print 'y <- data.frame(y)'
        print 'fit <- lm( X2 ~ X1 , data=y )'
        print 'par(mfrow=c(2,2))'
        print 'plot(fit)'
        print 'predict(fit, newdata=data.frame(X1=c(%s)))' % (t_predict)
        print
    predict = metric.predict( t_predict , False )
    if abs( predict - 76.77483 ) > tol :
        print "Wrong prediction : %s vs. 76.77483" % predict
    else :
        print "Acceptable prediction : %s vs. 76.77483" % predict

    metric = test_weighted.from_datastream ( 'weight_test' , config , wstream1 , wstream2 )
    mean , sd = metric.mean(-1)
    if abs( mean - 18.2 ) > tol or abs( sd - 6.11228 ) > tol :
        print "Wrong weighted average : %s / %s vs. 18.2 / 6.11" % metric.mean(-1)
    else :
        print "Acceptable weighted average : %s / %s vs. 18.2 / 6.11" % metric.mean(-1)