web_log plugin: initial version added
Ilya committed
Feb 8, 2017 at 21:18 UTC
e5b4041f9b5c04b51b9d73b0f7239d69d8e52a25
2 files changed
+337
conf.d/python.d/web_log.conf
new
+77
@@ -0,0 +1,77 @@
1
+# netdata python.d.plugin configuration for web log
2
+#
3
+# This file is in YaML format. Generally the format is:
4
+#
5
+# name: value
6
+#
7
+# There are 2 sections:
8
+# - global variables
9
+# - one or more JOBS
10
+#
11
+# JOBS allow you to collect values from multiple sources.
12
+# Each source will have its own set of charts.
13
+#
14
+# JOB parameters have to be indented (using spaces only, example below).
15
+
16
+# ----------------------------------------------------------------------
17
+# Global Variables
18
+# These variables set the defaults for all JOBs, however each JOB
19
+# may define its own, overriding the defaults.
20
+
21
+# update_every sets the default data collection frequency.
22
+# If unset, the python.d.plugin default is used.
23
+# update_every: 1
24
+
25
+# priority controls the order of charts at the netdata dashboard.
26
+# Lower numbers move the charts towards the top of the page.
27
+# If unset, the default for python.d.plugin is used.
28
+# priority: 60000
29
+
30
+# retries sets the number of retries to be made in case of failures.
31
+# If unset, the default for python.d.plugin is used.
32
+# Attempts to restore the service are made once every update_every
33
+# and only if the module has collected values in the past.
34
+# retries: 5
35
+
36
+# ----------------------------------------------------------------------
37
+# JOBS (data collection sources)
38
+#
39
+# The default JOBS share the same *name*. JOBS with the same name
40
+# are mutually exclusive. Only one of them will be allowed running at
41
+# any time. This allows autodetection to try several alternatives and
42
+# pick the one that works.
43
+#
44
+# Any number of jobs is supported.
45
+#
46
+# All python.d.plugin JOBS (for all its modules) support a set of
47
+# predefined parameters. These are:
48
+#
49
+# job_name:
50
+# name: myname # the JOB's name as it will appear at the
51
+# # dashboard (by default is the job_name)
52
+# # JOBs sharing a name are mutually exclusive
53
+# update_every: 1 # the JOB's data collection frequency
54
+# priority: 60000 # the JOB's order on the dashboard
55
+# retries: 5 # the JOB's number of restoration attempts
56
+#
57
+# Additionally to the above, web_log also supports the following:
58
+#
59
+# path: 'PATH' # the path to web server log file
60
+# detailed_response_codes: yes/no # Default: yes. Additional chart where response codes are not grouped
61
+# all_time : yes/no # Default: yes. All time unique client IPs chart (50000 addresses ~ 400KB)
62
+# categories: # requests per url chart configuration
63
+# cacti: 'cacti.*' # name(dimension): REGEX to match
64
+# observium: 'observium.*' # name(dimension): REGEX to match
65
+# stub_status: 'stub_status' # name(dimension): REGEX to match
66
+#
67
+
68
+# ----------------------------------------------------------------------
69
+# AUTO-DETECTION JOBS
70
+# only one of them will run (they have the same name)
71
+nginx_log:
72
+ name: 'nginx'
73
+ path: '/var/log/nginx/access.log'
74
+
75
+apache_log:
76
+ name: 'apache_vhosts'
77
+ path: '/var/log/apache2/other_vhosts_access.log'
python.d/web_log.chart.py
new
+260
@@ -0,0 +1,260 @@
1
+# -*- coding: utf-8 -*-
2
+# Description: web log netdata python.d module
3
+# Author: l2isbad
4
+
5
+from base import LogService
6
+import re
7
+import bisect
8
+from os import access, R_OK
9
+from os.path import getsize
10
+from collections import defaultdict, namedtuple
11
+
12
+priority = 60000
13
+retries = 60
14
+
15
+ORDER = ['response_codes', 'request_time', 'requests_per_url', 'http_method', 'bandwidth', 'clients_cur', 'clients_all']
16
+CHARTS = {
17
+ 'response_codes': {
18
+ 'options': [None, 'Response Codes', 'requests/s', 'responses', 'web_log.response', 'stacked'],
19
+ 'lines': [
20
+ ['2xx', '2xx', 'absolute'],
21
+ ['5xx', '5xx', 'absolute'],
22
+ ['3xx', '3xx', 'absolute'],
23
+ ['4xx', '4xx', 'absolute'],
24
+ ['1xx', '1xx', 'absolute'],
25
+ ['0xx', 'other', 'absolute']
26
+ ]},
27
+ 'bandwidth': {
28
+ 'options': [None, 'Bandwidth', 'KB/s', 'bandwidth', 'web_log.bw', 'area'],
29
+ 'lines': [
30
+ ['resp_length', 'received', 'absolute', 1, 1024],
31
+ ['bytes_sent', 'sent', 'absolute', -1, 1024]
32
+ ]},
33
+ 'request_time': {
34
+ 'options': [None, 'Processing Time', 'milliseconds', 'timings', 'web_log.request', 'area'],
35
+ 'lines': [
36
+ ['resp_time_min', 'min', 'absolute', 1, 1],
37
+ ['resp_time_max', 'max', 'absolute', 1, 1],
38
+ ['resp_time_avg', 'avg', 'absolute', 1, 1]
39
+ ]},
40
+ 'clients_cur': {
41
+ 'options': [None, 'Current Poll Unique Client IPs', 'unique ips', 'unique clients', 'web_log.clients', 'line'],
42
+ 'lines': [
43
+ ['unique_cur', 'clients', 'absolute', 1, 1]
44
+ ]},
45
+ 'clients_all': {
46
+ 'options': [None, 'All Time Unique Client IPs', 'unique ips', 'unique clients', 'web_log.clients', 'line'],
47
+ 'lines': [
48
+ ['unique_tot', 'clients', 'absolute', 1, 1]
49
+ ]},
50
+ 'http_method': {
51
+ 'options': [None, 'Requests Per HTTP Method', 'requests/s', 'requests', 'web_log.http_method', 'stacked'],
52
+ 'lines': [
53
+ ]}
54
+}
55
+
56
+NAMED_URL_PATTERN = namedtuple('URL_PATTERN', ['description', 'pattern'])
57
+
58
+
59
+class Service(LogService):
60
+ def __init__(self, configuration=None, name=None):
61
+ LogService.__init__(self, configuration=configuration, name=name)
62
+ # Vars from module configuration file
63
+ self.log_path = self.configuration.get('path')
64
+ self.detailed_response_codes = self.configuration.get('detailed_response_codes', True)
65
+ self.all_time = self.configuration.get('all_time', True)
66
+ self.url_pattern = self.configuration.get('categories') # dict
67
+ # REGEX: 1.IPv4 address 2.HTTP method 3. URL 4. Response code
68
+ # 5. Bytes sent 6. Response length 7. Response process time
69
+ #self.regex = re.compile(r'(\d{1,3}(?:\.\d{1,3}){3}) .*?"([A-Z]+) (.*?)" ([1-9]\d{2}) (\d+) (\d+)? ?([\d.]+)? ?"(.*?)" "(.*?)"')
70
+ self.regex = re.compile(r'(\d{1,3}(?:\.\d{1,3}){3}) .*?"([A-Z]+) (.*?)" ([1-9]\d{2}) (\d+) (\d+)? ?([\d.]+)?')
71
+ # sorted list of unique IPs
72
+ self.unique_all_time = list()
73
+ # dict for values that should not be zeroed every poll
74
+ self.storage = {'unique_tot': 0}
75
+ # if there is no new logs this dict + self.storage returned to netdata
76
+ self.data = {'bytes_sent': 0, 'resp_length': 0, 'resp_time_min': 0,
77
+ 'resp_time_max': 0, 'resp_time_avg': 0, 'unique_cur': 0,
78
+ '2xx': 0, '5xx': 0, '3xx': 0, '4xx': 0, '1xx': 0, '0xx': 0}
79
+
80
+ def check(self):
81
+ if not self.log_path:
82
+ self.error('log path is not specified')
83
+ return False
84
+
85
+ # log_path must be readable
86
+ if not access(self.log_path, R_OK):
87
+ self.error('%s not readable or not exist' % self.log_path)
88
+ return False
89
+
90
+ # log_path file should not be empty
91
+ if not getsize(self.log_path):
92
+ self.error('%s is empty' % self.log_path)
93
+ return False
94
+
95
+ # Read last line (or first if there is only one line)
96
+ with open(self.log_path, 'rb') as logs:
97
+ logs.seek(-2, 2)
98
+ while logs.read(1) != b'\n':
99
+ logs.seek(-2, 1)
100
+ if logs.tell() == 0:
101
+ break
102
+ last_line = logs.readline().decode(encoding='utf-8')
103
+
104
+ # Parse last line
105
+ parsed_line = self.regex.findall(last_line)
106
+ if not parsed_line:
107
+ self.error('Can\'t parse output')
108
+ return False
109
+
110
+ # parsed_line[0][6] - response process time
111
+ self.create_charts(parsed_line[0][6])
112
+ return True
113
+
114
+ def create_charts(self, parsed_line):
115
+ def find_job_name(override_name, name):
116
+ add_to_name = override_name or name
117
+ if add_to_name:
118
+ return '_'.join(['web_log', add_to_name])
119
+ else:
120
+ return 'web_log'
121
+
122
+ job_name = find_job_name(self.override_name, self.name)
123
+ self.detailed_chart = 'CHART %s.detailed_response_codes ""' \
124
+ ' "Response Codes" requests/s responses' \
125
+ ' web_log.detailed_resp stacked 1 %s\n' % (job_name, self.update_every)
126
+ self.http_method_chart = 'CHART %s.http_method' \
127
+ ' "" "HTTP Methods" requests/s requests' \
128
+ ' web_log.http_method stacked 2 %s\n' % (job_name, self.update_every)
129
+ self.order = ORDER[:]
130
+ self.definitions = CHARTS
131
+
132
+ # Remove 'request_time' chart from ORDER if request_time not in logs
133
+ if parsed_line == '':
134
+ self.order.remove('request_time')
135
+ # Remove 'clients_all' chart from ORDER if specified in the configuration
136
+ if not self.all_time:
137
+ self.order.remove('clients_all')
138
+ # Add 'detailed_response_codes' chart if specified in the configuration
139
+ if self.detailed_response_codes:
140
+ self.order.append('detailed_response_codes')
141
+ self.definitions['detailed_response_codes'] = {'options': [None, 'Detailed Response Codes', 'requests/s',
142
+ 'responses', 'web_log.detailed_resp', 'stacked'],
143
+ 'lines': []}
144
+
145
+ # Add 'requests_per_url' chart if specified in the configuration
146
+ if self.url_pattern:
147
+ self.url_pattern = [NAMED_URL_PATTERN(description=k, pattern=re.compile(v)) for k, v in self.url_pattern.items()]
148
+ self.definitions['requests_per_url'] = {'options': [None, 'Requests Per Url', 'requests/s',
149
+ 'requests', 'web_log.url_pattern', 'stacked'],
150
+ 'lines': [['other_url', 'other', 'absolute']]}
151
+ for elem in self.url_pattern:
152
+ self.definitions['requests_per_url']['lines'].append([elem.description, elem.description, 'absolute'])
153
+ self.data.update({elem.description: 0})
154
+ self.data.update({'other_url': 0})
155
+ else:
156
+ self.order.remove('requests_per_url')
157
+
158
+ def add_new_dimension(self, dimension, line_list, chart_string, key):
159
+ self.storage.update({dimension: 0})
160
+ # SET method check if dim in _dimensions
161
+ self._dimensions.append(dimension)
162
+ # UPDATE method do SET only if dim in definitions
163
+ self.definitions[key]['lines'].append(line_list)
164
+ chart = chart_string
165
+ chart += "%s %s\n" % ('DIMENSION', ' '.join(line_list))
166
+ print(chart)
167
+ return chart
168
+
169
+ def _get_data(self):
170
+ """
171
+ Parse new log lines
172
+ :return: dict
173
+ """
174
+ raw = self._get_raw_data()
175
+ if raw is None:
176
+ return None
177
+
178
+ request_time, unique_current = list(), list()
179
+ request_counter = {'count': 0, 'sum': 0}
180
+ to_netdata = dict()
181
+ to_netdata.update(self.data)
182
+ default_dict = defaultdict(lambda: 0)
183
+
184
+ for line in raw:
185
+ match = self.regex.findall(line)
186
+ if match:
187
+ match_dict = dict(zip('address method url code sent resp_length resp_time'.split(), match[0]))
188
+ try:
189
+ code = ''.join([match_dict['code'][0], 'xx'])
190
+ to_netdata[code] += 1
191
+ except KeyError:
192
+ to_netdata['0xx'] += 1
193
+ # detailed response code
194
+ if self.detailed_response_codes:
195
+ self._get_data_detailed_response_codes(match_dict['code'], default_dict)
196
+ # requests per url
197
+ if self.url_pattern:
198
+ self._get_data_per_url(match_dict['url'], default_dict)
199
+ # requests per http method
200
+ self._get_data_http_method(match_dict['method'], default_dict)
201
+
202
+ to_netdata['bytes_sent'] += int(match_dict['sent'])
203
+
204
+ if match_dict['resp_length'] != '' and match_dict['resp_time'] != '':
205
+ to_netdata['resp_length'] += int(match_dict['resp_length'])
206
+ resp_time = float(match_dict['resp_time']) * 1000
207
+ bisect.insort_left(request_time, resp_time)
208
+ request_counter['count'] += 1
209
+ request_counter['sum'] += resp_time
210
+ # unique clients ips
211
+ if address_not_in_pool(self.unique_all_time, match_dict['address'], self.storage['unique_tot']):
212
+ self.storage['unique_tot'] += 1
213
+ if address_not_in_pool(unique_current, match_dict['address'], to_netdata['unique_cur']):
214
+ to_netdata['unique_cur'] += 1
215
+ # timings
216
+ if request_time:
217
+ to_netdata['resp_time_min'] = request_time[0]
218
+ to_netdata['resp_time_avg'] = float(request_counter['sum']) / request_counter['count']
219
+ to_netdata['resp_time_max'] = request_time[-1]
220
+
221
+ to_netdata.update(self.storage)
222
+ to_netdata.update(default_dict)
223
+ return to_netdata
224
+
225
+ def _get_data_detailed_response_codes(self, code, default_dict):
226
+ if code not in self.storage:
227
+ chart_string_copy = self.detailed_chart
228
+ self.detailed_chart = self.add_new_dimension(code, [code, code, 'absolute'],
229
+ chart_string_copy, 'detailed_response_codes')
230
+ default_dict[code] += 1
231
+
232
+ def _get_data_http_method(self, method, default_dict):
233
+ if method not in self.storage:
234
+ chart_string_copy = self.http_method_chart
235
+ self.http_method_chart = self.add_new_dimension(method, [method, method, 'absolute'],
236
+ chart_string_copy, 'http_method')
237
+ default_dict[method] += 1
238
+
239
+ def _get_data_per_url(self, url, default_dict):
240
+ match = None
241
+ for elem in self.url_pattern:
242
+ if elem.pattern.search(url):
243
+ default_dict[elem.description] += 1
244
+ match = True
245
+ break
246
+ if not match:
247
+ default_dict['other_url'] += 1
248
+
249
+
250
+def address_not_in_pool(pool, address, pool_size):
251
+ index = bisect.bisect_left(pool, address)
252
+ if index < pool_size:
253
+ if pool[index] == address:
254
+ return False
255
+ else:
256
+ bisect.insort_left(pool, address)
257
+ return True
258
+ else:
259
+ bisect.insort_left(pool, address)
260
+ return True