Class: Pwrake::Report

Inherits:
Object
  • Object
show all
Defined in:
lib/pwrake/report/report.rb

Constant Summary collapse

HTML_HEAD = 
"<html><head><style>\n<!--\nh2 {\n  background-color:#eee;\n}\nh1 {\n  background-color:#0ff;\n}\ntable {\n margin:0px;\n border-style:solid;\n border-width:1px;\n}\ntd {\n  margin:0px;\n  border-style:solid;\n  border-width:1px;\n}\n-->\n</style>\n</head>\n<body>\n"
@@id = 
0
@@id_fmt = 
nil

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(base, pattern) ⇒ Report

Returns a new instance of Report.



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
 
# File 'lib/pwrake/report/report.rb', line 33

def initialize(base,pattern)
  @base = base
  @pattern = pattern

  @@id = @@id.succ
  @id = @@id

  @csv_file = base+'.csv'
  @task_file = base+'.task'
  @html_file = base+'.html'

  open(@base+".log","r").each do |s|
    if /num_cores=(\d+)/ =~ s
      @ncore = $1.to_i
      break
    end
  end

  @sh_table = CSV.read(@csv_file,:headers=>true)
  h = {}
  @sh_table.each do |row|
    if host = row['host']
      h[host] = true
    end
  end
  @hosts = h.keys.sort
  @start_time = Time.parse(@sh_table[0]["start_time"])
  @end_time = Time.parse(@sh_table[-1]["start_time"])
  @elap = @end_time - @start_time
  read_elap_each_cmd
  make_cmd_stat
end

Instance Attribute Details

#base ⇒ Object (readonly)

Returns the value of attribute base.



66
67
68
 
# File 'lib/pwrake/report/report.rb', line 66

def base
  @base
end

#cmd_elap ⇒ Object (readonly)

Returns the value of attribute cmd_elap.



68
69
70
 
# File 'lib/pwrake/report/report.rb', line 68

def cmd_elap
  @cmd_elap
end

#cmd_stat ⇒ Object (readonly)

Returns the value of attribute cmd_stat.



68
69
70
 
# File 'lib/pwrake/report/report.rb', line 68

def cmd_stat
  @cmd_stat
end

#csv_file ⇒ Object (readonly)

Returns the value of attribute csv_file.



67
68
69
 
# File 'lib/pwrake/report/report.rb', line 67

def csv_file
  @csv_file
end

#elap ⇒ Object (readonly)

Returns the value of attribute elap.



66
67
68
 
# File 'lib/pwrake/report/report.rb', line 66

def elap
  @elap
end

#html_file ⇒ Object (readonly)

Returns the value of attribute html_file.



67
68
69
 
# File 'lib/pwrake/report/report.rb', line 67

def html_file
  @html_file
end

#id ⇒ Object (readonly)

Returns the value of attribute id.



70
71
72
 
# File 'lib/pwrake/report/report.rb', line 70

def id
  @id
end

#ncore ⇒ Object (readonly)

Returns the value of attribute ncore.



66
67
68
 
# File 'lib/pwrake/report/report.rb', line 66

def ncore
  @ncore
end

#sh_table ⇒ Object (readonly)

Returns the value of attribute sh_table.



69
70
71
 
# File 'lib/pwrake/report/report.rb', line 69

def sh_table
  @sh_table
end

#task_table ⇒ Object (readonly)

Returns the value of attribute task_table.



69
70
71
 
# File 'lib/pwrake/report/report.rb', line 69

def task_table
  @task_table
end

Instance Method Details

#format_comma(x) ⇒ Object 



137
138
139
 
# File 'lib/pwrake/report/report.rb', line 137

def format_comma(x)
  x.to_s.gsub(/(?<=\d)(?=(?:\d\d\d)+(?!\d))/, ',')
end

#get_command(s) ⇒ Object 



112
113
114
115
116
117
118
119
120
121
122
123
124
125
 
# File 'lib/pwrake/report/report.rb', line 112

def get_command(s)
  if /\(([^()]+)\)/ =~ s
    s = $1
  end
  a = s.split(/;/)
  a.each do |x|
    if /^\s*(\S+)/ =~ x
      k = $1
      next if k=='cd'
      return k
    end
  end
  nil
end

#histogram_plot ⇒ Object 



315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
 
# File 'lib/pwrake/report/report.rb', line 315

def histogram_plot
  command_list = []
  @cmd_stat.each do |cmd,stat|
    if stat.n > 2
      command_list << cmd
    end
  end
  hist_image = @base+"_hist.png"
  IO.popen("gnuplot","r+") do |f|
    f.puts "
set terminal png # size 480,360
set output '#{hist_image}'
set ylabel 'histogram'
set xlabel 'Execution time (sec)'
set logscale x
set title 'histogram of elapsed time'"
    a = []

    command_list.each_with_index do |n,i|
      a << "'-' w histeps ls #{i+1} title ''"
      a << "'-' w lines ls #{i+1} title '#{n}'"
    end
    f.puts "plot "+ a.join(',')

    command_list.each do |cmd|
      stat = @cmd_stat[cmd]
      2.times do
        stat.hist_each do |x1,x2,y|
          x = Math.sqrt(x1*x2)
          f.printf "%f %d\n", x, y
        end
        f.puts "e"
      end
    end
    hist_image
  end
end

#id_str ⇒ Object 



72
73
74
75
76
77
78
 
# File 'lib/pwrake/report/report.rb', line 72

def id_str
  if @@id_fmt.nil?
    id_len = Math.log10(@@id).floor + 1
    @@id_fmt = "#%0#{id_len}d"
  end
  @@id_fmt % @id
end

#make_cmd_stat ⇒ Object 



127
128
129
130
131
132
133
134
135
 
# File 'lib/pwrake/report/report.rb', line 127

def make_cmd_stat
  @cmd_stat = {}
  @cmd_elap.each do |cmd,elap|
    @cmd_stat[cmd] = s = Stat.new(elap)
    if elap.size > 1
      s.make_logx_histogram(1.0/8)
    end
  end
end

#read_elap_each_cmd ⇒ Object 



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
 
# File 'lib/pwrake/report/report.rb', line 80

def read_elap_each_cmd
  @cmd_elap = {}
  @sh_table.each do |row|
    command = row['command']
    elap = row['elap_time']
    if command && elap
      elap = elap.to_f
      found = nil
      @pattern.each do |cmd,regex|
        if regex =~ command
          if a = @cmd_elap[cmd]
            a << elap
          else
            @cmd_elap[cmd] = [elap]
          end
          found = true
        end
      end
      if !found
        if cmd = get_command( command )
          if a = @cmd_elap[cmd]
            a << elap
          else
            @cmd_elap[cmd] = [elap]
          end
        end
      end
    end
  end
  @cmd_elap
end

#report_html ⇒ Object 



150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
 
# File 'lib/pwrake/report/report.rb', line 150

def report_html
  html = HTML_HEAD + "<body><h1>Pwrake Statistics</h1>\n"
  html << "<h2>Workflow</h2>\n"
  html << "<table>\n"
  html << "<tr><th>log file</th><td>#{@base}</td><tr>\n"
  html << "<tr><th>ncore</th><td>#{@ncore}</td><tr>\n"
  html << "<tr><th>elapsed time(sec)</th><td>#{@elap}</td><tr>\n"
  html << "<tr><th>start time</th><td>#{@start_time}</td><tr>\n"
  html << "<tr><th>end time</th><td>#{@end_time}</td><tr>\n"
  html << "</table>\n"
  html << "<table>\n"
  html << "<tr><th>hosts</th><tr>\n"
  @hosts.each do |h|
    html << "<tr><td>#{h}</td><tr>\n"
  end
  html << "</table>\n"
  html << "<h2>Parallelism</h2>\n"
  fimg = Parallelism.plot_parallelism2(@csv_file)
  html << "<img src='#{fimg}' align='top'/></br>\n"

  html << "<h2>Parallelism by host</h2>\n"
  fimg2 = Parallelism.plot_parallelizm_by_host(@sh_table,@base)
  html << "<img src='#{fimg2}' align='top'/></br>\n"

  html << "<h2>Command statistics</h2>\n"
  html << "<table>\n"
  html << "<tr><th>command</th>"
  html << Stat.html_th
  html << "</tr>\n"
  @cmd_stat.each do |cmd,stat|
    html << "<tr><td>#{cmd}</td>"
    html << stat.html_td
    html << "</tr>\n"
  end
  html << "<table>\n"
  html << "<img src='#{histogram_plot}' align='top'/></br>\n"

  task_locality
  html << "<h2>Locality statistics</h2>\n"
  html << "<table>\n"

  html << "<tr><th></th><th rowspan=3>gross elapsed time (sec)</th><th></th>"
  html << "<th colspan=6>read</th>"
  html << "<th></th>"
  html << "<th colspan=6>write</th>"
  html << "</tr>\n"


  html << "<tr><th></th><th></th>"
  html << "<th colspan=3>count</th><th colspan=3>file size (bytes)</th>"
  html << "<th></th>"
  html << "<th colspan=3>count</th><th colspan=3>file size (bytes)</th>"
  html << "</tr>\n"

  html << "<tr><th>host</th><th></th>"
  html << "<th>local</th><th>remote</th><th>total</th>"
  html << "<th>local</th><th>remote</th><th>total</th>"
  html << "<th></th>"
  html << "<th>local</th><th>remote</th><th>total</th>"
  html << "<th>local</th><th>remote</th><th>total</th>"
  html << "</tr>\n"
  n_same_input     = 0
  size_same_input  = 0
  n_diff_input     = 0
  size_diff_input  = 0
  n_same_output    = 0
  size_same_output = 0
  n_diff_output    = 0
  size_diff_output = 0
  elap_host = 0
  @exec_hosts.each do |h|
    html << "<tr><td>#{h}</td>"
    html << "<td align='right'>%.3f</td>" % @elap_host[h]
    html << "<td></td>"
    html << tr_count(@n_same_input[h],@n_diff_input[h])
    html << tr_count(@size_same_input[h],@size_diff_input[h])
    html << "<td></td>"
    html << tr_count(@n_same_output[h],@n_diff_output[h])
    html << tr_count(@size_same_output[h],@size_diff_output[h])

    html << "</tr>\n"
    n_same_input     += @n_same_input[h]
    size_same_input  += @size_same_input[h]
    n_diff_input     += @n_diff_input[h]
    size_diff_input  += @size_diff_input[h]
    n_same_output    += @n_same_output[h]
    size_same_output += @size_same_output[h]
    n_diff_output    += @n_diff_output[h]
    size_diff_output += @size_diff_output[h]
    elap_host += @elap_host[h]
  end
  html << "<tr><td>total</td>"
  html << "<td align='right'>%.3f</td>" % elap_host
  html << "<td></td>"
  html << tr_count(n_same_input,n_diff_input)
  html << tr_count(size_same_input,size_diff_input)
  html << "<td></td>"
  html << tr_count(n_same_output,n_diff_output)
  html << tr_count(size_same_output,size_diff_output)

  html << "</tr>\n"
  html << "<table>\n"

  html << "</body></html>\n"
  File.open(@html_file,"w") do |f|
    f.puts html
  end
  puts "generate "+@html_file
end

#task_locality ⇒ Object 



260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
 
# File 'lib/pwrake/report/report.rb', line 260

def task_locality
  @task_table = CSV.read(@task_file,:headers=>true)
  file_size = {}
  file_host = {}
  @task_table.each do |row|
    name = row['task_name']
    file_size[name] = row['file_size'].to_i
    file_host[name] = (row['file_host']||'').split('|')
  end

  @n_same_output   = Hash.new(0)
  @size_same_output= Hash.new(0)
  @n_diff_output   = Hash.new(0)
  @size_diff_output= Hash.new(0)
  @n_same_input    = Hash.new(0)
  @size_same_input = Hash.new(0)
  @n_diff_input    = Hash.new(0)
  @size_diff_input = Hash.new(0)
  h = {}
  @task_table.each do |row|
    if row['executed']=='1'
      name = row['task_name']
      exec_host = row['exec_host']
      h[exec_host] = true
      if file_host[name].include?(exec_host)
        @n_same_output[exec_host] += 1
        @size_same_output[exec_host] += file_size[name]
      else
        @n_diff_output[exec_host] += 1
        @size_diff_output[exec_host] += file_size[name]
      end
      preq_files = (row['preq']||'').split('|')
      preq_files.each do |preq|
        if (sz=file_size[preq]) && sz > 0
          if file_host[preq].include?(exec_host)
            @n_same_input[exec_host] += 1
            @size_same_input[exec_host] += sz
          else
            @n_diff_input[exec_host] += 1
            @size_diff_input[exec_host] += sz
          end
        end
      end
    end
  end
  @exec_hosts = h.keys.sort

  @elap_host = Hash.new(0)
  @sh_table.each do |row|
    if (h = row['host']) && (t = row['elap_time'])
      @elap_host[h] += t.to_f
    end
  end
end

#tr_count(x, y) ⇒ Object 



141
142
143
144
145
146
147
148
 
# File 'lib/pwrake/report/report.rb', line 141

def tr_count(x,y)
  sum = x+y
  td = "<td align='right' valign='top'>"
  m  = td + "%s<br/>%.2f%%</td>" % [format_comma(x),x*100.0/sum]
  m << td + "%s<br/>%.2f%%</td>" % [format_comma(y),y*100.0/sum]
  m << td + "%s</td>" % format_comma(sum)
  m
end