File: analyze_logs.py

package info (click to toggle)
android-platform-development 8.1.0%2Br23-1
  • links: PTS, VCS
  • area: main
  • in suites: buster
  • size: 37,780 kB
  • sloc: ansic: 166,133; xml: 75,737; java: 19,329; python: 11,511; cpp: 8,221; sh: 2,075; lisp: 261; ruby: 183; asm: 132; perl: 129; makefile: 18
file content (262 lines) | stat: -rwxr-xr-x 6,972 bytes parent folder | download | duplicates (6)
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
141
142
143
144
145
146
147
148
149
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
259
260
261
262
#!/usr/bin/env python2.7

import argparse
import datetime
import re
import subprocess
import sys

import logs
import ps

DURATION_RE = re.compile("((\\d+)w)?((\\d+)d)?((\\d+)h)?((\\d+)m)?((\\d+)s)?")

class Bucket(object):
  """Bucket of stats for a particular key managed by the Stats object."""
  def __init__(self):
    self.count = 0
    self.memory = 0
    self.lines = []

  def __str__(self):
    return "(%s,%s)" % (self.count, self.memory)


class Stats(object):
  """A group of stats with a particular key, where both memory and count are tracked."""
  def __init__(self):
    self._data = dict()

  def add(self, key, logLine):
    bucket = self._data.get(key)
    if not bucket:
      bucket = Bucket()
      self._data[key] = bucket
    bucket.count += 1
    bucket.memory += logLine.memory()
    bucket.lines.append(logLine)

  def __iter__(self):
    return self._data.iteritems()

  def data(self):
    return [(key, bucket) for key, bucket in self._data.iteritems()]

  def byCount(self):
    result = self.data()
    result.sort(lambda a, b: -cmp(a[1].count, b[1].count))
    return result

  def byMemory(self):
    result = self.data()
    result.sort(lambda a, b: -cmp(a[1].memory, b[1].memory))
    return result


def ParseDuration(s):
  """Parse a date of the format .w.d.h.m.s into the number of seconds."""
  def make_int(index):
    val = m.group(index)
    if val:
      return int(val)
    else:
      return 0
  m = DURATION_RE.match(s)
  if m:
    weeks = make_int(2)
    days = make_int(4)
    hours = make_int(6)
    minutes = make_int(8)
    seconds = make_int(10)
    return (weeks * 604800) + (days * 86400) + (hours * 3600) + (minutes * 60) + seconds
  return 0

def FormatMemory(n):
  """Prettify the number of bytes into gb, mb, etc."""
  if n >= 1024 * 1024 * 1024:
    return "%10d gb" % (n / (1024 * 1024 * 1024))
  elif n >= 1024 * 1024:
    return "%10d mb" % (n / (1024 * 1024))
  elif n >= 1024:
    return "%10d kb" % (n / 1024)
  else:
    return "%10d b " % n

def FormateTimeDelta(td):
  """Format a time duration into the same format we accept on the commandline."""
  seconds = (td.days * 86400) + (td.seconds) + int(td.microseconds / 1000000)
  if seconds == 0:
    return "0s"
  result = ""
  if seconds >= 604800:
    weeks = int(seconds / 604800)
    seconds -= weeks * 604800
    result += "%dw" % weeks
  if seconds >= 86400:
    days = int(seconds / 86400)
    seconds -= days * 86400
    result += "%dd" % days
  if seconds >= 3600:
    hours = int(seconds / 3600)
    seconds -= hours * 3600
    result += "%dh" % hours
  if seconds >= 60:
    minutes = int(seconds / 60)
    seconds -= minutes * 60
    result += "%dm" % minutes
  if seconds > 0:
    result += "%ds" % seconds
  return result


def WriteResult(totalCount, totalMemory, bucket, text):
  """Write a bucket in the normalized format."""
  print "%7d (%2d%%) %s (%2d%%)  %s" % (bucket.count, (100 * bucket.count / totalCount),
      FormatMemory(bucket.memory), (100 * bucket.memory / totalMemory), text)
  

def ParseArgs(argv):
  parser = argparse.ArgumentParser(description="Process some integers.")
  parser.add_argument("input", type=str, nargs="?",
                      help="the logs file to read")
  parser.add_argument("--clear", action="store_true",
                      help="clear the log buffer before running logcat")
  parser.add_argument("--duration", type=str, nargs=1,
                      help="how long to run for (XdXhXmXs)")
  parser.add_argument("--rawlogs", type=str, nargs=1,
                      help="file to put the rawlogs into")

  args = parser.parse_args()

  args.durationSec = ParseDuration(args.duration[0]) if args.duration else 0

  return args


def main(argv):
  args = ParseArgs(argv)

  processes = ps.ProcessSet()

  if args.rawlogs:
    rawlogs = file(args.rawlogs[0], "w")
  else:
    rawlogs = None

  # Choose the input
  if args.input:
    # From a file of raw logs
    try:
      infile = file(args.input, "r")
    except IOError:
      sys.stderr.write("Error opening file for read: %s\n" % args.input[0])
      sys.exit(1)
  else:
    # From running adb logcat on an attached device
    if args.clear:
      subprocess.check_call(["adb", "logcat", "-c"])
    cmd = ["adb", "logcat", "-v", "long", "-D", "-v", "uid"]
    if not args.durationSec:
      cmd.append("-d")
    logcat = subprocess.Popen(cmd, stdout=subprocess.PIPE)
    infile = logcat.stdout

    # Do one update because we know we'll need it, but then don't do it again
    # if we're not streaming them.
    processes.Update(True)
    if args.durationSec:
      processes.doUpdates = True

  totalCount = 0
  totalMemory = 0
  byTag = Stats()
  byPid = Stats()
  byText = Stats()

  startTime = datetime.datetime.now()

  # Read the log lines from the parser and build a big mapping of everything
  for logLine in logs.ParseLogcat(infile, processes, args.durationSec):
    if rawlogs:
      rawlogs.write("%-10s %s %-6s %-6s %-6s %s/%s: %s\n" %(logLine.buf, logLine.timestamp,
          logLine.uid, logLine.pid, logLine.tid, logLine.level, logLine.tag, logLine.text))
    
    totalCount += 1
    totalMemory += logLine.memory()
    byTag.add(logLine.tag, logLine)
    byPid.add(logLine.pid, logLine)
    byText.add(logLine.text, logLine)

  endTime = datetime.datetime.now()

  # Print the log analysis

  # At this point, everything is loaded, don't bother looking
  # for new processes
  processes.doUpdates = False

  print "Top tags by count"
  print "-----------------"
  i = 0
  for k,v in byTag.byCount():
    WriteResult(totalCount, totalMemory, v, k)
    if i >= 10:
      break
    i += 1

  print
  print "Top tags by memory"
  print "------------------"
  i = 0
  for k,v in byTag.byMemory():
    WriteResult(totalCount, totalMemory, v, k)
    if i >= 10:
      break
    i += 1

  print
  print "Top Processes by memory"
  print "-----------------------"
  i = 0
  for k,v in byPid.byMemory():
    WriteResult(totalCount, totalMemory, v,
        "%-8s %s" % (k, processes.FindPid(k).DisplayName()))
    if i >= 10:
      break
    i += 1

  print
  print "Top Duplicates by count"
  print "-----------------------"
  i = 0
  for k,v in byText.byCount():
    logLine = v.lines[0]
    WriteResult(totalCount, totalMemory, v,
        "%s/%s: %s" % (logLine.level, logLine.tag, logLine.text))
    if i >= 10:
      break
    i += 1

  print
  print "Top Duplicates by memory"
  print "-----------------------"
  i = 0
  for k,v in byText.byCount():
    logLine = v.lines[0]
    WriteResult(totalCount, totalMemory, v,
        "%s/%s: %s" % (logLine.level, logLine.tag, logLine.text))
    if i >= 10:
      break
    i += 1

  print
  print "Totals"
  print "------"
  print "%7d  %s" % (totalCount, FormatMemory(totalMemory))

  print "Actual duration: %s" % FormateTimeDelta(endTime-startTime)

if __name__ == "__main__":
  main(sys.argv)

# vim: set ts=2 sw=2 sts=2 tw=100 nocindent autoindent smartindent expandtab: