1 # -*- coding: utf-8 -*-
2 # Description: elastic search node stats netdata python.d module
5 from base import UrlService
6 from requests import get
7 from socket import gethostbyname
9 from queue import Queue
11 from Queue import Queue
12 from threading import Thread
14 # default module values (can be overridden per job in `config`)
20 # charts order (can be overridden if you want less charts, or different order)
21 ORDER = ['search_perf_total', 'search_perf_time', 'search_latency', 'index_perf_total', 'index_perf_time',
22 'index_latency', 'jvm_mem_heap', 'jvm_gc_count', 'jvm_gc_time', 'host_metrics_file_descriptors',
23 'host_metrics_http', 'host_metrics_transport', 'thread_pool_qr', 'fdata_cache', 'fdata_ev_tr',
24 'cluster_health_status', 'cluster_health_nodes', 'cluster_health_shards', 'cluster_stats_nodes',
25 'cluster_stats_query_cache', 'cluster_stats_docs', 'cluster_stats_store', 'cluster_stats_indices_shards']
28 'search_perf_total': {
29 'options': [None, 'Number of queries, fetches', 'queries', 'Search performance', 'es.search_query', 'stacked'],
31 ['query_total', 'search_total', 'incremental'],
32 ['fetch_total', 'fetch_total', 'incremental'],
33 ['query_current', 'search_current', 'absolute'],
34 ['fetch_current', 'fetch_current', 'absolute']
37 'options': [None, 'Time spent on queries, fetches', 'seconds', 'Search performance', 'es.search_time', 'stacked'],
39 ['query_time_in_millis', 'query', 'incremental', 1, 1000],
40 ['fetch_time_in_millis', 'fetch', 'incremental', 1, 1000]
43 'options': [None, 'Query and fetch latency', 'ms', 'Search performance', 'es.search_latency', 'stacked'],
45 ['query_latency', 'query', 'absolute', 1, 1000],
46 ['fetch_latency', 'fetch', 'absolute', 1, 1000]
49 'options': [None, 'Number of documents indexed, index refreshes, flushes', 'documents/indexes',
50 'Indexing performance', 'es.index_doc', 'stacked'],
52 ['indexing_index_total', 'indexed', 'incremental'],
53 ['refresh_total', 'refreshes', 'incremental'],
54 ['flush_total', 'flushes', 'incremental'],
55 ['indexing_index_current', 'indexed_current', 'absolute'],
58 'options': [None, 'Time spent on indexing, refreshing, flushing', 'seconds', 'Indexing performance',
59 'es.search_time', 'stacked'],
61 ['indexing_index_time_in_millis', 'indexing', 'incremental', 1, 1000],
62 ['refresh_total_time_in_millis', 'refreshing', 'incremental', 1, 1000],
63 ['flush_total_time_in_millis', 'flushing', 'incremental', 1, 1000]
66 'options': [None, 'Indexing and flushing latency', 'ms', 'Indexing performance',
67 'es.index_latency', 'stacked'],
69 ['indexing_latency', 'indexing', 'absolute', 1, 1000],
70 ['flushing_latency', 'flushing', 'absolute', 1, 1000]
73 'options': [None, 'JVM heap currently in use/committed', 'percent/MB', 'Memory usage and gc',
74 'es.jvm_heap', 'area'],
76 ['jvm_heap_percent', 'inuse', 'absolute'],
77 ['jvm_heap_commit', 'commit', 'absolute', -1, 1048576]
80 'options': [None, 'Count of garbage collections', 'counts', 'Memory usage and gc', 'es.gc_count', 'stacked'],
82 ['young_collection_count', 'young', 'incremental'],
83 ['old_collection_count', 'old', 'incremental']
86 'options': [None, 'Time spent on garbage collections', 'ms', 'Memory usage and gc', 'es.gc_time', 'stacked'],
88 ['young_collection_time_in_millis', 'young', 'incremental'],
89 ['old_collection_time_in_millis', 'old', 'incremental']
92 'options': [None, 'Number of queued/rejected threads in thread pool', 'threads', 'Queues and rejections',
95 ['bulk_queue', 'bulk_queue', 'absolute'],
96 ['index_queue', 'index_queue', 'absolute'],
97 ['search_queue', 'search_queue', 'absolute'],
98 ['merge_queue', 'merge_queue', 'absolute'],
99 ['bulk_rejected', 'bulk_rej', 'absolute'],
100 ['index_rejected', 'index_rej', 'absolute'],
101 ['search_rejected', 'search_rej', 'absolute'],
102 ['merge_rejected', 'merge_rej', 'absolute']
105 'options': [None, 'Fielddata cache size', 'MB', 'Fielddata cache', 'es.fdata_cache', 'line'],
107 ['index_fdata_mem', 'mem_size', 'absolute', 1, 1048576]
110 'options': [None, 'Fielddata evictions and circuit breaker tripped count', 'number of events',
111 'Fielddata cache', 'es.fdata_ev_tr', 'line'],
113 ['index_fdata_evic', 'evictions', 'incremental'],
114 ['breakers_fdata_trip', 'tripped', 'incremental']
116 'cluster_health_nodes': {
117 'options': [None, 'Nodes and tasks statistics', 'units', 'Cluster health API',
118 'es.cluster_health', 'stacked'],
120 ['health_number_of_nodes', 'nodes', 'absolute'],
121 ['health_number_of_data_nodes', 'data_nodes', 'absolute'],
122 ['health_number_of_pending_tasks', 'pending_tasks', 'absolute'],
123 ['health_number_of_in_flight_fetch', 'inflight_fetch', 'absolute']
125 'cluster_health_status': {
126 'options': [None, 'Cluster status', 'status', 'Cluster health API',
127 'es.cluster_health_status', 'area'],
129 ['status_green', 'green', 'absolute'],
130 ['status_red', 'red', 'absolute'],
131 ['status_foo1', None, 'absolute'],
132 ['status_foo2', None, 'absolute'],
133 ['status_foo3', None, 'absolute'],
134 ['status_yellow', 'yellow', 'absolute']
136 'cluster_health_shards': {
137 'options': [None, 'Shards statistics', 'shards', 'Cluster health API',
138 'es.cluster_health_sharts', 'stacked'],
140 ['health_active_shards', 'active_shards', 'absolute'],
141 ['health_relocating_shards', 'relocating_shards', 'absolute'],
142 ['health_unassigned_shards', 'unassigned', 'absolute'],
143 ['health_delayed_unassigned_shards', 'delayed_unassigned', 'absolute'],
144 ['health_initializing_shards', 'initializing', 'absolute'],
145 ['health_active_shards_percent_as_number', 'active_percent', 'absolute']
147 'cluster_stats_nodes': {
148 'options': [None, 'Nodes statistics', 'nodes', 'Cluster stats API',
149 'es.cluster_stats_nodes', 'stacked'],
151 ['count_data_only', 'data_only', 'absolute'],
152 ['count_master_data', 'master_data', 'absolute'],
153 ['count_total', 'total', 'absolute'],
154 ['count_master_only', 'master_only', 'absolute'],
155 ['count_client', 'client', 'absolute']
157 'cluster_stats_query_cache': {
158 'options': [None, 'Query cache statistics', 'queries', 'Cluster stats API',
159 'es.cluster_stats_query_cache', 'stacked'],
161 ['query_cache_hit_count', 'hit', 'incremental'],
162 ['query_cache_miss_count', 'miss', 'incremental']
164 'cluster_stats_docs': {
165 'options': [None, 'Docs statistics', 'count', 'Cluster stats API',
166 'es.cluster_stats_docs', 'line'],
168 ['docs_count', 'docs', 'absolute']
170 'cluster_stats_store': {
171 'options': [None, 'Store statistics', 'MB', 'Cluster stats API',
172 'es.cluster_stats_store', 'line'],
174 ['store_size_in_bytes', 'size', 'absolute', 1, 1048567]
176 'cluster_stats_indices_shards': {
177 'options': [None, 'Indices and shards statistics', 'count', 'Cluster stats API',
178 'es.cluster_stats_ind_sha', 'stacked'],
180 ['indices_count', 'indices', 'absolute'],
181 ['shards_total', 'shards', 'absolute']
183 'host_metrics_transport': {
184 'options': [None, 'Cluster communication transport metrics', 'kbit/s', 'Host metrics',
185 'es.host_metrics_transport', 'area'],
187 ['transport_rx_size_in_bytes', 'in', 'incremental', 8, 1000],
188 ['transport_tx_size_in_bytes', 'out', 'incremental', -8, 1000]
190 'host_metrics_file_descriptors': {
191 'options': [None, 'Available file descriptors in percent', 'percent', 'Host metrics',
192 'es.host_metrics_descriptors', 'area'],
194 ['file_descriptors_used', 'used', 'absolute', 1, 10]
196 'host_metrics_http': {
197 'options': [None, 'Opened HTTP connections', 'connections', 'Host metrics',
198 'es.host_metrics_http', 'line'],
200 ['http_current_open', 'opened', 'absolute', 1, 1]
205 class Service(UrlService):
206 def __init__(self, configuration=None, name=None):
207 UrlService.__init__(self, configuration=configuration, name=name)
209 self.definitions = CHARTS
210 self.host = self.configuration.get('host')
211 self.port = self.configuration.get('port')
212 self.user = self.configuration.get('user')
213 self.password = self.configuration.get('pass')
214 self.latency = dict()
217 # We can't start if <host> AND <port> not specified
218 if not all([self.host, self.port]):
221 # It as a bad idea to use hostname.
222 # Hostname -> ipaddress
224 self.host = gethostbyname(self.host)
225 except Exception as e:
229 # HTTP Auth? NOT TESTED
230 self.auth = self.user and self.password
232 # Create URL for every Elasticsearch API
233 url_node_stats = 'http://%s:%s/_nodes/_local/stats' % (self.host, self.port)
234 url_cluster_health = 'http://%s:%s/_cluster/health' % (self.host, self.port)
235 url_cluster_stats = 'http://%s:%s/_cluster/stats' % (self.host, self.port)
237 # Create list of enabled API calls
238 user_choice = [bool(self.configuration.get('node_stats', True)),
239 bool(self.configuration.get('cluster_health', True)),
240 bool(self.configuration.get('cluster_stats', True))]
242 avail_methods = [(self._get_node_stats, url_node_stats),
243 (self._get_cluster_health, url_cluster_health),
244 (self._get_cluster_stats, url_cluster_stats)]
246 # Remove disabled API calls from 'avail methods'
247 self.methods = [avail_methods[_] for _ in range(len(avail_methods)) if user_choice[_]]
249 # Run _get_data for ALL active API calls.
251 for method in self.methods:
252 api_result[method[1]] = (bool(self._get_raw_data(method[1])))
254 # We can start ONLY if all active API calls returned NOT None
255 if not all(api_result.values()):
256 self.error('Plugin could not get data from all APIs')
257 self.error('%s' % api_result)
260 self.info('%s' % api_result)
261 self.info('Plugin was started successfully')
265 def _get_raw_data(self, url):
270 raw_data = get(url, auth=(self.user, self.password))
281 for method in self.methods:
282 th = Thread(target=method[0], args=(queue, method[1]))
286 for thread in threads:
288 result.update(queue.get())
290 return result or None
292 def _get_cluster_health(self, queue, url):
294 Format data received from http request
298 data = self._get_raw_data(url)
306 to_netdata.update(update_key('health', data))
307 to_netdata.update({'status_green': 0, 'status_red': 0, 'status_yellow': 0,
308 'status_foo1': 0, 'status_foo2': 0, 'status_foo3': 0})
309 to_netdata[''.join(['status_', to_netdata.get('health_status', '')])] = 1
311 queue.put(to_netdata)
313 def _get_cluster_stats(self, queue, url):
315 Format data received from http request
319 data = self._get_raw_data(url)
327 to_netdata.update(update_key('count', data['nodes']['count']))
328 to_netdata.update(update_key('query_cache', data['indices']['query_cache']))
329 to_netdata.update(update_key('docs', data['indices']['docs']))
330 to_netdata.update(update_key('store', data['indices']['store']))
331 to_netdata['indices_count'] = data['indices']['count']
332 to_netdata['shards_total'] = data['indices']['shards']['total']
334 queue.put(to_netdata)
336 def _get_node_stats(self, queue, url):
338 Format data received from http request
342 data = self._get_raw_data(url)
348 node = list(data['nodes'].keys())[0]
350 # Search performance metrics
351 to_netdata.update(data['nodes'][node]['indices']['search'])
352 to_netdata['query_latency'] = self.find_avg(to_netdata['query_total'],
353 to_netdata['query_time_in_millis'], 'query_latency')
354 to_netdata['fetch_latency'] = self.find_avg(to_netdata['fetch_total'],
355 to_netdata['fetch_time_in_millis'], 'fetch_latency')
357 # Indexing performance metrics
358 for key in ['indexing', 'refresh', 'flush']:
359 to_netdata.update(update_key(key, data['nodes'][node]['indices'].get(key, {})))
360 to_netdata['indexing_latency'] = self.find_avg(to_netdata['indexing_index_total'],
361 to_netdata['indexing_index_time_in_millis'], 'index_latency')
362 to_netdata['flushing_latency'] = self.find_avg(to_netdata['flush_total'],
363 to_netdata['flush_total_time_in_millis'], 'flush_latency')
364 # Memory usage and garbage collection
365 to_netdata.update(update_key('young', data['nodes'][node]['jvm']['gc']['collectors']['young']))
366 to_netdata.update(update_key('old', data['nodes'][node]['jvm']['gc']['collectors']['old']))
367 to_netdata['jvm_heap_percent'] = data['nodes'][node]['jvm']['mem']['heap_used_percent']
368 to_netdata['jvm_heap_commit'] = data['nodes'][node]['jvm']['mem']['heap_committed_in_bytes']
370 # Thread pool queues and rejections
371 for key in ['bulk', 'index', 'search', 'merge']:
372 to_netdata.update(update_key(key, data['nodes'][node]['thread_pool'].get(key, {})))
375 to_netdata['index_fdata_mem'] = data['nodes'][node]['indices']['fielddata']['memory_size_in_bytes']
376 to_netdata['index_fdata_evic'] = data['nodes'][node]['indices']['fielddata']['evictions']
377 to_netdata['breakers_fdata_trip'] = data['nodes'][node]['breakers']['fielddata']['tripped']
380 to_netdata.update(update_key('http', data['nodes'][node]['http']))
381 to_netdata.update(update_key('transport', data['nodes'][node]['transport']))
382 to_netdata['file_descriptors_used'] = round(float(data['nodes'][node]['process']['open_file_descriptors'])
383 / data['nodes'][node]['process']['max_file_descriptors'] * 1000)
385 queue.put(to_netdata)
387 def find_avg(self, value1, value2, key):
388 if key not in self.latency:
389 self.latency.update({key: [value1, value2]})
392 if not self.latency[key][0] == value1:
393 latency = round(float(value2 - self.latency[key][1]) / float(value1 - self.latency[key][0]) * 1000)
394 self.latency.update({key: [value1, value2]})
397 self.latency.update({key: [value1, value2]})
401 def update_key(string, dictionary):
402 return {'_'.join([string, k]): v for k, v in dictionary.items()}