elasticsearch: bugfix and conf file update
Ilya committed
Jun 16, 2017 at 22:14 UTC
9cb561b8e9af94409ec06daf7116926cb5e47a30
2 files changed
+106
-105
conf.d/python.d/elasticsearch.conf
+7
-10
@@ -61,19 +61,16 @@
61
# cluster_health: False/True # Calls to cluster health elasticsearch API. Enabled by default.
62
# cluster_stats: False/True # Calls to cluster stats elasticsearch API. Enabled by default.
63
#
64
-# ----------------------------------------------------------------------
65
-# IMPORTANT Information
66
-#
67
-# Module uses python `requests` package
64
#
69
-# You need to install it manually. (python-requests or python3-requests depending on the version of python).
65
+# if the URL is password protected, the following are supported:
66
#
67
+# user: 'username'
68
+# pass: 'password'
69
#
70
+# ----------------------------------------------------------------------
71
# AUTO-DETECTION JOBS
72
# only one of them will run (they have the same name)
73
#
75
-#local:
76
-# host: '127.0.0.1'
77
-# port: '9200'
78
-# cluster_health: True
79
-# cluster_stats: True
74
+local:
75
+ host: '127.0.0.1'
76
+ port: '9200'
python.d/elasticsearch.chart.py
+99
-95
@@ -85,6 +85,21 @@ HEALTH_STATS = [
85
('active_shards_percent_as_number', 'health_active_shards_percent_as_number', None)
86
]
87
88
+LATENCY = {
89
+ 'query_latency':
90
+ {'total': 'query_total',
91
+ 'spent_time': 'query_time_in_millis'},
92
+ 'fetch_latency':
93
+ {'total': 'fetch_total',
94
+ 'spent_time': 'fetch_time_in_millis'},
95
+ 'indexing_latency':
96
+ {'total': 'indexing_index_total',
97
+ 'spent_time': 'indexing_index_time_in_millis'},
98
+ 'flushing_latency':
99
+ {'total': 'flush_total',
100
+ 'spent_time': 'flush_total_time_in_millis'}
101
+}
102
+
103
# charts order (can be overridden if you want less charts, or different order)
104
ORDER = ['search_perf_total', 'search_perf_current', 'search_perf_time', 'search_latency', 'index_perf_total',
105
'index_perf_current', 'index_perf_time', 'index_latency', 'jvm_mem_heap', 'jvm_gc_count',
@@ -95,34 +110,34 @@ ORDER = ['search_perf_total', 'search_perf_current', 'search_perf_time', 'search
110
111
CHARTS = {
112
'search_perf_total': {
98
- 'options': [None, 'Total number of queries, fetches', 'number of', 'search performance',
113
+ 'options': [None, 'Queries And Fetches', 'number of', 'search performance',
114
'es.search_query_total', 'stacked'],
115
'lines': [
116
['query_total', 'queries', 'incremental'],
117
['fetch_total', 'fetches', 'incremental']
118
]},
119
'search_perf_current': {
105
- 'options': [None, 'Number of queries, fetches in progress', 'number of', 'search performance',
120
+ 'options': [None, 'Queries and Fetches In Progress', 'number of', 'search performance',
121
'es.search_query_current', 'stacked'],
122
'lines': [
123
['query_current', 'queries', 'absolute'],
124
['fetch_current', 'fetches', 'absolute']
125
]},
126
'search_perf_time': {
112
- 'options': [None, 'Time spent on queries, fetches', 'seconds', 'search performance',
127
+ 'options': [None, 'Time Spent On Queries And Fetches', 'seconds', 'search performance',
128
'es.search_time', 'stacked'],
129
'lines': [
130
['query_time_in_millis', 'query', 'incremental', 1, 1000],
131
['fetch_time_in_millis', 'fetch', 'incremental', 1, 1000]
132
]},
133
'search_latency': {
119
- 'options': [None, 'Query and fetch latency', 'ms', 'search performance', 'es.search_latency', 'stacked'],
134
+ 'options': [None, 'Query And Fetch Latency', 'ms', 'search performance', 'es.search_latency', 'stacked'],
135
'lines': [
136
['query_latency', 'query', 'absolute', 1, 1000],
137
['fetch_latency', 'fetch', 'absolute', 1, 1000]
138
]},
139
'index_perf_total': {
125
- 'options': [None, 'Total number of documents indexed, index refreshes, index flushes to disk', 'number of',
140
+ 'options': [None, 'Indexed Documents, Index Refreshes, Index Flushes To Disk', 'number of',
141
'indexing performance', 'es.index_performance_total', 'stacked'],
142
'lines': [
143
['indexing_index_total', 'indexed', 'incremental'],
@@ -130,13 +145,13 @@ CHARTS = {
145
['flush_total', 'flushes', 'incremental']
146
]},
147
'index_perf_current': {
133
- 'options': [None, 'Number of documents currently being indexed', 'currently indexed',
148
+ 'options': [None, 'Number Of Documents Currently Being Indexed', 'currently indexed',
149
'indexing performance', 'es.index_performance_current', 'stacked'],
150
'lines': [
151
['indexing_index_current', 'documents', 'absolute']
152
]},
153
'index_perf_time': {
139
- 'options': [None, 'Time spent on indexing, refreshing, flushing', 'seconds', 'indexing performance',
154
+ 'options': [None, 'Time Spent On Indexing, Refreshing, Flushing', 'seconds', 'indexing performance',
155
'es.search_time', 'stacked'],
156
'lines': [
157
['indexing_index_time_in_millis', 'indexing', 'incremental', 1, 1000],
@@ -144,33 +159,33 @@ CHARTS = {
159
['flush_total_time_in_millis', 'flushing', 'incremental', 1, 1000]
160
]},
161
'index_latency': {
147
- 'options': [None, 'Indexing and flushing latency', 'ms', 'indexing performance',
162
+ 'options': [None, 'Indexing And Flushing Latency', 'ms', 'indexing performance',
163
'es.index_latency', 'stacked'],
164
'lines': [
165
['indexing_latency', 'indexing', 'absolute', 1, 1000],
166
['flushing_latency', 'flushing', 'absolute', 1, 1000]
167
]},
168
'jvm_mem_heap': {
154
- 'options': [None, 'JVM heap currently in use/committed', 'percent/MB', 'memory usage and gc',
169
+ 'options': [None, 'JVM Heap Currently in Use/Committed', 'percent/MB', 'memory usage and gc',
170
'es.jvm_heap', 'area'],
171
'lines': [
172
['jvm_heap_percent', 'inuse', 'absolute'],
173
['jvm_heap_commit', 'commit', 'absolute', -1, 1048576]
174
]},
175
'jvm_gc_count': {
161
- 'options': [None, 'Count of garbage collections', 'counts', 'memory usage and gc', 'es.gc_count', 'stacked'],
176
+ 'options': [None, 'Garbage Collections', 'counts', 'memory usage and gc', 'es.gc_count', 'stacked'],
177
'lines': [
178
['young_collection_count', 'young', 'incremental'],
179
['old_collection_count', 'old', 'incremental']
180
]},
181
'jvm_gc_time': {
167
- 'options': [None, 'Time spent on garbage collections', 'ms', 'memory usage and gc', 'es.gc_time', 'stacked'],
182
+ 'options': [None, 'Time Spent On Garbage Collections', 'ms', 'memory usage and gc', 'es.gc_time', 'stacked'],
183
'lines': [
184
['young_collection_time_in_millis', 'young', 'incremental'],
185
['old_collection_time_in_millis', 'old', 'incremental']
186
]},
187
'thread_pool_qr_q': {
173
- 'options': [None, 'Number of queued threads in thread pool', 'queued threads', 'queues and rejections',
188
+ 'options': [None, 'Number Of Queued Threads In Thread Pool', 'queued threads', 'queues and rejections',
189
'es.thread_pool_queued', 'stacked'],
190
'lines': [
191
['bulk_queue', 'bulk', 'absolute'],
@@ -179,7 +194,7 @@ CHARTS = {
194
['merge_queue', 'merge', 'absolute']
195
]},
196
'thread_pool_qr_r': {
182
- 'options': [None, 'Number of rejected threads in thread pool', 'rejected threads', 'queues and rejections',
197
+ 'options': [None, 'Rejected Threads In Thread Pool', 'rejected threads', 'queues and rejections',
198
'es.thread_pool_rejected', 'stacked'],
199
'lines': [
200
['bulk_rejected', 'bulk', 'absolute'],
@@ -188,19 +203,19 @@ CHARTS = {
203
['merge_rejected', 'merge', 'absolute']
204
]},
205
'fdata_cache': {
191
- 'options': [None, 'Fielddata cache size', 'MB', 'fielddata cache', 'es.fdata_cache', 'line'],
206
+ 'options': [None, 'Fielddata Cache', 'MB', 'fielddata cache', 'es.fdata_cache', 'line'],
207
'lines': [
208
['index_fdata_memory', 'cache', 'absolute', 1, 1048576]
209
]},
210
'fdata_ev_tr': {
196
- 'options': [None, 'Fielddata evictions and circuit breaker tripped count', 'number of events',
211
+ 'options': [None, 'Fielddata Evictions And Circuit Breaker Tripped Count', 'number of events',
212
'fielddata cache', 'es.evictions_tripped', 'line'],
213
'lines': [
214
['evictions', None, 'incremental'],
215
['tripped', None, 'incremental']
216
]},
217
'cluster_health_nodes': {
203
- 'options': [None, 'Nodes and tasks statistics', 'units', 'cluster health API',
218
+ 'options': [None, 'Nodes And Tasks Statistics', 'units', 'cluster health API',
219
'es.cluster_health_nodes', 'stacked'],
220
'lines': [
221
['health_number_of_nodes', 'nodes', 'absolute'],
@@ -209,7 +224,7 @@ CHARTS = {
224
['health_number_of_in_flight_fetch', 'in_flight_fetch', 'absolute']
225
]},
226
'cluster_health_status': {
212
- 'options': [None, 'Cluster status', 'status', 'cluster health API',
227
+ 'options': [None, 'Cluster Status', 'status', 'cluster health API',
228
'es.cluster_health_status', 'area'],
229
'lines': [
230
['status_green', 'green', 'absolute'],
@@ -220,7 +235,7 @@ CHARTS = {
235
['status_yellow', 'yellow', 'absolute']
236
]},
237
'cluster_health_shards': {
223
- 'options': [None, 'Shards statistics', 'shards', 'cluster health API',
238
+ 'options': [None, 'Shards Statistics', 'shards', 'cluster health API',
239
'es.cluster_health_shards', 'stacked'],
240
'lines': [
241
['health_active_shards', 'active_shards', 'absolute'],
@@ -231,7 +246,7 @@ CHARTS = {
246
['health_active_shards_percent_as_number', 'active_percent', 'absolute']
247
]},
248
'cluster_stats_nodes': {
234
- 'options': [None, 'Nodes statistics', 'nodes', 'cluster stats API',
249
+ 'options': [None, 'Nodes Statistics', 'nodes', 'cluster stats API',
250
'es.cluster_nodes', 'stacked'],
251
'lines': [
252
['count_data_only', 'data_only', 'absolute'],
@@ -241,46 +256,46 @@ CHARTS = {
256
['count_client', 'client', 'absolute']
257
]},
258
'cluster_stats_query_cache': {
244
- 'options': [None, 'Query cache statistics', 'queries', 'cluster stats API',
259
+ 'options': [None, 'Query Cache Statistics', 'queries', 'cluster stats API',
260
'es.cluster_query_cache', 'stacked'],
261
'lines': [
262
['query_cache_hit_count', 'hit', 'incremental'],
263
['query_cache_miss_count', 'miss', 'incremental']
264
]},
265
'cluster_stats_docs': {
251
- 'options': [None, 'Docs statistics', 'count', 'cluster stats API',
266
+ 'options': [None, 'Docs Statistics', 'count', 'cluster stats API',
267
'es.cluster_docs', 'line'],
268
'lines': [
269
['docs_count', 'docs', 'absolute']
270
]},
271
'cluster_stats_store': {
257
- 'options': [None, 'Store statistics', 'MB', 'cluster stats API',
272
+ 'options': [None, 'Store Statistics', 'MB', 'cluster stats API',
273
'es.cluster_store', 'line'],
274
'lines': [
275
['store_size_in_bytes', 'size', 'absolute', 1, 1048567]
276
]},
277
'cluster_stats_indices_shards': {
263
- 'options': [None, 'Indices and shards statistics', 'count', 'cluster stats API',
278
+ 'options': [None, 'Indices And Shards Statistics', 'count', 'cluster stats API',
279
'es.cluster_indices_shards', 'stacked'],
280
'lines': [
281
['indices_count', 'indices', 'absolute'],
282
['shards_total', 'shards', 'absolute']
283
]},
284
'host_metrics_transport': {
270
- 'options': [None, 'Cluster communication transport metrics', 'kbit/s', 'host metrics',
285
+ 'options': [None, 'Cluster Communication Transport Metrics', 'kilobit/s', 'host metrics',
286
'es.host_transport', 'area'],
287
'lines': [
288
['transport_rx_size_in_bytes', 'in', 'incremental', 8, 1000],
289
['transport_tx_size_in_bytes', 'out', 'incremental', -8, 1000]
290
]},
291
'host_metrics_file_descriptors': {
277
- 'options': [None, 'Available file descriptors in percent', 'percent', 'host metrics',
292
+ 'options': [None, 'Available File Descriptors In Percent', 'percent', 'host metrics',
293
'es.host_descriptors', 'area'],
294
'lines': [
295
['file_descriptors_used', 'used', 'absolute', 1, 10]
296
]},
297
'host_metrics_http': {
283
- 'options': [None, 'Opened HTTP connections', 'connections', 'host metrics',
298
+ 'options': [None, 'Opened HTTP Connections', 'connections', 'host metrics',
299
'es.host_http_connections', 'line'],
300
'lines': [
301
['http_current_open', 'opened', 'absolute', 1, 1]
@@ -300,12 +315,13 @@ class Service(UrlService):
315
self.methods = list()
316
317
def check(self):
303
- # We can't start if <host> AND <port> not specified
304
- if not all([self.host, self.port, isinstance(self.host, str), isinstance(self.port, (str, int))]):
318
+ if not all([self.host,
319
+ self.port,
320
+ isinstance(self.host, str),
321
+ isinstance(self.port, (str, int))]):
322
self.error('Host is not defined in the module configuration file')
323
return False
324
308
- # It as a bad idea to use hostname.
325
# Hostname -> ip address
326
try:
327
self.host = gethostbyname(self.host)
@@ -322,36 +338,24 @@ class Service(UrlService):
338
url_cluster_health = '%s://%s:%s/_cluster/health' % (scheme, self.host, self.port)
339
url_cluster_stats = '%s://%s:%s/_cluster/stats' % (scheme, self.host, self.port)
340
325
- # Create list of enabled API calls
341
user_choice = [bool(self.configuration.get('node_stats', True)),
342
bool(self.configuration.get('cluster_health', True)),
343
bool(self.configuration.get('cluster_stats', True))]
344
330
- avail_methods = [METHODS(get_data_function=self._get_node_stats_, url=url_node_stats),
331
- METHODS(get_data_function=self._get_cluster_health_, url=url_cluster_health),
332
- METHODS(get_data_function=self._get_cluster_stats_, url=url_cluster_stats)]
345
+ avail_methods = [METHODS(get_data_function=self._get_node_stats_,
346
+ url=url_node_stats),
347
+ METHODS(get_data_function=self._get_cluster_health_,
348
+ url=url_cluster_health),
349
+ METHODS(get_data_function=self._get_cluster_stats_,
350
+ url=url_cluster_stats)]
351
352
# Remove disabled API calls from 'avail methods'
353
self.methods = [avail_methods[e[0]] for e in enumerate(avail_methods) if user_choice[e[0]]]
336
-
337
- # Run _get_data for ALL active API calls.
338
- api_check_result = dict()
339
- data_from_check = dict()
340
- for method in self.methods:
341
- try:
342
- api_check_result[method.url] = method.get_data_function(None, method.url)
343
- data_from_check.update(api_check_result[method.url] or dict())
344
- except KeyError as error:
345
- self.error('Failed to parse %s. Error: %s' % (method.url, str(error)))
346
- return False
347
-
348
- # We can start ONLY if all active API calls returned NOT None
349
- if not all(api_check_result.values()):
350
- self.error('Plugin could not get data from all APIs')
354
+ data = self._get_data()
355
+ if not data:
356
return False
352
- else:
353
- self._data_from_check = data_from_check
354
- return True
357
+ self._data_from_check = data
358
+ return True
359
360
def _get_data(self):
361
threads = list()
@@ -359,7 +363,8 @@ class Service(UrlService):
363
result = dict()
364
365
for method in self.methods:
362
- th = Thread(target=method.get_data_function, args=(queue, method.url))
366
+ th = Thread(target=method.get_data_function,
367
+ args=(queue, method.url))
368
th.start()
369
threads.append(th)
370
@@ -378,18 +383,18 @@ class Service(UrlService):
383
raw_data = self._get_raw_data(url)
384
385
if not raw_data:
381
- return queue.put(dict()) if queue else None
382
- else:
383
- data = loads(raw_data)
386
+ return queue.put(dict())
387
385
- to_netdata = fetch_data_(raw_data=data, metrics_list=HEALTH_STATS)
388
+ data = loads(raw_data)
389
+ to_netdata = fetch_data_(raw_data=data,
390
+ metrics_list=HEALTH_STATS)
391
387
- to_netdata.update({'status_green': 0, 'status_red': 0, 'status_yellow': 0,
388
- 'status_foo1': 0, 'status_foo2': 0, 'status_foo3': 0})
389
- current_status = 'status_' + data['status']
390
- to_netdata[current_status] = 1
392
+ to_netdata.update({'status_green': 0, 'status_red': 0, 'status_yellow': 0,
393
+ 'status_foo1': 0, 'status_foo2': 0, 'status_foo3': 0})
394
+ current_status = 'status_' + data['status']
395
+ to_netdata[current_status] = 1
396
392
- return queue.put(to_netdata) if queue else to_netdata
397
+ return queue.put(to_netdata)
398
399
def _get_cluster_stats_(self, queue, url):
400
"""
@@ -400,13 +405,13 @@ class Service(UrlService):
405
raw_data = self._get_raw_data(url)
406
407
if not raw_data:
403
- return queue.put(dict()) if queue else None
404
- else:
405
- data = loads(raw_data)
408
+ return queue.put(dict())
409
407
- to_netdata = fetch_data_(raw_data=data, metrics_list=CLUSTER_STATS)
410
+ data = loads(raw_data)
411
+ to_netdata = fetch_data_(raw_data=data,
412
+ metrics_list=CLUSTER_STATS)
413
409
- return queue.put(to_netdata) if queue else to_netdata
414
+ return queue.put(to_netdata)
415
416
def _get_node_stats_(self, queue, url):
417
"""
@@ -417,47 +422,46 @@ class Service(UrlService):
422
raw_data = self._get_raw_data(url)
423
424
if not raw_data:
420
- return queue.put(dict()) if queue else None
421
- else:
422
- data = loads(raw_data)
423
-
424
- node = list(data['nodes'].keys())[0]
425
- to_netdata = fetch_data_(raw_data=data['nodes'][node], metrics_list=NODE_STATS)
425
+ return queue.put(dict())
426
427
- # Search performance latency
428
- to_netdata['query_latency'] = self.find_avg_(to_netdata['query_total'],
429
- to_netdata['query_time_in_millis'], 'query_latency')
430
- to_netdata['fetch_latency'] = self.find_avg_(to_netdata['fetch_total'],
431
- to_netdata['fetch_time_in_millis'], 'fetch_latency')
427
+ data = loads(raw_data)
428
433
- # Indexing performance latency
434
- to_netdata['indexing_latency'] = self.find_avg_(to_netdata['indexing_index_total'],
435
- to_netdata['indexing_index_time_in_millis'], 'index_latency')
436
- to_netdata['flushing_latency'] = self.find_avg_(to_netdata['flush_total'],
437
- to_netdata['flush_total_time_in_millis'], 'flush_latency')
429
+ node = list(data['nodes'].keys())[0]
430
+ to_netdata = fetch_data_(raw_data=data['nodes'][node],
431
+ metrics_list=NODE_STATS)
432
433
+ # Search, index, flush, fetch performance latency
434
+ for key in LATENCY:
435
+ try:
436
+ to_netdata[key] = self.find_avg_(total=to_netdata[LATENCY[key]['total']],
437
+ spent_time=to_netdata[LATENCY[key]['spent_time']],
438
+ key=key)
439
+ except KeyError:
440
+ continue
441
+ if 'open_file_descriptors' in to_netdata:
442
to_netdata['file_descriptors_used'] = round(float(to_netdata['open_file_descriptors'])
443
/ to_netdata['max_file_descriptors'] * 1000)
444
442
- return queue.put(to_netdata) if queue else to_netdata
445
+ return queue.put(to_netdata)
446
444
- def find_avg_(self, value1, value2, key):
447
+ def find_avg_(self, total, spent_time, key):
448
if key not in self.latency:
446
- self.latency.update({key: [value1, value2]})
449
+ self.latency[key] = dict(total=total,
450
+ spent_time=spent_time)
451
return 0
448
- else:
449
- if not self.latency[key][0] == value1:
450
- latency = round(float(value2 - self.latency[key][1]) / float(value1 - self.latency[key][0]) * 1000)
451
- self.latency.update({key: [value1, value2]})
452
- return latency
453
- else:
454
- self.latency.update({key: [value1, value2]})
455
- return 0
452
+ if self.latency[key]['total'] != total:
453
+ latency = float(spent_time - self.latency[key]['spent_time'])\
454
+ / float(total - self.latency[key]['total']) * 1000
455
+ self.latency[key]['total'] = total
456
+ self.latency[key]['spent_time'] = spent_time
457
+ return latency
458
+ self.latency[key]['spent_time'] = spent_time
459
+ return 0
460
461
462
def fetch_data_(raw_data, metrics_list):
463
to_netdata = dict()
460
- for metric, new_name, function in metrics_list:
464
+ for metric, new_name, func in metrics_list:
465
value = raw_data
466
for key in metric.split('.'):
467
try:
@@ -465,7 +469,7 @@ def fetch_data_(raw_data, metrics_list):
469
except KeyError:
470
break
471
if not isinstance(value, dict) and key:
468
- to_netdata[new_name or key] = value if not function else function(value)
472
+ to_netdata[new_name or key] = value if not func else func(value)
473
474
return to_netdata
475