watchd/bokehwatch.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
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
#! /usr/bin/env python
# -*- coding: utf-8 -*-

# PHASE 1 : create datasource objects

import datetime

now = datetime.datetime.now()

def watchd_data ( input_file , threshold , low ) :
    data = dict(date=[], time=[], count=[], healthy=[], mean=[], sigma=[], two_sigma=[], sigma_up=[], sigma_down=[], decile=[], threshold=[], low=[])
    predict = dict(date=[], predict=[])
    with open(input_file) as fd :
        line = fd.readline()
        while line :
            items = line[:-1].split()
            date = datetime.datetime.strptime( "%s %s" % (items[0], items[1].split('.')[0]) , "%Y-%m-%d %H:%M:%S" )
            if (now - date).total_seconds() < 3*3600 :
              try :
                healthy = int(items[4]) - int(items[3])
                data['date'].append( date )
                data['time'].append( float(items[2]) )
                data['count'].append( int(items[4]) )
                #data['healthy'].append( int(items[4]) - int(items[3]) )
                #data['mean'].append( float(items[5]) * ( int(items[4]) - int(items[3]) ) )
                #data['sigma'].append( float(items[6]) * ( int(items[4]) - int(items[3]) ) )
                #data['decile'].append( float(items[7]) * ( int(items[4]) - int(items[3]) ) )
                data['healthy'].append( healthy )
                data['mean'].append( float(items[5]) * healthy )
                data['sigma'].append( float(items[6]) * healthy )
                data['decile'].append( float(items[7]) * healthy )
                #data['two_sigma'].append( ( float(items[5]) + 2 * float(items[6]) ) * ( int(items[4]) - int(items[3]) ) )
                #data['threshold'].append( threshold * ( int(items[4]) - int(items[3]) ) )
                #data['low'].append( low * ( int(items[4]) - int(items[3]) ) )
                data['two_sigma'].append( ( float(items[5]) + 2 * float(items[6]) ) * healthy )
                data['sigma_up'].append( ( float(items[5]) + float(items[6]) ) * healthy )
                data['sigma_down'].append( ( float(items[5]) - float(items[6]) ) * healthy )
                data['threshold'].append( threshold * healthy )
                data['low'].append( low * healthy )

                #predict['date'].append( date + datetime.timedelta(minutes=5) )
                predict['date'].append( date )
                predict['predict'].append( float(items[8]) * ( int(items[4]) - int(items[3]) ) )
              except Exception , ex :
                print "Se ha roto : '%s'" % ex
                print "           : %s" % line
            line = fd.readline()
    return ColumnDataSource(data) , ColumnDataSource(predict)

from bokeh.models import ColumnDataSource

#backend = ColumnDataSource( watchd_data('front-balancer.out', 80, 30) )
#recommender = ColumnDataSource( watchd_data('recommender.out', 60, 25) )
backend = watchd_data('backend.out', 80, 30)
#backend = watchd_data('front-balancer.out', 80, 30)
recommender = watchd_data('recommender.out', 60, 25)
#alpha = watchd_data('alpha-recommender.out', 60, 25)


# PHASE 2 : page/figure definition

from bokeh.plotting import figure, curdoc

p1 = figure(width=800, plot_height=600, title="backend", x_axis_type="datetime")
p2 = figure(width=800, plot_height=600, title="recommender", x_axis_type="datetime")
#p3 = figure(width=800, plot_height=600, title="alpha", x_axis_type="datetime")

for p,d in ( p1 , backend[0] ) , ( p2 , recommender[0] ) :
#for p,d in ( p1 , backend[0] ) , ( p2 , recommender[0] ) , ( p3 , alpha[0] ) :
    p.line(x='date', y='mean', source=d, color='navy', line_width=1, legend="mean")
    p.line(x='date', y='sigma', source=d, color='crimson', line_width=1, legend="sd")
    p.line(x='date', y='decile', source=d, color='chocolate', line_width=2, legend="10%")
    p.line(x='date', y='two_sigma', source=d, color='darkcyan', line_width=2, legend="two sigma")
    p.line(x='date', y='sigma_up', source=d, color='navy', line_width=2, legend="avg + sigma")
    #p.line(x='date', y='sigma_down', source=d, color='navy', line_width=2, legend="avg - sigma")
    p.line(x='date', y='threshold', source=d, color='red', line_width=3, legend="scaling threshold")
    jilito = p.line(x='date', y='low', source=d, color='pink', line_width=4, legend="lower scale")
    #p.line(x='date', y='count', source=d, color='crimson', line_width=2, legend="# nodes")

#p3.line(x='date', y='mean', source=alpha[0], color='navy', line_width=1, legend="mean")
#p3.line(x='date', y='sigma', source=alpha[0], color='crimson', line_width=1, legend="sd")
#p3.line(x='date', y='decile', source=alpha[0], color='chocolate', line_width=2, legend="10%")
#p3.line(x='date', y='two_sigma', source=alpha[0], color='darkcyan', line_width=2, legend="two sigma")

for p,d in ( p1 , backend[1] ) , ( p2 , recommender[1] ) :
    p.line(x='date', y='predict', source=d, color='green', line_width=2, legend="5 minute prediction")

for p in p1 , p2 : # , p3 :
    p.legend[0].location = "top_left"
    #p.legend[0].label_text_font_size = '6pt'
    #p.legend[0].legend_spacing = 0

#p.xaxis.formatter=DatetimeTickFormatter(formats=dict(
#        hours=["%d %B %Y"],
#        days=["%d %B %Y"],
#        months=["%d %B %Y"],
#        years=["%d %B %Y"],
#    ))
    p.xaxis.major_label_orientation = 3.141592/4

from bokeh.models import LinearAxis, Range1d

#for p,d,threshold in ( p1 , backend[0] , 80 ) , ( p2 , recommender[0] , 60 ) :
#    p.extra_y_ranges['elb_size'] = Range1d(start=0, end=max(d.data['count'])+1)
#    p.line(x='date', y='count', source=d, color='crimson', y_range_name="elb_size", line_width=2, legend="# nodes")
#    p.add_layout(LinearAxis(y_range_name="elb_size"), 'right')

#p3.extra_y_ranges['elb_size'] = Range1d(start=0, end=max(alpha[0].data['count'])+1)
#p3.line(x='date', y='count', source=alpha[0], color='crimson', y_range_name="elb_size", line_width=2, legend="# nodes")
#p3.add_layout(LinearAxis(y_range_name="elb_size"), 'right')


from bokeh.io import gridplot
#p = gridplot([[p1, p2],[p3]])
p = gridplot([[p1, p2],[]])
p.toolbar_location='above'


# PHASE 3 : application/session setup

from bokeh.plotting import output_file, show

output_file("/var/www/html/watchd.html", title='ELB cpu usage')#, mode='inline')
show(p)


#from bokeh.embed import file_html
#from bokeh.resources import Resources
#with open( '/var/www/html/watcher.html' , 'w' ) as fd :
#    fd.write( file_html( p3 , Resources() ) )


#import os
#from bokeh.embed import components
#
#for p,f in ( ( p1 , 'backend.js' ) , ( p2 , 'recommender.js' ) , ( p3 , 'alpha.js' ) ) :
#    with open(os.path.join('/var/www/html',f), 'w') as fd :
#        script, div = components(p)
#        print div
#        fd.write( script )