osm_policy_module/autoscaling/service.py

   1 # -*- coding: utf-8 -*-
   2 # pylint: disable=no-member
   3
   4 # Copyright 2018 Whitestack, LLC
   5 # *************************************************************
   6
   7 # This file is part of OSM Monitoring module
   8 # All Rights Reserved to Whitestack, LLC
   9
  10 # Licensed under the Apache License, Version 2.0 (the "License"); you may
  11 # not use this file except in compliance with the License. You may obtain
  12 # a copy of the License at
  13
  14 #         http://www.apache.org/licenses/LICENSE-2.0
  15
  16 # Unless required by applicable law or agreed to in writing, software
  17 # distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
  18 # WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
  19 # License for the specific language governing permissions and limitations
  20 # under the License.
  21
  22 # For those usages not covered by the Apache License, Version 2.0 please
  23 # contact: bdiaz@whitestack.com or glavado@whitestack.com
  24 ##
  25 import asyncio
  26 import datetime
  27 import json
  28 import logging
  29 from typing import List
  30
  31 from osm_policy_module.common.common_db_client import CommonDbClient
  32 from osm_policy_module.common.lcm_client import LcmClient
  33 from osm_policy_module.common.mon_client import MonClient
  34 from osm_policy_module.core import database
  35 from osm_policy_module.core.config import Config
  36 from osm_policy_module.core.database import ScalingGroup, ScalingAlarm, ScalingPolicy, ScalingCriteria, \
  37     ScalingAlarmRepository, ScalingGroupRepository, ScalingPolicyRepository, ScalingCriteriaRepository
  38 from osm_policy_module.core.exceptions import VdurNotFound
  39 from osm_policy_module.utils.vnfd import VnfdUtils
  40
  41 log = logging.getLogger(__name__)
  42
  43
  44 class AutoscalingService:
  45
  46     def __init__(self, config: Config, loop=None):
  47         self.conf = config
  48         if not loop:
  49             loop = asyncio.get_event_loop()
  50         self.loop = loop
  51         self.db_client = CommonDbClient(config)
  52         self.mon_client = MonClient(config, loop=self.loop)
  53         self.lcm_client = LcmClient(config, loop=self.loop)
  54
  55     async def configure_scaling_groups(self, nsr_id: str):
  56         """
  57         Configures scaling groups for a network service. Creates records in DB. Creates alarms in MON.
  58         :param nsr_id: Network service record id
  59         :return:
  60         """
  61         log.info("Configuring scaling groups for network service with nsr_id: %s",
  62                  nsr_id)
  63         alarms_created = []
  64         database.db.connect()
  65         try:
  66             with database.db.atomic() as tx:
  67                 try:
  68                     vnfrs = self.db_client.get_vnfrs(nsr_id)
  69                     for vnfr in vnfrs:
  70                         log.debug("Processing vnfr: %s", vnfr)
  71                         vnfd = self.db_client.get_vnfd(vnfr['vnfd-id'])
  72                         if 'scaling-group-descriptor' not in vnfd:
  73                             log.debug("No scaling group present in vnfd")
  74                             continue
  75                         scaling_groups = vnfd['scaling-group-descriptor']
  76                         vnf_monitoring_params = vnfd['monitoring-param']
  77                         for scaling_group in scaling_groups:
  78                             scaling_group_record = self._get_or_create_scaling_group(nsr_id,
  79                                                                                      vnfr['member-vnf-index-ref'],
  80                                                                                      scaling_group)
  81                             for scaling_policy in scaling_group['scaling-policy']:
  82                                 if scaling_policy['scaling-type'] != 'automatic':
  83                                     continue
  84                                 scaling_policy_record = self._get_or_create_scaling_policy(nsr_id,
  85                                                                                            scaling_policy,
  86                                                                                            scaling_group_record)
  87
  88                                 for scaling_criteria in scaling_policy['scaling-criteria']:
  89                                     scaling_criteria_record = self._get_or_create_scaling_criteria(
  90                                         nsr_id,
  91                                         scaling_criteria,
  92                                         scaling_policy_record
  93                                     )
  94                                     vnf_monitoring_param = next(
  95                                         filter(
  96                                             lambda param: param['id'] == scaling_criteria[
  97                                                 'vnf-monitoring-param-ref'
  98                                             ],
  99                                             vnf_monitoring_params)
 100                                     )
 101                                     vdurs = self._get_monitored_vdurs(vnf_monitoring_param, vnfr['vdur'], vnfd)
 102                                     for vdur in vdurs:
 103                                         log.debug("Creating alarm for vdur %s ", vdur)
 104                                         try:
 105                                             ScalingAlarmRepository.get(ScalingAlarm.vdu_name == vdur['name'],
 106                                                                        ScalingCriteria.name == scaling_criteria['name'],
 107                                                                        ScalingPolicy.name == scaling_policy['name'],
 108                                                                        ScalingGroup.nsr_id == nsr_id,
 109                                                                        join_classes=[ScalingCriteria,
 110                                                                                      ScalingPolicy,
 111                                                                                      ScalingGroup])
 112                                             log.debug("vdu %s already has an alarm configured", vdur['name'])
 113                                             continue
 114                                         except ScalingAlarm.DoesNotExist:
 115                                             pass
 116                                         metric_name = self._get_metric_name(vnf_monitoring_param, vdur, vnfd)
 117                                         alarm_uuid = await self.mon_client.create_alarm(
 118                                             metric_name=metric_name,
 119                                             ns_id=nsr_id,
 120                                             vdu_name=vdur['name'],
 121                                             vnf_member_index=vnfr['member-vnf-index-ref'],
 122                                             threshold=scaling_criteria['scale-in-threshold'],
 123                                             operation=scaling_criteria['scale-in-relational-operation'],
 124                                             statistic=vnf_monitoring_param['aggregation-type']
 125                                         )
 126                                         alarm = ScalingAlarmRepository.create(
 127                                             alarm_uuid=alarm_uuid,
 128                                             action='scale_in',
 129                                             vnf_member_index=vnfr['member-vnf-index-ref'],
 130                                             vdu_name=vdur['name'],
 131                                             scaling_criteria=scaling_criteria_record
 132                                         )
 133                                         alarms_created.append(alarm)
 134                                         alarm_uuid = await self.mon_client.create_alarm(
 135                                             metric_name=metric_name,
 136                                             ns_id=nsr_id,
 137                                             vdu_name=vdur['name'],
 138                                             vnf_member_index=vnfr['member-vnf-index-ref'],
 139                                             threshold=scaling_criteria['scale-out-threshold'],
 140                                             operation=scaling_criteria['scale-out-relational-operation'],
 141                                             statistic=vnf_monitoring_param['aggregation-type']
 142                                         )
 143                                         alarm = ScalingAlarmRepository.create(
 144                                             alarm_uuid=alarm_uuid,
 145                                             action='scale_out',
 146                                             vnf_member_index=vnfr['member-vnf-index-ref'],
 147                                             vdu_name=vdur['name'],
 148                                             scaling_criteria=scaling_criteria_record
 149                                         )
 150                                         alarms_created.append(alarm)
 151
 152                 except Exception as e:
 153                     log.exception("Error configuring scaling groups:")
 154                     tx.rollback()
 155                     if len(alarms_created) > 0:
 156                         log.info("Cleaning alarm resources in MON")
 157                         for alarm in alarms_created:
 158                             await self.mon_client.delete_alarm(
 159                                 alarm.scaling_criteria.scaling_policy.scaling_group.nsr_id,
 160                                 alarm.vnf_member_index,
 161                                 alarm.vdu_name,
 162                                 alarm.alarm_uuid)
 163                     raise e
 164         finally:
 165             database.db.close()
 166
 167     async def delete_scaling_groups(self, nsr_id: str):
 168         log.debug("Deleting scaling groups for network service %s", nsr_id)
 169         database.db.connect()
 170         try:
 171             with database.db.atomic() as tx:
 172                 try:
 173                     for scaling_group in ScalingGroupRepository.list(ScalingGroup.nsr_id == nsr_id):
 174                         for scaling_policy in scaling_group.scaling_policies:
 175                             for scaling_criteria in scaling_policy.scaling_criterias:
 176                                 for alarm in scaling_criteria.scaling_alarms:
 177                                     try:
 178                                         await self.mon_client.delete_alarm(
 179                                             alarm.scaling_criteria.scaling_policy.scaling_group.nsr_id,
 180                                             alarm.vnf_member_index,
 181                                             alarm.vdu_name,
 182                                             alarm.alarm_uuid)
 183                                     except ValueError:
 184                                         log.exception("Error deleting alarm in MON %s", alarm.alarm_uuid)
 185                                     alarm.delete_instance()
 186                                 scaling_criteria.delete_instance()
 187                             scaling_policy.delete_instance()
 188                         scaling_group.delete_instance()
 189
 190                 except Exception as e:
 191                     log.exception("Error deleting scaling groups and alarms:")
 192                     tx.rollback()
 193                     raise e
 194         finally:
 195             database.db.close()
 196
 197     async def delete_orphaned_alarms(self, nsr_id):
 198         log.info("Deleting orphaned scaling alarms for network service %s", nsr_id)
 199         database.db.connect()
 200         try:
 201             with database.db.atomic() as tx:
 202                 try:
 203                     for scaling_group in ScalingGroupRepository.list(ScalingGroup.nsr_id == nsr_id):
 204                         for scaling_policy in scaling_group.scaling_policies:
 205                             for scaling_criteria in scaling_policy.scaling_criterias:
 206                                 for alarm in scaling_criteria.scaling_alarms:
 207                                     try:
 208                                         self.db_client.get_vdur(nsr_id, alarm.vnf_member_index, alarm.vdu_name)
 209                                     except VdurNotFound:
 210                                         log.debug("Deleting orphaned scaling alarm %s", alarm.alarm_uuid)
 211                                         try:
 212                                             await self.mon_client.delete_alarm(
 213                                                 alarm.scaling_criteria.scaling_policy.scaling_group.nsr_id,
 214                                                 alarm.vnf_member_index,
 215                                                 alarm.vdu_name,
 216                                                 alarm.alarm_uuid)
 217                                         except ValueError:
 218                                             log.exception("Error deleting alarm in MON %s", alarm.alarm_uuid)
 219                                         alarm.delete_instance()
 220
 221                 except Exception as e:
 222                     log.exception("Error deleting orphaned alarms:")
 223                     tx.rollback()
 224                     raise e
 225         finally:
 226             database.db.close()
 227
 228     async def handle_alarm(self, alarm_uuid: str, status: str):
 229         await self.update_alarm_status(alarm_uuid, status)
 230         await self.evaluate_policy(alarm_uuid)
 231
 232     async def update_alarm_status(self, alarm_uuid: str, status: str):
 233         database.db.connect()
 234         try:
 235             with database.db.atomic():
 236                 alarm = ScalingAlarmRepository.get(ScalingAlarm.alarm_uuid == alarm_uuid)
 237                 alarm.last_status = status
 238                 alarm.save()
 239         except ScalingAlarm.DoesNotExist:
 240             log.debug("There is no autoscaling action configured for alarm %s.", alarm_uuid)
 241         finally:
 242             database.db.close()
 243
 244     async def evaluate_policy(self, alarm_uuid):
 245         database.db.connect()
 246         try:
 247             with database.db.atomic():
 248                 alarm = ScalingAlarmRepository.get(ScalingAlarm.alarm_uuid == alarm_uuid)
 249                 vnf_member_index = alarm.vnf_member_index
 250                 action = alarm.action
 251                 scaling_policy = alarm.scaling_criteria.scaling_policy
 252                 if not scaling_policy.enabled:
 253                     return
 254                 if action == 'scale_in':
 255                     operation = scaling_policy.scale_in_operation
 256                 elif action == 'scale_out':
 257                     operation = scaling_policy.scale_out_operation
 258                 else:
 259                     raise Exception('Unknown alarm action {}'.format(alarm.action))
 260                 alarms = ScalingAlarmRepository.list(ScalingAlarm.scaling_criteria == alarm.scaling_criteria,
 261                                                      ScalingAlarm.action == alarm.action,
 262                                                      ScalingAlarm.vnf_member_index == vnf_member_index,
 263                                                      ScalingAlarm.vdu_name == alarm.vdu_name)
 264                 statuses = []
 265                 for alarm in alarms:
 266                     statuses.append(alarm.last_status)
 267                 if (operation == 'AND' and set(statuses) == {'alarm'}) or (operation == 'OR' and 'alarm' in statuses):
 268                     delta = datetime.datetime.now() - scaling_policy.last_scale
 269                     if delta.total_seconds() > scaling_policy.cooldown_time:
 270                         log.info("Sending %s action message for ns: %s",
 271                                  alarm.action,
 272                                  scaling_policy.scaling_group.nsr_id)
 273                         await self.lcm_client.scale(scaling_policy.scaling_group.nsr_id,
 274                                                     scaling_policy.scaling_group.name,
 275                                                     vnf_member_index,
 276                                                     action)
 277                         scaling_policy.last_scale = datetime.datetime.now()
 278                         scaling_policy.save()
 279
 280         except ScalingAlarm.DoesNotExist:
 281             log.debug("There is no autoscaling action configured for alarm %s.", alarm_uuid)
 282         finally:
 283             database.db.close()
 284
 285     def _get_or_create_scaling_group(self, nsr_id: str, vnf_member_index: str, scaling_group: dict):
 286         try:
 287             scaling_group_record = ScalingGroupRepository.get(
 288                 ScalingGroup.nsr_id == nsr_id,
 289                 ScalingGroup.vnf_member_index == vnf_member_index,
 290                 ScalingGroup.name == scaling_group['name']
 291             )
 292             log.debug("Found existing scaling group record in DB...")
 293         except ScalingGroup.DoesNotExist:
 294             log.debug("Creating scaling group record in DB...")
 295             scaling_group_record = ScalingGroupRepository.create(
 296                 nsr_id=nsr_id,
 297                 vnf_member_index=vnf_member_index,
 298                 name=scaling_group['name'],
 299                 content=json.dumps(scaling_group)
 300             )
 301             log.debug(
 302                 "Created scaling group record in DB : nsr_id=%s, vnf_member_index=%s, name=%s",
 303                 scaling_group_record.nsr_id,
 304                 scaling_group_record.vnf_member_index,
 305                 scaling_group_record.name)
 306         return scaling_group_record
 307
 308     def _get_or_create_scaling_policy(self, nsr_id: str, scaling_policy: dict, scaling_group_record: ScalingGroup):
 309         try:
 310             scaling_policy_record = ScalingPolicyRepository.get(
 311                 ScalingPolicy.name == scaling_policy['name'],
 312                 ScalingGroup.id == scaling_group_record.id,
 313                 join_classes=[ScalingGroup]
 314             )
 315             log.debug("Found existing scaling policy record in DB...")
 316         except ScalingPolicy.DoesNotExist:
 317             log.debug("Creating scaling policy record in DB...")
 318             scaling_policy_record = ScalingPolicyRepository.create(
 319                 nsr_id=nsr_id,
 320                 name=scaling_policy['name'],
 321                 cooldown_time=scaling_policy['cooldown-time'],
 322                 scaling_group=scaling_group_record,
 323             )
 324             if 'scale-in-operation-type' in scaling_policy:
 325                 scaling_policy_record.scale_in_operation = scaling_policy[
 326                     'scale-in-operation-type']
 327             if 'scale-out-operation-type' in scaling_policy:
 328                 scaling_policy_record.scale_out_operation = scaling_policy[
 329                     'scale-out-operation-type']
 330             if 'enabled' in scaling_policy:
 331                 scaling_policy_record.enabled = scaling_policy['enabled']
 332             scaling_policy_record.save()
 333             log.debug("Created scaling policy record in DB : name=%s, scaling_group.name=%s",
 334                       scaling_policy_record.name,
 335                       scaling_policy_record.scaling_group.name)
 336         return scaling_policy_record
 337
 338     def _get_or_create_scaling_criteria(self, nsr_id: str, scaling_criteria: dict,
 339                                         scaling_policy_record: ScalingPolicy):
 340         try:
 341             scaling_criteria_record = ScalingCriteriaRepository.get(
 342                 ScalingPolicy.id == scaling_policy_record.id,
 343                 ScalingCriteria.name == scaling_criteria['name'],
 344                 join_classes=[ScalingPolicy]
 345             )
 346             log.debug("Found existing scaling criteria record in DB...")
 347         except ScalingCriteria.DoesNotExist:
 348             log.debug("Creating scaling criteria record in DB...")
 349             scaling_criteria_record = ScalingCriteriaRepository.create(
 350                 nsr_id=nsr_id,
 351                 name=scaling_criteria['name'],
 352                 scaling_policy=scaling_policy_record
 353             )
 354             log.debug(
 355                 "Created scaling criteria record in DB : name=%s, scaling_policy.name=%s",
 356                 scaling_criteria_record.name,
 357                 scaling_criteria_record.scaling_policy.name)
 358         return scaling_criteria_record
 359
 360     def _get_monitored_vdurs(self, vnf_monitoring_param: dict, vdurs: List[dict], vnfd: dict):
 361         monitored_vdurs = []
 362         if 'vdu-monitoring-param' in vnf_monitoring_param:
 363             monitored_vdurs = list(
 364                 filter(
 365                     lambda vdur: vdur['vdu-id-ref'] == vnf_monitoring_param
 366                     ['vdu-monitoring-param']
 367                     ['vdu-ref'],
 368                     vdurs
 369                 )
 370             )
 371         elif 'vdu-metric' in vnf_monitoring_param:
 372             monitored_vdurs = list(
 373                 filter(
 374                     lambda vdur: vdur['vdu-id-ref'] == vnf_monitoring_param
 375                     ['vdu-metric']
 376                     ['vdu-ref'],
 377                     vdurs
 378                 )
 379             )
 380         elif 'vnf-metric' in vnf_monitoring_param:
 381             vdu = VnfdUtils.get_mgmt_vdu(vnfd)
 382             monitored_vdurs = list(
 383                 filter(
 384                     lambda vdur: vdur['vdu-id-ref'] == vdu['id'],
 385                     vdurs
 386                 )
 387             )
 388         else:
 389             log.warning(
 390                 "Scaling criteria is referring to a vnf-monitoring-param that does not "
 391                 "contain a reference to a vdu or vnf metric.")
 392         return monitored_vdurs
 393
 394     def _get_metric_name(self, vnf_monitoring_param: dict, vdur: dict, vnfd: dict):
 395         vdu = next(
 396             filter(lambda vdu: vdu['id'] == vdur['vdu-id-ref'], vnfd['vdu'])
 397         )
 398         if 'vdu-monitoring-param' in vnf_monitoring_param:
 399             vdu_monitoring_param = next(filter(
 400                 lambda param: param['id'] == vnf_monitoring_param['vdu-monitoring-param'][
 401                     'vdu-monitoring-param-ref'], vdu['monitoring-param']))
 402             nfvi_metric = vdu_monitoring_param['nfvi-metric']
 403             return nfvi_metric
 404         if 'vdu-metric' in vnf_monitoring_param:
 405             vnf_metric_name = vnf_monitoring_param['vdu-metric']['vdu-metric-name-ref']
 406             return vnf_metric_name
 407         if 'vnf-metric' in vnf_monitoring_param:
 408             vnf_metric_name = vnf_monitoring_param['vnf-metric']['vnf-metric-name-ref']
 409             return vnf_metric_name
 410         raise ValueError('No metric name found for vnf_monitoring_param %s' % vnf_monitoring_param['id'])