Show More
@@ -1,662 +1,663 b'' | |||
|
1 | 1 | # -*- coding: utf-8 -*- |
|
2 | 2 | |
|
3 | 3 | # Copyright (C) 2010-2016 RhodeCode GmbH |
|
4 | 4 | # |
|
5 | 5 | # This program is free software: you can redistribute it and/or modify |
|
6 | 6 | # it under the terms of the GNU Affero General Public License, version 3 |
|
7 | 7 | # (only), as published by the Free Software Foundation. |
|
8 | 8 | # |
|
9 | 9 | # This program is distributed in the hope that it will be useful, |
|
10 | 10 | # but WITHOUT ANY WARRANTY; without even the implied warranty of |
|
11 | 11 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
|
12 | 12 | # GNU General Public License for more details. |
|
13 | 13 | # |
|
14 | 14 | # You should have received a copy of the GNU Affero General Public License |
|
15 | 15 | # along with this program. If not, see <http://www.gnu.org/licenses/>. |
|
16 | 16 | # |
|
17 | 17 | # This program is dual-licensed. If you wish to learn more about the |
|
18 | 18 | # AppEnlight Enterprise Edition, including its added features, Support |
|
19 | 19 | # services, and proprietary license terms, please see |
|
20 | 20 | # https://rhodecode.com/licenses/ |
|
21 | 21 | |
|
22 | 22 | import bisect |
|
23 | 23 | import collections |
|
24 | 24 | import math |
|
25 | 25 | from datetime import datetime, timedelta |
|
26 | 26 | |
|
27 | 27 | import sqlalchemy as sa |
|
28 | 28 | import pyelasticsearch |
|
29 | 29 | |
|
30 | 30 | from celery.utils.log import get_task_logger |
|
31 | 31 | from zope.sqlalchemy import mark_changed |
|
32 | 32 | from pyramid.threadlocal import get_current_request, get_current_registry |
|
33 | 33 | from appenlight.celery import celery |
|
34 | 34 | from appenlight.models.report_group import ReportGroup |
|
35 | 35 | from appenlight.models import DBSession, Datastores |
|
36 | 36 | from appenlight.models.report import Report |
|
37 | 37 | from appenlight.models.log import Log |
|
38 | 38 | from appenlight.models.request_metric import Metric |
|
39 | 39 | from appenlight.models.event import Event |
|
40 | 40 | |
|
41 | 41 | from appenlight.models.services.application import ApplicationService |
|
42 | 42 | from appenlight.models.services.event import EventService |
|
43 | 43 | from appenlight.models.services.log import LogService |
|
44 | 44 | from appenlight.models.services.report import ReportService |
|
45 | 45 | from appenlight.models.services.report_group import ReportGroupService |
|
46 | 46 | from appenlight.models.services.user import UserService |
|
47 | 47 | from appenlight.models.tag import Tag |
|
48 | 48 | from appenlight.lib import print_traceback |
|
49 | 49 | from appenlight.lib.utils import parse_proto, in_batches |
|
50 | 50 | from appenlight.lib.ext_json import json |
|
51 | 51 | from appenlight.lib.redis_keys import REDIS_KEYS |
|
52 | 52 | from appenlight.lib.enums import ReportType |
|
53 | 53 | |
|
54 | 54 | log = get_task_logger(__name__) |
|
55 | 55 | |
|
56 |
sample_boundries = list(range(100, 1000 |
|
|
56 | sample_boundries = list(range(100, 1000, 100)) + \ | |
|
57 | list(range(1000, 10000, 1000)) + \ | |
|
58 | list(range(10000, 100000, 5000)) | |
|
57 | 59 | |
|
58 | 60 | |
|
59 |
def pick_sample(total_occurences, report_type= |
|
|
61 | def pick_sample(total_occurences, report_type=None): | |
|
60 | 62 | every = 1.0 |
|
61 | 63 | position = bisect.bisect_left(sample_boundries, total_occurences) |
|
62 | 64 | if position > 0: |
|
63 | # 404 | |
|
64 | if report_type == 2: | |
|
65 | if report_type == ReportType.not_found: | |
|
65 | 66 | divide = 10.0 |
|
66 | 67 | else: |
|
67 | 68 | divide = 100.0 |
|
68 | 69 | every = sample_boundries[position - 1] / divide |
|
69 | 70 | return total_occurences % every == 0 |
|
70 | 71 | |
|
71 | 72 | |
|
72 | 73 | @celery.task(queue="default", default_retry_delay=1, max_retries=2) |
|
73 | 74 | def test_exception_task(): |
|
74 | 75 | log.error('test celery log', extra={'location': 'celery'}) |
|
75 | 76 | log.warning('test celery log', extra={'location': 'celery'}) |
|
76 | 77 | raise Exception('Celery exception test') |
|
77 | 78 | |
|
78 | 79 | |
|
79 | 80 | @celery.task(queue="default", default_retry_delay=1, max_retries=2) |
|
80 | 81 | def test_retry_exception_task(): |
|
81 | 82 | try: |
|
82 | 83 | import time |
|
83 | 84 | |
|
84 | 85 | time.sleep(1.3) |
|
85 | 86 | log.error('test retry celery log', extra={'location': 'celery'}) |
|
86 | 87 | log.warning('test retry celery log', extra={'location': 'celery'}) |
|
87 | 88 | raise Exception('Celery exception test') |
|
88 | 89 | except Exception as exc: |
|
89 | 90 | test_retry_exception_task.retry(exc=exc) |
|
90 | 91 | |
|
91 | 92 | |
|
92 | 93 | @celery.task(queue="reports", default_retry_delay=600, max_retries=999) |
|
93 | 94 | def add_reports(resource_id, params, dataset, environ=None, **kwargs): |
|
94 | 95 | proto_version = parse_proto(params.get('protocol_version', '')) |
|
95 | 96 | current_time = datetime.utcnow().replace(second=0, microsecond=0) |
|
96 | 97 | try: |
|
97 | 98 | # we will store solr docs here for single insert |
|
98 | 99 | es_report_docs = {} |
|
99 | 100 | es_report_group_docs = {} |
|
100 | 101 | resource = ApplicationService.by_id(resource_id) |
|
101 | 102 | reports = [] |
|
102 | 103 | |
|
103 | 104 | if proto_version.major < 1 and proto_version.minor < 5: |
|
104 | 105 | for report_data in dataset: |
|
105 | 106 | report_details = report_data.get('report_details', []) |
|
106 | 107 | for i, detail_data in enumerate(report_details): |
|
107 | 108 | report_data.update(detail_data) |
|
108 | 109 | report_data.pop('report_details') |
|
109 | 110 | traceback = report_data.get('traceback') |
|
110 | 111 | if traceback is None: |
|
111 | 112 | report_data['traceback'] = report_data.get('frameinfo') |
|
112 | 113 | # for 0.3 api |
|
113 | 114 | error = report_data.pop('error_type', '') |
|
114 | 115 | if error: |
|
115 | 116 | report_data['error'] = error |
|
116 | 117 | if proto_version.minor < 4: |
|
117 | 118 | # convert "Unknown" slow reports to |
|
118 | 119 | # '' (from older clients) |
|
119 | 120 | if (report_data['error'] and |
|
120 | 121 | report_data['http_status'] < 500): |
|
121 | 122 | report_data['error'] = '' |
|
122 | 123 | message = report_data.get('message') |
|
123 | 124 | if 'extra' not in report_data: |
|
124 | 125 | report_data['extra'] = [] |
|
125 | 126 | if message: |
|
126 | 127 | report_data['extra'] = [('message', message), ] |
|
127 | 128 | reports.append(report_data) |
|
128 | 129 | else: |
|
129 | 130 | reports = dataset |
|
130 | 131 | |
|
131 | 132 | tags = [] |
|
132 | 133 | es_slow_calls_docs = {} |
|
133 | 134 | es_reports_stats_rows = {} |
|
134 | 135 | for report_data in reports: |
|
135 | 136 | # build report details for later |
|
136 | 137 | added_details = 0 |
|
137 | 138 | report = Report() |
|
138 | 139 | report.set_data(report_data, resource, proto_version) |
|
139 | 140 | report._skip_ft_index = True |
|
140 | 141 | |
|
141 | 142 | report_group = ReportGroupService.by_hash_and_resource( |
|
142 | 143 | report.resource_id, |
|
143 | 144 | report.grouping_hash |
|
144 | 145 | ) |
|
145 | 146 | occurences = report_data.get('occurences', 1) |
|
146 | 147 | if not report_group: |
|
147 | 148 | # total reports will be +1 moment later |
|
148 | 149 | report_group = ReportGroup(grouping_hash=report.grouping_hash, |
|
149 | 150 | occurences=0, total_reports=0, |
|
150 | 151 | last_report=0, |
|
151 | 152 | priority=report.priority, |
|
152 | 153 | error=report.error, |
|
153 | 154 | first_timestamp=report.start_time) |
|
154 | 155 | report_group._skip_ft_index = True |
|
155 | 156 | report_group.report_type = report.report_type |
|
156 | 157 | report.report_group_time = report_group.first_timestamp |
|
157 | 158 | add_sample = pick_sample(report_group.occurences, |
|
158 | 159 | report_type=report_group.report_type) |
|
159 | 160 | if add_sample: |
|
160 | 161 | resource.report_groups.append(report_group) |
|
161 | 162 | report_group.reports.append(report) |
|
162 | 163 | added_details += 1 |
|
163 | 164 | DBSession.flush() |
|
164 | 165 | if report.partition_id not in es_report_docs: |
|
165 | 166 | es_report_docs[report.partition_id] = [] |
|
166 | 167 | es_report_docs[report.partition_id].append(report.es_doc()) |
|
167 | 168 | tags.extend(list(report.tags.items())) |
|
168 | 169 | slow_calls = report.add_slow_calls(report_data, report_group) |
|
169 | 170 | DBSession.flush() |
|
170 | 171 | for s_call in slow_calls: |
|
171 | 172 | if s_call.partition_id not in es_slow_calls_docs: |
|
172 | 173 | es_slow_calls_docs[s_call.partition_id] = [] |
|
173 | 174 | es_slow_calls_docs[s_call.partition_id].append( |
|
174 | 175 | s_call.es_doc()) |
|
175 | 176 | # try generating new stat rows if needed |
|
176 | 177 | else: |
|
177 | 178 | # required for postprocessing to not fail later |
|
178 | 179 | report.report_group = report_group |
|
179 | 180 | |
|
180 | 181 | stat_row = ReportService.generate_stat_rows( |
|
181 | 182 | report, resource, report_group) |
|
182 | 183 | if stat_row.partition_id not in es_reports_stats_rows: |
|
183 | 184 | es_reports_stats_rows[stat_row.partition_id] = [] |
|
184 | 185 | es_reports_stats_rows[stat_row.partition_id].append( |
|
185 | 186 | stat_row.es_doc()) |
|
186 | 187 | |
|
187 | 188 | # see if we should mark 10th occurence of report |
|
188 | 189 | last_occurences_10 = int(math.floor(report_group.occurences / 10)) |
|
189 | 190 | curr_occurences_10 = int(math.floor( |
|
190 | 191 | (report_group.occurences + report.occurences) / 10)) |
|
191 | 192 | last_occurences_100 = int( |
|
192 | 193 | math.floor(report_group.occurences / 100)) |
|
193 | 194 | curr_occurences_100 = int(math.floor( |
|
194 | 195 | (report_group.occurences + report.occurences) / 100)) |
|
195 | 196 | notify_occurences_10 = last_occurences_10 != curr_occurences_10 |
|
196 | 197 | notify_occurences_100 = last_occurences_100 != curr_occurences_100 |
|
197 | 198 | report_group.occurences = ReportGroup.occurences + occurences |
|
198 | 199 | report_group.last_timestamp = report.start_time |
|
199 | 200 | report_group.summed_duration = ReportGroup.summed_duration + report.duration |
|
200 | 201 | summed_duration = ReportGroup.summed_duration + report.duration |
|
201 | 202 | summed_occurences = ReportGroup.occurences + occurences |
|
202 | 203 | report_group.average_duration = summed_duration / summed_occurences |
|
203 | 204 | report_group.run_postprocessing(report) |
|
204 | 205 | if added_details: |
|
205 | 206 | report_group.total_reports = ReportGroup.total_reports + 1 |
|
206 | 207 | report_group.last_report = report.id |
|
207 | 208 | report_group.set_notification_info(notify_10=notify_occurences_10, |
|
208 | 209 | notify_100=notify_occurences_100) |
|
209 | 210 | DBSession.flush() |
|
210 | 211 | report_group.get_report().notify_channel(report_group) |
|
211 | 212 | if report_group.partition_id not in es_report_group_docs: |
|
212 | 213 | es_report_group_docs[report_group.partition_id] = [] |
|
213 | 214 | es_report_group_docs[report_group.partition_id].append( |
|
214 | 215 | report_group.es_doc()) |
|
215 | 216 | |
|
216 | 217 | action = 'REPORT' |
|
217 | 218 | log_msg = '%s: %s %s, client: %s, proto: %s' % ( |
|
218 | 219 | action, |
|
219 | 220 | report_data.get('http_status', 'unknown'), |
|
220 | 221 | str(resource), |
|
221 | 222 | report_data.get('client'), |
|
222 | 223 | proto_version) |
|
223 | 224 | log.info(log_msg) |
|
224 | 225 | total_reports = len(dataset) |
|
225 | 226 | key = REDIS_KEYS['counters']['reports_per_minute'].format(current_time) |
|
226 | 227 | Datastores.redis.incr(key, total_reports) |
|
227 | 228 | Datastores.redis.expire(key, 3600 * 24) |
|
228 | 229 | key = REDIS_KEYS['counters']['reports_per_minute_per_app'].format( |
|
229 | 230 | resource_id, current_time) |
|
230 | 231 | Datastores.redis.incr(key, total_reports) |
|
231 | 232 | Datastores.redis.expire(key, 3600 * 24) |
|
232 | 233 | |
|
233 | 234 | add_reports_es(es_report_group_docs, es_report_docs) |
|
234 | 235 | add_reports_slow_calls_es(es_slow_calls_docs) |
|
235 | 236 | add_reports_stats_rows_es(es_reports_stats_rows) |
|
236 | 237 | return True |
|
237 | 238 | except Exception as exc: |
|
238 | 239 | print_traceback(log) |
|
239 | 240 | add_reports.retry(exc=exc) |
|
240 | 241 | |
|
241 | 242 | |
|
242 | 243 | @celery.task(queue="es", default_retry_delay=600, max_retries=999) |
|
243 | 244 | def add_reports_es(report_group_docs, report_docs): |
|
244 | 245 | for k, v in report_group_docs.items(): |
|
245 | 246 | Datastores.es.bulk_index(k, 'report_group', v, id_field="_id") |
|
246 | 247 | for k, v in report_docs.items(): |
|
247 | 248 | Datastores.es.bulk_index(k, 'report', v, id_field="_id", |
|
248 | 249 | parent_field='_parent') |
|
249 | 250 | |
|
250 | 251 | |
|
251 | 252 | @celery.task(queue="es", default_retry_delay=600, max_retries=999) |
|
252 | 253 | def add_reports_slow_calls_es(es_docs): |
|
253 | 254 | for k, v in es_docs.items(): |
|
254 | 255 | Datastores.es.bulk_index(k, 'log', v) |
|
255 | 256 | |
|
256 | 257 | |
|
257 | 258 | @celery.task(queue="es", default_retry_delay=600, max_retries=999) |
|
258 | 259 | def add_reports_stats_rows_es(es_docs): |
|
259 | 260 | for k, v in es_docs.items(): |
|
260 | 261 | Datastores.es.bulk_index(k, 'log', v) |
|
261 | 262 | |
|
262 | 263 | |
|
263 | 264 | @celery.task(queue="logs", default_retry_delay=600, max_retries=999) |
|
264 | 265 | def add_logs(resource_id, request, dataset, environ=None, **kwargs): |
|
265 | 266 | proto_version = request.get('protocol_version') |
|
266 | 267 | current_time = datetime.utcnow().replace(second=0, microsecond=0) |
|
267 | 268 | |
|
268 | 269 | try: |
|
269 | 270 | es_docs = collections.defaultdict(list) |
|
270 | 271 | application = ApplicationService.by_id(resource_id) |
|
271 | 272 | ns_pairs = [] |
|
272 | 273 | for entry in dataset: |
|
273 | 274 | # gather pk and ns so we can remove older versions of row later |
|
274 | 275 | if entry['primary_key'] is not None: |
|
275 | 276 | ns_pairs.append({"pk": entry['primary_key'], |
|
276 | 277 | "ns": entry['namespace']}) |
|
277 | 278 | log_entry = Log() |
|
278 | 279 | log_entry.set_data(entry, resource=application) |
|
279 | 280 | log_entry._skip_ft_index = True |
|
280 | 281 | application.logs.append(log_entry) |
|
281 | 282 | DBSession.flush() |
|
282 | 283 | # insert non pk rows first |
|
283 | 284 | if entry['primary_key'] is None: |
|
284 | 285 | es_docs[log_entry.partition_id].append(log_entry.es_doc()) |
|
285 | 286 | |
|
286 | 287 | # 2nd pass to delete all log entries from db foe same pk/ns pair |
|
287 | 288 | if ns_pairs: |
|
288 | 289 | ids_to_delete = [] |
|
289 | 290 | es_docs = collections.defaultdict(list) |
|
290 | 291 | es_docs_to_delete = collections.defaultdict(list) |
|
291 | 292 | found_pkey_logs = LogService.query_by_primary_key_and_namespace( |
|
292 | 293 | list_of_pairs=ns_pairs) |
|
293 | 294 | log_dict = {} |
|
294 | 295 | for log_entry in found_pkey_logs: |
|
295 | 296 | log_key = (log_entry.primary_key, log_entry.namespace) |
|
296 | 297 | if log_key not in log_dict: |
|
297 | 298 | log_dict[log_key] = [] |
|
298 | 299 | log_dict[log_key].append(log_entry) |
|
299 | 300 | |
|
300 | 301 | for ns, entry_list in log_dict.items(): |
|
301 | 302 | entry_list = sorted(entry_list, key=lambda x: x.timestamp) |
|
302 | 303 | # newest row needs to be indexed in es |
|
303 | 304 | log_entry = entry_list[-1] |
|
304 | 305 | # delete everything from pg and ES, leave the last row in pg |
|
305 | 306 | for e in entry_list[:-1]: |
|
306 | 307 | ids_to_delete.append(e.log_id) |
|
307 | 308 | es_docs_to_delete[e.partition_id].append(e.delete_hash) |
|
308 | 309 | |
|
309 | 310 | es_docs_to_delete[log_entry.partition_id].append( |
|
310 | 311 | log_entry.delete_hash) |
|
311 | 312 | |
|
312 | 313 | es_docs[log_entry.partition_id].append(log_entry.es_doc()) |
|
313 | 314 | |
|
314 | 315 | if ids_to_delete: |
|
315 | 316 | query = DBSession.query(Log).filter( |
|
316 | 317 | Log.log_id.in_(ids_to_delete)) |
|
317 | 318 | query.delete(synchronize_session=False) |
|
318 | 319 | if es_docs_to_delete: |
|
319 | 320 | # batch this to avoid problems with default ES bulk limits |
|
320 | 321 | for es_index in es_docs_to_delete.keys(): |
|
321 | 322 | for batch in in_batches(es_docs_to_delete[es_index], 20): |
|
322 | 323 | query = {'terms': {'delete_hash': batch}} |
|
323 | 324 | |
|
324 | 325 | try: |
|
325 | 326 | Datastores.es.delete_by_query( |
|
326 | 327 | es_index, 'log', query) |
|
327 | 328 | except pyelasticsearch.ElasticHttpNotFoundError as exc: |
|
328 | 329 | log.error(exc) |
|
329 | 330 | |
|
330 | 331 | total_logs = len(dataset) |
|
331 | 332 | |
|
332 | 333 | log_msg = 'LOG_NEW: %s, entries: %s, proto:%s' % ( |
|
333 | 334 | str(application), |
|
334 | 335 | total_logs, |
|
335 | 336 | proto_version) |
|
336 | 337 | log.info(log_msg) |
|
337 | 338 | # mark_changed(session) |
|
338 | 339 | key = REDIS_KEYS['counters']['logs_per_minute'].format(current_time) |
|
339 | 340 | Datastores.redis.incr(key, total_logs) |
|
340 | 341 | Datastores.redis.expire(key, 3600 * 24) |
|
341 | 342 | key = REDIS_KEYS['counters']['logs_per_minute_per_app'].format( |
|
342 | 343 | resource_id, current_time) |
|
343 | 344 | Datastores.redis.incr(key, total_logs) |
|
344 | 345 | Datastores.redis.expire(key, 3600 * 24) |
|
345 | 346 | add_logs_es(es_docs) |
|
346 | 347 | return True |
|
347 | 348 | except Exception as exc: |
|
348 | 349 | print_traceback(log) |
|
349 | 350 | add_logs.retry(exc=exc) |
|
350 | 351 | |
|
351 | 352 | |
|
352 | 353 | @celery.task(queue="es", default_retry_delay=600, max_retries=999) |
|
353 | 354 | def add_logs_es(es_docs): |
|
354 | 355 | for k, v in es_docs.items(): |
|
355 | 356 | Datastores.es.bulk_index(k, 'log', v) |
|
356 | 357 | |
|
357 | 358 | |
|
358 | 359 | @celery.task(queue="metrics", default_retry_delay=600, max_retries=999) |
|
359 | 360 | def add_metrics(resource_id, request, dataset, proto_version): |
|
360 | 361 | current_time = datetime.utcnow().replace(second=0, microsecond=0) |
|
361 | 362 | try: |
|
362 | 363 | application = ApplicationService.by_id_cached()(resource_id) |
|
363 | 364 | application = DBSession.merge(application, load=False) |
|
364 | 365 | es_docs = [] |
|
365 | 366 | rows = [] |
|
366 | 367 | for metric in dataset: |
|
367 | 368 | tags = dict(metric['tags']) |
|
368 | 369 | server_n = tags.get('server_name', metric['server_name']).lower() |
|
369 | 370 | tags['server_name'] = server_n or 'unknown' |
|
370 | 371 | new_metric = Metric( |
|
371 | 372 | timestamp=metric['timestamp'], |
|
372 | 373 | resource_id=application.resource_id, |
|
373 | 374 | namespace=metric['namespace'], |
|
374 | 375 | tags=tags) |
|
375 | 376 | rows.append(new_metric) |
|
376 | 377 | es_docs.append(new_metric.es_doc()) |
|
377 | 378 | session = DBSession() |
|
378 | 379 | session.bulk_save_objects(rows) |
|
379 | 380 | session.flush() |
|
380 | 381 | |
|
381 | 382 | action = 'METRICS' |
|
382 | 383 | metrics_msg = '%s: %s, metrics: %s, proto:%s' % ( |
|
383 | 384 | action, |
|
384 | 385 | str(application), |
|
385 | 386 | len(dataset), |
|
386 | 387 | proto_version |
|
387 | 388 | ) |
|
388 | 389 | log.info(metrics_msg) |
|
389 | 390 | |
|
390 | 391 | mark_changed(session) |
|
391 | 392 | key = REDIS_KEYS['counters']['metrics_per_minute'].format(current_time) |
|
392 | 393 | Datastores.redis.incr(key, len(rows)) |
|
393 | 394 | Datastores.redis.expire(key, 3600 * 24) |
|
394 | 395 | key = REDIS_KEYS['counters']['metrics_per_minute_per_app'].format( |
|
395 | 396 | resource_id, current_time) |
|
396 | 397 | Datastores.redis.incr(key, len(rows)) |
|
397 | 398 | Datastores.redis.expire(key, 3600 * 24) |
|
398 | 399 | add_metrics_es(es_docs) |
|
399 | 400 | return True |
|
400 | 401 | except Exception as exc: |
|
401 | 402 | print_traceback(log) |
|
402 | 403 | add_metrics.retry(exc=exc) |
|
403 | 404 | |
|
404 | 405 | |
|
405 | 406 | @celery.task(queue="es", default_retry_delay=600, max_retries=999) |
|
406 | 407 | def add_metrics_es(es_docs): |
|
407 | 408 | for doc in es_docs: |
|
408 | 409 | partition = 'rcae_m_%s' % doc['timestamp'].strftime('%Y_%m_%d') |
|
409 | 410 | Datastores.es.index(partition, 'log', doc) |
|
410 | 411 | |
|
411 | 412 | |
|
412 | 413 | @celery.task(queue="default", default_retry_delay=5, max_retries=2) |
|
413 | 414 | def check_user_report_notifications(resource_id): |
|
414 | 415 | since_when = datetime.utcnow() |
|
415 | 416 | try: |
|
416 | 417 | request = get_current_request() |
|
417 | 418 | application = ApplicationService.by_id(resource_id) |
|
418 | 419 | if not application: |
|
419 | 420 | return |
|
420 | 421 | error_key = REDIS_KEYS['reports_to_notify_per_type_per_app'].format( |
|
421 | 422 | ReportType.error, resource_id) |
|
422 | 423 | slow_key = REDIS_KEYS['reports_to_notify_per_type_per_app'].format( |
|
423 | 424 | ReportType.slow, resource_id) |
|
424 | 425 | error_group_ids = Datastores.redis.smembers(error_key) |
|
425 | 426 | slow_group_ids = Datastores.redis.smembers(slow_key) |
|
426 | 427 | Datastores.redis.delete(error_key) |
|
427 | 428 | Datastores.redis.delete(slow_key) |
|
428 | 429 | err_gids = [int(g_id) for g_id in error_group_ids] |
|
429 | 430 | slow_gids = [int(g_id) for g_id in list(slow_group_ids)] |
|
430 | 431 | group_ids = err_gids + slow_gids |
|
431 | 432 | occurence_dict = {} |
|
432 | 433 | for g_id in group_ids: |
|
433 | 434 | key = REDIS_KEYS['counters']['report_group_occurences'].format( |
|
434 | 435 | g_id) |
|
435 | 436 | val = Datastores.redis.get(key) |
|
436 | 437 | Datastores.redis.delete(key) |
|
437 | 438 | if val: |
|
438 | 439 | occurence_dict[g_id] = int(val) |
|
439 | 440 | else: |
|
440 | 441 | occurence_dict[g_id] = 1 |
|
441 | 442 | report_groups = ReportGroupService.by_ids(group_ids) |
|
442 | 443 | report_groups.options(sa.orm.joinedload(ReportGroup.last_report_ref)) |
|
443 | 444 | |
|
444 | 445 | ApplicationService.check_for_groups_alert( |
|
445 | 446 | application, 'alert', report_groups=report_groups, |
|
446 | 447 | occurence_dict=occurence_dict) |
|
447 | 448 | users = set([p.user for p in application.users_for_perm('view')]) |
|
448 | 449 | report_groups = report_groups.all() |
|
449 | 450 | for user in users: |
|
450 | 451 | UserService.report_notify(user, request, application, |
|
451 | 452 | report_groups=report_groups, |
|
452 | 453 | occurence_dict=occurence_dict) |
|
453 | 454 | for group in report_groups: |
|
454 | 455 | # marks report_groups as notified |
|
455 | 456 | if not group.notified: |
|
456 | 457 | group.notified = True |
|
457 | 458 | except Exception as exc: |
|
458 | 459 | print_traceback(log) |
|
459 | 460 | raise |
|
460 | 461 | |
|
461 | 462 | |
|
462 | 463 | @celery.task(queue="default", default_retry_delay=5, max_retries=2) |
|
463 | 464 | def check_alerts(resource_id): |
|
464 | 465 | since_when = datetime.utcnow() |
|
465 | 466 | try: |
|
466 | 467 | request = get_current_request() |
|
467 | 468 | application = ApplicationService.by_id(resource_id) |
|
468 | 469 | if not application: |
|
469 | 470 | return |
|
470 | 471 | error_key = REDIS_KEYS[ |
|
471 | 472 | 'reports_to_notify_per_type_per_app_alerting'].format( |
|
472 | 473 | ReportType.error, resource_id) |
|
473 | 474 | slow_key = REDIS_KEYS[ |
|
474 | 475 | 'reports_to_notify_per_type_per_app_alerting'].format( |
|
475 | 476 | ReportType.slow, resource_id) |
|
476 | 477 | error_group_ids = Datastores.redis.smembers(error_key) |
|
477 | 478 | slow_group_ids = Datastores.redis.smembers(slow_key) |
|
478 | 479 | Datastores.redis.delete(error_key) |
|
479 | 480 | Datastores.redis.delete(slow_key) |
|
480 | 481 | err_gids = [int(g_id) for g_id in error_group_ids] |
|
481 | 482 | slow_gids = [int(g_id) for g_id in list(slow_group_ids)] |
|
482 | 483 | group_ids = err_gids + slow_gids |
|
483 | 484 | occurence_dict = {} |
|
484 | 485 | for g_id in group_ids: |
|
485 | 486 | key = REDIS_KEYS['counters'][ |
|
486 | 487 | 'report_group_occurences_alerting'].format( |
|
487 | 488 | g_id) |
|
488 | 489 | val = Datastores.redis.get(key) |
|
489 | 490 | Datastores.redis.delete(key) |
|
490 | 491 | if val: |
|
491 | 492 | occurence_dict[g_id] = int(val) |
|
492 | 493 | else: |
|
493 | 494 | occurence_dict[g_id] = 1 |
|
494 | 495 | report_groups = ReportGroupService.by_ids(group_ids) |
|
495 | 496 | report_groups.options(sa.orm.joinedload(ReportGroup.last_report_ref)) |
|
496 | 497 | |
|
497 | 498 | ApplicationService.check_for_groups_alert( |
|
498 | 499 | application, 'alert', report_groups=report_groups, |
|
499 | 500 | occurence_dict=occurence_dict, since_when=since_when) |
|
500 | 501 | except Exception as exc: |
|
501 | 502 | print_traceback(log) |
|
502 | 503 | raise |
|
503 | 504 | |
|
504 | 505 | |
|
505 | 506 | @celery.task(queue="default", default_retry_delay=1, max_retries=2) |
|
506 | 507 | def close_alerts(): |
|
507 | 508 | log.warning('Checking alerts') |
|
508 | 509 | since_when = datetime.utcnow() |
|
509 | 510 | try: |
|
510 | 511 | event_types = [Event.types['error_report_alert'], |
|
511 | 512 | Event.types['slow_report_alert'], ] |
|
512 | 513 | statuses = [Event.statuses['active']] |
|
513 | 514 | # get events older than 5 min |
|
514 | 515 | events = EventService.by_type_and_status( |
|
515 | 516 | event_types, |
|
516 | 517 | statuses, |
|
517 | 518 | older_than=(since_when - timedelta(minutes=5))) |
|
518 | 519 | for event in events: |
|
519 | 520 | # see if we can close them |
|
520 | 521 | event.validate_or_close( |
|
521 | 522 | since_when=(since_when - timedelta(minutes=1))) |
|
522 | 523 | except Exception as exc: |
|
523 | 524 | print_traceback(log) |
|
524 | 525 | raise |
|
525 | 526 | |
|
526 | 527 | |
|
527 | 528 | @celery.task(queue="default", default_retry_delay=600, max_retries=999) |
|
528 | 529 | def update_tag_counter(tag_name, tag_value, count): |
|
529 | 530 | try: |
|
530 | 531 | query = DBSession.query(Tag).filter(Tag.name == tag_name).filter( |
|
531 | 532 | sa.cast(Tag.value, sa.types.TEXT) == sa.cast(json.dumps(tag_value), |
|
532 | 533 | sa.types.TEXT)) |
|
533 | 534 | query.update({'times_seen': Tag.times_seen + count, |
|
534 | 535 | 'last_timestamp': datetime.utcnow()}, |
|
535 | 536 | synchronize_session=False) |
|
536 | 537 | session = DBSession() |
|
537 | 538 | mark_changed(session) |
|
538 | 539 | return True |
|
539 | 540 | except Exception as exc: |
|
540 | 541 | print_traceback(log) |
|
541 | 542 | update_tag_counter.retry(exc=exc) |
|
542 | 543 | |
|
543 | 544 | |
|
544 | 545 | @celery.task(queue="default") |
|
545 | 546 | def update_tag_counters(): |
|
546 | 547 | """ |
|
547 | 548 | Sets task to update counters for application tags |
|
548 | 549 | """ |
|
549 | 550 | tags = Datastores.redis.lrange(REDIS_KEYS['seen_tag_list'], 0, -1) |
|
550 | 551 | Datastores.redis.delete(REDIS_KEYS['seen_tag_list']) |
|
551 | 552 | c = collections.Counter(tags) |
|
552 | 553 | for t_json, count in c.items(): |
|
553 | 554 | tag_info = json.loads(t_json) |
|
554 | 555 | update_tag_counter.delay(tag_info[0], tag_info[1], count) |
|
555 | 556 | |
|
556 | 557 | |
|
557 | 558 | @celery.task(queue="default") |
|
558 | 559 | def daily_digest(): |
|
559 | 560 | """ |
|
560 | 561 | Sends daily digest with top 50 error reports |
|
561 | 562 | """ |
|
562 | 563 | request = get_current_request() |
|
563 | 564 | apps = Datastores.redis.smembers(REDIS_KEYS['apps_that_had_reports']) |
|
564 | 565 | Datastores.redis.delete(REDIS_KEYS['apps_that_had_reports']) |
|
565 | 566 | since_when = datetime.utcnow() - timedelta(hours=8) |
|
566 | 567 | log.warning('Generating daily digests') |
|
567 | 568 | for resource_id in apps: |
|
568 | 569 | resource_id = resource_id.decode('utf8') |
|
569 | 570 | end_date = datetime.utcnow().replace(microsecond=0, second=0) |
|
570 | 571 | filter_settings = {'resource': [resource_id], |
|
571 | 572 | 'tags': [{'name': 'type', |
|
572 | 573 | 'value': ['error'], 'op': None}], |
|
573 | 574 | 'type': 'error', 'start_date': since_when, |
|
574 | 575 | 'end_date': end_date} |
|
575 | 576 | |
|
576 | 577 | reports = ReportGroupService.get_trending( |
|
577 | 578 | request, filter_settings=filter_settings, limit=50) |
|
578 | 579 | |
|
579 | 580 | application = ApplicationService.by_id(resource_id) |
|
580 | 581 | if application: |
|
581 | 582 | users = set([p.user for p in application.users_for_perm('view')]) |
|
582 | 583 | for user in users: |
|
583 | 584 | user.send_digest(request, application, reports=reports, |
|
584 | 585 | since_when=since_when) |
|
585 | 586 | |
|
586 | 587 | |
|
587 | 588 | @celery.task(queue="default") |
|
588 | 589 | def notifications_reports(): |
|
589 | 590 | """ |
|
590 | 591 | Loop that checks redis for info and then issues new tasks to celery to |
|
591 | 592 | issue notifications |
|
592 | 593 | """ |
|
593 | 594 | apps = Datastores.redis.smembers(REDIS_KEYS['apps_that_had_reports']) |
|
594 | 595 | Datastores.redis.delete(REDIS_KEYS['apps_that_had_reports']) |
|
595 | 596 | for app in apps: |
|
596 | 597 | log.warning('Notify for app: %s' % app) |
|
597 | 598 | check_user_report_notifications.delay(app.decode('utf8')) |
|
598 | 599 | |
|
599 | 600 | @celery.task(queue="default") |
|
600 | 601 | def alerting_reports(): |
|
601 | 602 | """ |
|
602 | 603 | Loop that checks redis for info and then issues new tasks to celery to |
|
603 | 604 | perform the following: |
|
604 | 605 | - which applications should have new alerts opened |
|
605 | 606 | """ |
|
606 | 607 | |
|
607 | 608 | apps = Datastores.redis.smembers(REDIS_KEYS['apps_that_had_reports_alerting']) |
|
608 | 609 | Datastores.redis.delete(REDIS_KEYS['apps_that_had_reports_alerting']) |
|
609 | 610 | for app in apps: |
|
610 | 611 | log.warning('Notify for app: %s' % app) |
|
611 | 612 | check_alerts.delay(app.decode('utf8')) |
|
612 | 613 | |
|
613 | 614 | |
|
614 | 615 | @celery.task(queue="default", soft_time_limit=3600 * 4, |
|
615 | 616 | hard_time_limit=3600 * 4, max_retries=999) |
|
616 | 617 | def logs_cleanup(resource_id, filter_settings): |
|
617 | 618 | request = get_current_request() |
|
618 | 619 | request.tm.begin() |
|
619 | 620 | es_query = { |
|
620 | 621 | "_source": False, |
|
621 | 622 | "size": 5000, |
|
622 | 623 | "query": { |
|
623 | 624 | "filtered": { |
|
624 | 625 | "filter": { |
|
625 | 626 | "and": [{"term": {"resource_id": resource_id}}] |
|
626 | 627 | } |
|
627 | 628 | } |
|
628 | 629 | } |
|
629 | 630 | } |
|
630 | 631 | |
|
631 | 632 | query = DBSession.query(Log).filter(Log.resource_id == resource_id) |
|
632 | 633 | if filter_settings['namespace']: |
|
633 | 634 | query = query.filter(Log.namespace == filter_settings['namespace'][0]) |
|
634 | 635 | es_query['query']['filtered']['filter']['and'].append( |
|
635 | 636 | {"term": {"namespace": filter_settings['namespace'][0]}} |
|
636 | 637 | ) |
|
637 | 638 | query.delete(synchronize_session=False) |
|
638 | 639 | request.tm.commit() |
|
639 | 640 | result = request.es_conn.search(es_query, index='rcae_l_*', |
|
640 | 641 | doc_type='log', es_scroll='1m', |
|
641 | 642 | es_search_type='scan') |
|
642 | 643 | scroll_id = result['_scroll_id'] |
|
643 | 644 | while True: |
|
644 | 645 | log.warning('log_cleanup, app:{} ns:{} batch'.format( |
|
645 | 646 | resource_id, |
|
646 | 647 | filter_settings['namespace'] |
|
647 | 648 | )) |
|
648 | 649 | es_docs_to_delete = [] |
|
649 | 650 | result = request.es_conn.send_request( |
|
650 | 651 | 'POST', ['_search', 'scroll'], |
|
651 | 652 | body=scroll_id, query_params={"scroll": '1m'}) |
|
652 | 653 | scroll_id = result['_scroll_id'] |
|
653 | 654 | if not result['hits']['hits']: |
|
654 | 655 | break |
|
655 | 656 | for doc in result['hits']['hits']: |
|
656 | 657 | es_docs_to_delete.append({"id": doc['_id'], |
|
657 | 658 | "index": doc['_index']}) |
|
658 | 659 | |
|
659 | 660 | for batch in in_batches(es_docs_to_delete, 10): |
|
660 | 661 | Datastores.es.bulk([Datastores.es.delete_op(doc_type='log', |
|
661 | 662 | **to_del) |
|
662 | 663 | for to_del in batch]) |
General Comments 0
You need to be logged in to leave comments.
Login now