forked from netdata/netdata
-
Notifications
You must be signed in to change notification settings - Fork 0
/
varnish.chart.py
308 lines (265 loc) · 10.3 KB
/
varnish.chart.py
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
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
# -*- coding: utf-8 -*-
# Description: varnish netdata python.d module
# Author: ilyam8
# SPDX-License-Identifier: GPL-3.0-or-later
import re
from bases.FrameworkServices.ExecutableService import ExecutableService
from bases.collection import find_binary
ORDER = [
'session_connections',
'client_requests',
'all_time_hit_rate',
'current_poll_hit_rate',
'cached_objects_expired',
'cached_objects_nuked',
'threads_total',
'threads_statistics',
'threads_queue_len',
'backend_connections',
'backend_requests',
'esi_statistics',
'memory_usage',
'uptime'
]
CHARTS = {
'session_connections': {
'options': [None, 'Connections Statistics', 'connections/s',
'client metrics', 'varnish.session_connection', 'line'],
'lines': [
['sess_conn', 'accepted', 'incremental'],
['sess_dropped', 'dropped', 'incremental']
]
},
'client_requests': {
'options': [None, 'Client Requests', 'requests/s',
'client metrics', 'varnish.client_requests', 'line'],
'lines': [
['client_req', 'received', 'incremental']
]
},
'all_time_hit_rate': {
'options': [None, 'All History Hit Rate Ratio', 'percentage', 'cache performance',
'varnish.all_time_hit_rate', 'stacked'],
'lines': [
['cache_hit', 'hit', 'percentage-of-absolute-row'],
['cache_miss', 'miss', 'percentage-of-absolute-row'],
['cache_hitpass', 'hitpass', 'percentage-of-absolute-row']]
},
'current_poll_hit_rate': {
'options': [None, 'Current Poll Hit Rate Ratio', 'percentage', 'cache performance',
'varnish.current_poll_hit_rate', 'stacked'],
'lines': [
['cache_hit', 'hit', 'percentage-of-incremental-row'],
['cache_miss', 'miss', 'percentage-of-incremental-row'],
['cache_hitpass', 'hitpass', 'percentage-of-incremental-row']
]
},
'cached_objects_expired': {
'options': [None, 'Expired Objects', 'expired/s', 'cache performance',
'varnish.cached_objects_expired', 'line'],
'lines': [
['n_expired', 'objects', 'incremental']
]
},
'cached_objects_nuked': {
'options': [None, 'Least Recently Used Nuked Objects', 'nuked/s', 'cache performance',
'varnish.cached_objects_nuked', 'line'],
'lines': [
['n_lru_nuked', 'objects', 'incremental']
]
},
'threads_total': {
'options': [None, 'Number Of Threads In All Pools', 'number', 'thread related metrics',
'varnish.threads_total', 'line'],
'lines': [
['threads', None, 'absolute']
]
},
'threads_statistics': {
'options': [None, 'Threads Statistics', 'threads/s', 'thread related metrics',
'varnish.threads_statistics', 'line'],
'lines': [
['threads_created', 'created', 'incremental'],
['threads_failed', 'failed', 'incremental'],
['threads_limited', 'limited', 'incremental']
]
},
'threads_queue_len': {
'options': [None, 'Current Queue Length', 'requests', 'thread related metrics',
'varnish.threads_queue_len', 'line'],
'lines': [
['thread_queue_len', 'in queue']
]
},
'backend_connections': {
'options': [None, 'Backend Connections Statistics', 'connections/s', 'backend metrics',
'varnish.backend_connections', 'line'],
'lines': [
['backend_conn', 'successful', 'incremental'],
['backend_unhealthy', 'unhealthy', 'incremental'],
['backend_reuse', 'reused', 'incremental'],
['backend_toolate', 'closed', 'incremental'],
['backend_recycle', 'resycled', 'incremental'],
['backend_fail', 'failed', 'incremental']
]
},
'backend_requests': {
'options': [None, 'Requests To The Backend', 'requests/s', 'backend metrics',
'varnish.backend_requests', 'line'],
'lines': [
['backend_req', 'sent', 'incremental']
]
},
'esi_statistics': {
'options': [None, 'ESI Statistics', 'problems/s', 'esi related metrics', 'varnish.esi_statistics', 'line'],
'lines': [
['esi_errors', 'errors', 'incremental'],
['esi_warnings', 'warnings', 'incremental']
]
},
'memory_usage': {
'options': [None, 'Memory Usage', 'MiB', 'memory usage', 'varnish.memory_usage', 'stacked'],
'lines': [
['memory_free', 'free', 'absolute', 1, 1 << 20],
['memory_allocated', 'allocated', 'absolute', 1, 1 << 20]]
},
'uptime': {
'lines': [
['uptime', None, 'absolute']
],
'options': [None, 'Uptime', 'seconds', 'uptime', 'varnish.uptime', 'line']
}
}
VARNISHSTAT = 'varnishstat'
re_version = re.compile(r'varnish-(?P<major>\d+)\.(?P<minor>\d+)\.(?P<patch>\d+)')
class VarnishVersion:
def __init__(self, major, minor, patch):
self.major = major
self.minor = minor
self.patch = patch
def __str__(self):
return '{0}.{1}.{2}'.format(self.major, self.minor, self.patch)
class Parser:
_backend_new = re.compile(r'VBE.([\d\w_.]+)\(.*?\).(beresp[\w_]+)\s+(\d+)')
_backend_old = re.compile(r'VBE\.[\d\w-]+\.([\w\d_]+).(beresp[\w_]+)\s+(\d+)')
_default = re.compile(r'([A-Z]+\.)?([\d\w_.]+)\s+(\d+)')
def __init__(self):
self.re_default = None
self.re_backend = None
def init(self, data):
data = ''.join(data)
parsed_main = Parser._default.findall(data)
if parsed_main:
self.re_default = Parser._default
parsed_backend = Parser._backend_new.findall(data)
if parsed_backend:
self.re_backend = Parser._backend_new
else:
parsed_backend = Parser._backend_old.findall(data)
if parsed_backend:
self.re_backend = Parser._backend_old
def server_stats(self, data):
return self.re_default.findall(''.join(data))
def backend_stats(self, data):
return self.re_backend.findall(''.join(data))
class Service(ExecutableService):
def __init__(self, configuration=None, name=None):
ExecutableService.__init__(self, configuration=configuration, name=name)
self.order = ORDER
self.definitions = CHARTS
self.instance_name = configuration.get('instance_name')
self.parser = Parser()
self.command = None
def create_command(self):
varnishstat = find_binary(VARNISHSTAT)
if not varnishstat:
self.error("can't locate '{0}' binary or binary is not executable by user netdata".format(VARNISHSTAT))
return False
command = [varnishstat, '-V']
reply = self._get_raw_data(stderr=True, command=command)
if not reply:
self.error(
"no output from '{0}'. Is varnish running? Not enough privileges?".format(' '.join(self.command)))
return False
ver = parse_varnish_version(reply)
if not ver:
self.error("failed to parse reply from '{0}', used regex :'{1}', reply : {2}".format(
' '.join(command),
re_version.pattern,
reply,
))
return False
if self.instance_name:
self.command = [varnishstat, '-1', '-n', self.instance_name]
else:
self.command = [varnishstat, '-1']
if ver.major > 4:
self.command.extend(['-t', '1'])
self.info("varnish version: {0}, will use command: '{1}'".format(ver, ' '.join(self.command)))
return True
def check(self):
if not self.create_command():
return False
# STDOUT is not empty
reply = self._get_raw_data()
if not reply:
self.error("no output from '{0}'. Is it running? Not enough privileges?".format(' '.join(self.command)))
return False
self.parser.init(reply)
# Output is parsable
if not self.parser.re_default:
self.error('cant parse the output...')
return False
if self.parser.re_backend:
backends = [b[0] for b in self.parser.backend_stats(reply)[::2]]
self.create_backends_charts(backends)
return True
def get_data(self):
"""
Format data received from shell command
:return: dict
"""
raw = self._get_raw_data()
if not raw:
return None
data = dict()
server_stats = self.parser.server_stats(raw)
if not server_stats:
return None
if self.parser.re_backend:
backend_stats = self.parser.backend_stats(raw)
data.update(dict(('_'.join([name, param]), value) for name, param, value in backend_stats))
data.update(dict((param, value) for _, param, value in server_stats))
# varnish 5 uses default.g_bytes and default.g_space
data['memory_allocated'] = data.get('s0.g_bytes') or data.get('default.g_bytes')
data['memory_free'] = data.get('s0.g_space') or data.get('default.g_space')
return data
def create_backends_charts(self, backends):
for backend in backends:
chart_name = ''.join([backend, '_response_statistics'])
title = 'Backend "{0}"'.format(backend.capitalize())
hdr_bytes = ''.join([backend, '_beresp_hdrbytes'])
body_bytes = ''.join([backend, '_beresp_bodybytes'])
chart = {
chart_name:
{
'options': [None, title, 'kilobits/s', 'backend response statistics',
'varnish.backend', 'area'],
'lines': [
[hdr_bytes, 'header', 'incremental', 8, 1000],
[body_bytes, 'body', 'incremental', -8, 1000]
]
}
}
self.order.insert(0, chart_name)
self.definitions.update(chart)
def parse_varnish_version(lines):
m = re_version.search(lines[0])
if not m:
return None
m = m.groupdict()
return VarnishVersion(
int(m['major']),
int(m['minor']),
int(m['patch']),
)