octavia/octavia/controller/housekeeping/house_keeping.py

182 lines
8.1 KiB
Python

# Copyright 2015 Hewlett-Packard Development Company, L.P.
#
# Licensed under the Apache License, Version 2.0 (the "License"); you may
# not use this file except in compliance with the License. You may obtain
# a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
# License for the specific language governing permissions and limitations
# under the License.
from concurrent import futures
import datetime
from oslo_config import cfg
from oslo_log import log as logging
from oslo_utils import timeutils
from sqlalchemy.orm import exc as sqlalchemy_exceptions
from octavia.common import constants
from octavia.controller.worker.v1 import controller_worker as cw1
from octavia.controller.worker.v2 import controller_worker as cw2
from octavia.db import api as db_api
from octavia.db import repositories as repo
LOG = logging.getLogger(__name__)
CONF = cfg.CONF
class SpareAmphora(object):
def __init__(self):
self.amp_repo = repo.AmphoraRepository()
self.spares_repo = repo.SparesPoolRepository()
self.az_repo = repo.AvailabilityZoneRepository()
if CONF.api_settings.default_provider_driver == constants.AMPHORAV2:
self.cw = cw2.ControllerWorker()
self.check_booting_amphora = True
else:
self.cw = cw1.ControllerWorker()
self.check_booting_amphora = False
def spare_check(self):
"""Checks the DB for the Spare amphora count.
If it's less than the requirement, starts new amphora.
"""
lock_session = db_api.get_session(autocommit=False)
session = db_api.get_session()
try:
# Lock the spares_pool record for read and write
spare_amp_row = self.spares_repo.get_for_update(lock_session)
conf_spare_cnt = CONF.house_keeping.spare_amphora_pool_size
LOG.debug("Required Spare Amphora count : %d", conf_spare_cnt)
availability_zones, links = self.az_repo.get_all(session,
enabled=True)
compute_zones = set()
for az in availability_zones:
az_meta = self.az_repo.get_availability_zone_metadata_dict(
session, az.name)
compute_zones.add(az_meta.get(constants.COMPUTE_ZONE))
# If no AZs objects then build in the configured AZ (even if None)
# Also if configured AZ is not None then also build in there
# as could be different to the current AZs objects.
if CONF.nova.availability_zone or not compute_zones:
compute_zones.add(CONF.nova.availability_zone)
amp_booting = []
for az_name in compute_zones:
# TODO(rm_work): If az_name is None, this will get ALL spares
# across all AZs. This is the safest/most backwards compatible
# way I can think of, as cached_zone on the amphora records
# won't ever match. This should not impact any existing deploys
# with no AZ configured, as the behavior should be identical
# in that case. In the case of multiple AZs configured, it will
# simply ensure there are at least <N> spares *somewhere*, but
# will function more accurately if the operator actually
# configures the AZ setting properly.
curr_spare_cnt = self.amp_repo.get_spare_amphora_count(
session, availability_zone=az_name,
check_booting_amphora=self.check_booting_amphora)
LOG.debug("Current Spare Amphora count for AZ %s: %d",
az_name, curr_spare_cnt)
diff_count = conf_spare_cnt - curr_spare_cnt
# When the current spare amphora is less than required
if diff_count > 0:
LOG.info("Initiating creation of %d spare amphora "
"for az %s.", diff_count, az_name)
# Call Amphora Create Flow diff_count times
with futures.ThreadPoolExecutor(
max_workers=conf_spare_cnt) as executor:
for i in range(1, diff_count + 1):
LOG.debug("Starting amphorae number %d ...", i)
amp_booting.append(executor.submit(
self.cw.create_amphora, az_name))
else:
LOG.debug("Current spare amphora count for AZ %s "
"satisfies the requirement", az_name)
# Wait for the amphora boot threads to finish
futures.wait(amp_booting)
spare_amp_row.updated_at = timeutils.utcnow()
lock_session.commit()
except Exception:
lock_session.rollback()
class DatabaseCleanup(object):
def __init__(self):
self.amp_repo = repo.AmphoraRepository()
self.amp_health_repo = repo.AmphoraHealthRepository()
self.lb_repo = repo.LoadBalancerRepository()
def delete_old_amphorae(self):
"""Checks the DB for old amphora and deletes them based on its age."""
exp_age = datetime.timedelta(
seconds=CONF.house_keeping.amphora_expiry_age)
session = db_api.get_session()
amp_ids = self.amp_repo.get_all_deleted_expiring(session,
exp_age=exp_age)
for amp_id in amp_ids:
# If we're here, we already think the amp is expiring according to
# the amphora table. Now check it is expired in the health table.
# In this way, we ensure that amps aren't deleted unless they are
# both expired AND no longer receiving zombie heartbeats.
if self.amp_health_repo.check_amphora_health_expired(
session, amp_id, exp_age):
LOG.debug('Attempting to purge db record for Amphora ID: %s',
amp_id)
self.amp_repo.delete(session, id=amp_id)
try:
self.amp_health_repo.delete(session, amphora_id=amp_id)
except sqlalchemy_exceptions.NoResultFound:
pass # Best effort delete, this record might not exist
LOG.info('Purged db record for Amphora ID: %s', amp_id)
def cleanup_load_balancers(self):
"""Checks the DB for old load balancers and triggers their removal."""
exp_age = datetime.timedelta(
seconds=CONF.house_keeping.load_balancer_expiry_age)
session = db_api.get_session()
lb_ids = self.lb_repo.get_all_deleted_expiring(session,
exp_age=exp_age)
for lb_id in lb_ids:
LOG.info('Attempting to delete load balancer id : %s', lb_id)
self.lb_repo.delete(session, id=lb_id)
LOG.info('Deleted load balancer id : %s', lb_id)
class CertRotation(object):
def __init__(self):
self.threads = CONF.house_keeping.cert_rotate_threads
if CONF.api_settings.default_provider_driver == constants.AMPHORAV2:
self.cw = cw2.ControllerWorker()
else:
self.cw = cw1.ControllerWorker()
def rotate(self):
"""Check the amphora db table for expiring auth certs."""
amp_repo = repo.AmphoraRepository()
with futures.ThreadPoolExecutor(max_workers=self.threads) as executor:
session = db_api.get_session()
rotation_count = 0
while True:
amp = amp_repo.get_cert_expiring_amphora(session)
if not amp:
break
rotation_count += 1
LOG.debug("Cert expired amphora's id is: %s", amp.id)
executor.submit(self.cw.amphora_cert_rotation, amp.id)
if rotation_count > 0:
LOG.info("Rotated certificates for %s amphora", rotation_count)