def notify_zaqar(self, action, message, headers=None):
queue_info = urlparse.parse_qs(action.query)
try:
# NOTE(flwang): Try to get build a pre-signed client if user has
# provide enough information about that. Otherwise, go to build
# a client with service account and queue name for this alarm.
conf, queue_name = self._get_presigned_client_conf(queue_info)
if conf is not None:
zaqar_client = self.get_zaqar_client(conf)
if conf is None or queue_name is None or zaqar_client is None:
zaqar_client = self.client
# queue_name is a combination of <alarm-id>-<topic>
queue_name = "%s-%s" % (message['body']['alarm_id'],
queue_info.get('topic')[-1])
# create a queue in zaqar
queue = zaqar_client.queue(queue_name)
subscriber_list = queue_info.get('subscriber', [])
ttl = int(queue_info.get('ttl', ['3600'])[-1])
for subscriber in subscriber_list:
# add subscriber to the zaqar queue
subscription_data = dict(subscriber=subscriber, ttl=ttl)
zaqar_client.subscription(queue_name, **subscription_data)
# post the message to the queue
queue.post(message)
except IndexError:
LOG.error(
_LE("Required query option missing in action %s"), action)
except Exception:
LOG.error(
_LE("Unknown error occurred; Failed to post message to"
" Zaqar queue"),
exc_info=True)
def notify_zaqar(self, action, message):
queue_info = urlparse.parse_qs(action.query)
try:
# queue_name is a combination of <alarm-id>-<topic>
queue_name = "%s-%s" % (message['body']['alarm_id'],
queue_info.get('topic')[-1])
# create a queue in zaqar
queue = self.client.queue(queue_name, force_create=True)
subscriber_list = queue_info.get('subscriber', [])
ttl = queue_info.get('ttl', [3600])[-1]
for subscriber in subscriber_list:
# add subscriber to the zaqar queue
subscription_data = dict(subscriber=subscriber, ttl=ttl)
self.client.subscription(queue_name, **subscription_data)
# post the message to the queue
queue.post(message)
except IndexError:
LOG.error(
_LE("Required topic query option missing in action %s") %
action)
except Exception:
LOG.error(_LE("Unknown error occurred; Failed to post message to"
" Zaqar queue"),
exc_info=True)
def notify_zaqar(self, action, message, headers=None):
queue_info = urlparse.parse_qs(action.query)
try:
# NOTE(flwang): Try to get build a pre-signed client if user has
# provide enough information about that. Otherwise, go to build
# a client with service account and queue name for this alarm.
conf, queue_name = self._get_presigned_client_conf(queue_info)
if conf is not None:
zaqar_client = self.get_zaqar_client(conf)
if conf is None or queue_name is None or zaqar_client is None:
zaqar_client = self.client
# queue_name is a combination of <alarm-id>-<topic>
queue_name = "%s-%s" % (message['body']['alarm_id'],
queue_info.get('topic')[-1])
# create a queue in zaqar
queue = zaqar_client.queue(queue_name)
subscriber_list = queue_info.get('subscriber', [])
ttl = int(queue_info.get('ttl', ['3600'])[-1])
for subscriber in subscriber_list:
# add subscriber to the zaqar queue
subscription_data = dict(subscriber=subscriber, ttl=ttl)
zaqar_client.subscription(queue_name, **subscription_data)
# post the message to the queue
queue.post(message)
except IndexError:
LOG.error(_LE("Required query option missing in action %s"),
action)
except Exception:
LOG.error(_LE("Unknown error occurred; Failed to post message to"
" Zaqar queue"),
exc_info=True)
def notify_zaqar(self, action, message):
queue_info = urlparse.parse_qs(action.query)
try:
# queue_name is a combination of <alarm-id>-<topic>
queue_name = "%s-%s" % (message['body']['alarm_id'],
queue_info.get('topic')[-1])
# create a queue in zaqar
queue = self.client.queue(queue_name, force_create=True)
subscriber_list = queue_info.get('subscriber', [])
ttl = queue_info.get('ttl', [3600])[-1]
for subscriber in subscriber_list:
# add subscriber to the zaqar queue
subscription_data = dict(subscriber=subscriber,
ttl=ttl)
self.client.subscription(queue_name,
**subscription_data)
# post the message to the queue
queue.post(message)
except IndexError:
LOG.error(_LE("Required topic query option missing in action %s")
% action)
except Exception:
LOG.error(_LE("Unknown error occurred; Failed to post message to"
" Zaqar queue"),
exc_info=True)
def _get_alarm_state(self, alarm_id):
try:
alarms = self._storage_conn.get_alarms(alarm_id=alarm_id)
except Exception:
LOG.exception(_LE('alarm %s retrieval failed'), alarm_id)
return None
if not alarms:
LOG.error(_LE("alarm %s doesn't exists anymore"), alarm_id)
return None
return list(alarms)[0].state
def _get_alarm_state(self, alarm_id):
try:
alarms = self._storage_conn.get_alarms(alarm_id=alarm_id)
except Exception:
LOG.exception(_LE('alarm %s retrieval failed'), alarm_id)
return None
if not alarms:
LOG.error(_LE("alarm %s doesn't exist anymore"), alarm_id)
return None
return list(alarms)[0].state
def _validate_conn_options(args):
nosql_scheme = urlparse.urlparse(args.nosql_conn).scheme
sql_scheme = urlparse.urlparse(args.sql_conn).scheme
if nosql_scheme not in ('mongodb', 'hbase'):
root_logger.error(_LE('Invalid source DB type %s, the source database '
'connection should be one of: [mongodb, hbase]'
), nosql_scheme)
sys.exit(1)
if sql_scheme not in ('mysql', 'mysql+pymysql', 'postgresql',
'sqlite'):
root_logger.error(_LE('Invalid destination DB type %s, the destination'
' database connection should be one of: '
'[mysql, postgresql, sqlite]'), sql_scheme)
sys.exit(1)
def _refresh(self, alarm, state, reason, reason_data, always_record=False):
"""Refresh alarm state."""
try:
previous = alarm.state
alarm.state = state
if previous != state or always_record:
LOG.info(_LI('alarm %(id)s transitioning to %(state)s because '
'%(reason)s'), {'id': alarm.alarm_id,
'state': state,
'reason': reason})
try:
self._storage_conn.update_alarm(alarm)
except storage.AlarmNotFound:
LOG.warning(_LW("Skip updating this alarm's state, the"
"alarm: %s has been deleted"),
alarm.alarm_id)
else:
self._record_change(alarm, reason)
self.notifier.notify(alarm, previous, reason, reason_data)
elif alarm.repeat_actions:
self.notifier.notify(alarm, previous, reason, reason_data)
except Exception:
# retry will occur naturally on the next evaluation
# cycle (unless alarm state reverts in the meantime)
LOG.exception(_LE('alarm state update failed'))
def extract_my_subset(self, group_id, universal_set):
"""Filters an iterable, returning only objects assigned to this agent.
We have a list of objects and get a list of active group members from
`tooz`. We then hash all the objects into buckets and return only
the ones that hashed into *our* bucket.
"""
if not group_id:
return universal_set
if group_id not in self._groups:
self.join_group(group_id)
try:
members = self._get_members(group_id)
LOG.debug('Members of group: %s, Me: %s', members, self._my_id)
if self._my_id not in members:
LOG.warning(_LW('Cannot extract tasks because agent failed to '
'join group properly. Rejoining group.'))
self.join_group(group_id)
members = self._get_members(group_id)
if self._my_id not in members:
raise MemberNotInGroupError(group_id, members, self._my_id)
LOG.debug('Members of group: %s, Me: %s', members, self._my_id)
hr = HashRing(members)
LOG.debug('Universal set: %s', universal_set)
my_subset = [v for v in universal_set
if hr.get_node(str(v)) == self._my_id]
LOG.debug('My subset: %s', my_subset)
return my_subset
except tooz.coordination.ToozError:
LOG.exception(_LE('Error getting group membership info from '
'coordination backend.'))
return []
def extract_my_subset(self, group_id, iterable):
"""Filters an iterable, returning only objects assigned to this agent.
We have a list of objects and get a list of active group members from
`tooz`. We then hash all the objects into buckets and return only
the ones that hashed into *our* bucket.
"""
if not group_id:
return iterable
if group_id not in self._groups:
self.join_group(group_id)
try:
members = self._get_members(group_id)
LOG.debug('Members of group: %s', members)
hr = HashRing(members)
filtered = [
v for v in iterable if hr.get_node(str(v)) == self._my_id
]
LOG.debug('My subset: %s', filtered)
return filtered
except tooz.coordination.ToozError:
LOG.exception(
_LE('Error getting group membership info from '
'coordination backend.'))
return []
def _validate(self):
"""Validate received event has mandatory parameters."""
if not self.obj:
LOG.error(_LE('Received invalid event (empty or None)'))
raise InvalidEvent()
if not self.obj.get('event_type'):
LOG.error(_LE('Failed to extract event_type from event = %s'),
self.obj)
raise InvalidEvent()
if not self.obj.get('message_id'):
LOG.error(_LE('Failed to extract message_id from event = %s'),
self.obj)
raise InvalidEvent()
def _refresh(self, alarm, state, reason, reason_data, always_record=False):
"""Refresh alarm state."""
try:
previous = alarm.state
alarm.state = state
if previous != state or always_record:
LOG.info(
_LI('alarm %(id)s transitioning to %(state)s because '
'%(reason)s'), {
'id': alarm.alarm_id,
'state': state,
'reason': reason
})
try:
self._storage_conn.update_alarm(alarm)
except storage.AlarmNotFound:
LOG.warning(
_LW("Skip updating this alarm's state, the"
"alarm: %s has been deleted"), alarm.alarm_id)
else:
self._record_change(alarm, reason)
self.notifier.notify(alarm, previous, reason, reason_data)
elif alarm.repeat_actions:
self.notifier.notify(alarm, previous, reason, reason_data)
except Exception:
# retry will occur naturally on the next evaluation
# cycle (unless alarm state reverts in the meantime)
LOG.exception(_LE('alarm state update failed'))
def evaluate_events(self, events):
"""Evaluate the events by referring related alarms."""
if not isinstance(events, list):
events = [events]
LOG.debug('Starting event alarm evaluation: #events = %d', len(events))
for e in events:
LOG.debug('Evaluating event: event = %s', e)
try:
event = Event(e)
except InvalidEvent:
LOG.debug('Aborting evaluation of the event.')
continue
for id, alarm in six.iteritems(
self._get_project_alarms(event.project)):
try:
self._evaluate_alarm(alarm, event)
except Exception:
LOG.exception(
_LE('Failed to evaluate alarm (id=%(a)s) '
'triggered by event = %(e)s.'), {
'a': id,
'e': e
})
LOG.debug('Finished event alarm evaluation.')
def _validate(self):
"""Validate received event has mandatory parameters."""
if not self.obj:
LOG.error(_LE('Received invalid event (empty or None)'))
raise InvalidEvent()
if not self.obj.get('event_type'):
LOG.error(_LE('Failed to extract event_type from event = %s'),
self.obj)
raise InvalidEvent()
if not self.obj.get('message_id'):
LOG.error(_LE('Failed to extract message_id from event = %s'),
self.obj)
raise InvalidEvent()
def evaluate_events(self, events):
"""Evaluate the events by referring related alarms."""
if not isinstance(events, list):
events = [events]
LOG.debug('Starting event alarm evaluation: #events = %d',
len(events))
for e in events:
LOG.debug('Evaluating event: event = %s', e)
try:
event = Event(e)
except InvalidEvent:
LOG.debug('Aborting evaluation of the event.')
continue
for id, alarm in six.iteritems(
self._get_project_alarms(event.project)):
try:
self._evaluate_alarm(alarm, event)
except Exception:
LOG.exception(_LE('Failed to evaluate alarm (id=%(a)s) '
'triggered by event = %(e)s.'),
{'a': id, 'e': e})
LOG.debug('Finished event alarm evaluation.')
def extract_my_subset(self, group_id, universal_set):
"""Filters an iterable, returning only objects assigned to this agent.
We have a list of objects and get a list of active group members from
`tooz`. We then hash all the objects into buckets and return only
the ones that hashed into *our* bucket.
"""
if not group_id:
return universal_set
if group_id not in self._groups:
self.join_group(group_id)
try:
members = self._get_members(group_id)
LOG.debug('Members of group: %s, Me: %s', members, self._my_id)
if self._my_id not in members:
LOG.warning(
_LW('Cannot extract tasks because agent failed to '
'join group properly. Rejoining group.'))
self.join_group(group_id)
members = self._get_members(group_id)
if self._my_id not in members:
raise MemberNotInGroupError(group_id, members, self._my_id)
LOG.debug('Members of group: %s, Me: %s', members, self._my_id)
hr = HashRing(members)
LOG.debug('Universal set: %s', universal_set)
my_subset = [
v for v in universal_set if hr.get_node(str(v)) == self._my_id
]
LOG.debug('My subset: %s', my_subset)
return my_subset
except tooz.coordination.ToozError:
LOG.exception(
_LE('Error getting group membership info from '
'coordination backend.'))
return []
def _handle_action(notifiers, action, alarm_id, alarm_name, severity,
previous, current, reason, reason_data):
"""Process action on alarm
:param notifiers: list of possible notifiers.
:param action: The action that is being attended, as a parsed URL.
:param alarm_id: The triggered alarm.
:param alarm_name: The name of triggered alarm.
:param severity: The level of triggered alarm
:param previous: The previous state of the alarm.
:param current: The current state of the alarm.
:param reason: The reason the alarm changed its state.
:param reason_data: A dict representation of the reason.
"""
try:
action = netutils.urlsplit(action)
except Exception:
LOG.error(
_LE("Unable to parse action %(action)s for alarm "
"%(alarm_id)s"), {
'action': action,
'alarm_id': alarm_id
})
return
try:
notifier = notifiers[action.scheme].obj
except KeyError:
scheme = action.scheme
LOG.error(
_LE("Action %(scheme)s for alarm %(alarm_id)s is unknown, "
"cannot notify"), {
'scheme': scheme,
'alarm_id': alarm_id
})
return
try:
LOG.debug("Notifying alarm %(id)s with action %(act)s", {
'id': alarm_id,
'act': action
})
notifier.notify(action, alarm_id, alarm_name, severity, previous,
current, reason, reason_data)
except Exception:
LOG.exception(_LE("Unable to notify alarm %s"), alarm_id)
def get_zaqar_client(self, conf):
try:
from zaqarclient.queues import client as zaqar_client
return zaqar_client.Client(
self._get_endpoint(), version=2, conf=conf)
except Exception:
LOG.error(
_LE("Failed to connect to Zaqar service "), exc_info=True)
def __init__(self, group_id, members, my_id):
super(MemberNotInGroupError, self).__init__(
_LE('Group ID: %{group_id}s, Members: %{members}s, Me: %{me}s: '
'Current agent is not part of group and cannot take tasks') % {
'group_id': group_id,
'members': members,
'me': my_id
})
def _evaluate_assigned_alarms(self):
try:
alarms = self._assigned_alarms()
LOG.info(_LI('initiating evaluation cycle on %d alarms'),
len(alarms))
for alarm in alarms:
self._evaluate_alarm(alarm)
except Exception:
LOG.exception(_LE('alarm evaluation cycle failed'))
def start(self):
if self.backend_url:
try:
self._coordinator = tooz.coordination.get_coordinator(
self.backend_url, self._my_id)
self._coordinator.start()
LOG.info(_LI('Coordination backend started successfully.'))
except tooz.coordination.ToozError:
LOG.exception(_LE('Error connecting to coordination backend.'))
def start(self):
if self.backend_url:
try:
self._coordinator = tooz.coordination.get_coordinator(
self.backend_url, self._my_id)
self._coordinator.start()
LOG.info(_LI('Coordination backend started successfully.'))
except tooz.coordination.ToozError:
LOG.exception(_LE('Error connecting to coordination backend.'))
def _convert_to_datetime(isotime):
try:
date_time = timeutils.parse_isotime(isotime)
date_time = date_time.replace(tzinfo=None)
return date_time
except ValueError:
LOG.exception(_LE("String %s is not a valid isotime"), isotime)
msg = _('Failed to parse the timestamp value %s') % isotime
raise base.ClientSideError(msg)
def get_zaqar_client(self, conf):
try:
from zaqarclient.queues import client as zaqar_client
return zaqar_client.Client(self._get_endpoint(),
version=2,
conf=conf)
except Exception:
LOG.error(_LE("Failed to connect to Zaqar service "),
exc_info=True)
def _evaluate_assigned_alarms(self):
try:
alarms = self._assigned_alarms()
LOG.info(_LI('initiating evaluation cycle on %d alarms'),
len(alarms))
for alarm in alarms:
self._evaluate_alarm(alarm)
except Exception:
LOG.exception(_LE('alarm evaluation cycle failed'))
def heartbeat(self):
if self._coordinator:
if not self._coordinator.is_started:
# re-connect
self.start()
try:
self._coordinator.heartbeat()
except tooz.coordination.ToozError:
LOG.exception(_LE('Error sending a heartbeat to coordination '
'backend.'))
def notify_zaqar(self, action, message, headers):
queue_info = urlparse.parse_qs(action.query)
try:
queue_name = queue_info.get('queue_name')[-1]
except IndexError:
LOG.error(_LE("Required 'queue_name' query option missing in"
" action %s"),
action)
return
try:
conf = self._get_client_conf(headers['X-Auth-Token'])
client = self.get_zaqar_client(conf)
queue = client.queue(queue_name)
queue.post(message)
except Exception:
LOG.error(_LE("Unknown error occurred; Failed to post message to"
" Zaqar queue"),
exc_info=True)
def notify_zaqar(self, action, message, headers):
queue_info = urlparse.parse_qs(action.query)
try:
queue_name = queue_info.get('queue_name')[-1]
except IndexError:
LOG.error(
_LE("Required 'queue_name' query option missing in"
" action %s"), action)
return
try:
conf = self._get_client_conf(headers['X-Auth-Token'])
client = self.get_zaqar_client(conf)
queue = client.queue(queue_name)
queue.post(message)
except Exception:
LOG.error(_LE("Unknown error occurred; Failed to post message to"
" Zaqar queue"),
exc_info=True)
def _get_endpoint(self):
try:
ks_client = keystone_client.get_client(self.conf)
return ks_client.service_catalog.url_for(
service_type=cfg.CONF.service_types.zaqar,
endpoint_type=self.conf.service_credentials.os_endpoint_type)
except Exception:
LOG.error(_LE("Aodh was configured to use zaqar:// action,"
" but Zaqar endpoint could not be found in Keystone"
" service catalog."))
def heartbeat(self):
if self._coordinator:
if not self._coordinator.is_started:
# re-connect
self.start()
try:
self._coordinator.heartbeat()
except tooz.coordination.ToozError:
LOG.exception(_LE('Error sending a heartbeat to coordination '
'backend.'))
def post(self, data):
"""Create a new alarm.
:param data: an alarm within the request body.
"""
rbac.enforce('create_alarm', pecan.request.headers,
pecan.request.enforcer, {})
conn = pecan.request.storage
now = timeutils.utcnow()
data.alarm_id = uuidutils.generate_uuid()
user_limit, project_limit = rbac.get_limited_to(
pecan.request.headers, pecan.request.enforcer)
def _set_ownership(aspect, owner_limitation, header):
attr = '%s_id' % aspect
requested_owner = getattr(data, attr)
explicit_owner = requested_owner != wtypes.Unset
caller = pecan.request.headers.get(header)
if (owner_limitation and explicit_owner
and requested_owner != caller):
raise base.ProjectNotAuthorized(requested_owner, aspect)
actual_owner = (owner_limitation or requested_owner
if explicit_owner else caller)
setattr(data, attr, actual_owner)
_set_ownership('user', user_limit, 'X-User-Id')
_set_ownership('project', project_limit, 'X-Project-Id')
# Check if there's room for one more alarm
if is_over_quota(conn, data.project_id, data.user_id):
raise OverQuota(data)
data.timestamp = now
data.state_timestamp = now
ALARMS_RULES[data.type].plugin.create_hook(data)
change = data.as_dict(models.Alarm)
data.update_actions()
try:
alarm_in = models.Alarm(**change)
except Exception:
LOG.exception(_LE("Error while posting alarm: %s"), change)
raise base.ClientSideError(_("Alarm incorrect"))
alarm = conn.create_alarm(alarm_in)
self._record_creation(conn, change, alarm.alarm_id, now)
v2_utils.set_resp_location_hdr("/alarms/" + alarm.alarm_id)
return Alarm.from_db_model(alarm)
def post(self, data):
"""Create a new alarm.
:param data: an alarm within the request body.
"""
rbac.enforce('create_alarm', pecan.request.headers,
pecan.request.enforcer, {})
conn = pecan.request.storage
now = timeutils.utcnow()
data.alarm_id = str(uuid.uuid4())
user_limit, project_limit = rbac.get_limited_to(pecan.request.headers,
pecan.request.enforcer)
def _set_ownership(aspect, owner_limitation, header):
attr = '%s_id' % aspect
requested_owner = getattr(data, attr)
explicit_owner = requested_owner != wtypes.Unset
caller = pecan.request.headers.get(header)
if (owner_limitation and explicit_owner
and requested_owner != caller):
raise base.ProjectNotAuthorized(requested_owner, aspect)
actual_owner = (owner_limitation or
requested_owner if explicit_owner else caller)
setattr(data, attr, actual_owner)
_set_ownership('user', user_limit, 'X-User-Id')
_set_ownership('project', project_limit, 'X-Project-Id')
# Check if there's room for one more alarm
if is_over_quota(conn, data.project_id, data.user_id):
raise OverQuota(data)
data.timestamp = now
data.state_timestamp = now
ALARMS_RULES[data.type].plugin.create_hook(data)
change = data.as_dict(models.Alarm)
data.update_actions()
try:
alarm_in = models.Alarm(**change)
except Exception:
LOG.exception(_LE("Error while posting alarm: %s"), change)
raise base.ClientSideError(_("Alarm incorrect"))
alarm = conn.create_alarm(alarm_in)
self._record_creation(conn, change, alarm.alarm_id, now)
v2_utils.set_resp_location_hdr("/v2/alarms/" + alarm.alarm_id)
return Alarm.from_db_model(alarm)
def _get_endpoint(self):
try:
ks_client = keystone_client.get_client(self.conf)
return ks_client.service_catalog.url_for(
service_type=cfg.CONF.service_types.zaqar,
endpoint_type=self.conf.service_credentials.os_endpoint_type)
except Exception:
LOG.error(
_LE("Aodh was configured to use zaqar:// action,"
" but Zaqar endpoint could not be found in Keystone"
" service catalog."))
def _evaluate_alarm(self, alarm):
"""Evaluate the alarms assigned to this evaluator."""
if alarm.type not in self.evaluators:
LOG.debug('skipping alarm %s: type unsupported', alarm.alarm_id)
return
LOG.debug('evaluating alarm %s', alarm.alarm_id)
try:
self.evaluators[alarm.type].obj.evaluate(alarm)
except Exception:
LOG.exception(_LE('Failed to evaluate alarm %s'), alarm.alarm_id)
def _evaluate_alarm(self, alarm):
"""Evaluate the alarms assigned to this evaluator."""
if alarm.type not in self.evaluators:
LOG.debug('skipping alarm %s: type unsupported', alarm.alarm_id)
return
LOG.debug('evaluating alarm %s', alarm.alarm_id)
try:
self.evaluators[alarm.type].obj.evaluate(alarm)
except Exception:
LOG.exception(_LE('Failed to evaluate alarm %s'), alarm.alarm_id)
def start(self):
backend_url = cfg.CONF.coordination.backend_url
if backend_url:
try:
self._coordinator = tooz.coordination.get_coordinator(
backend_url, self._my_id)
self._coordinator.start()
self._started = True
LOG.info(_LI('Coordination backend started successfully.'))
except tooz.coordination.ToozError:
self._started = False
LOG.exception(_LE('Error connecting to coordination backend.'))
def _handle_action(notifiers, action, alarm_id, alarm_name, severity,
previous, current, reason, reason_data):
"""Process action on alarm
:param notifiers: list of possible notifiers.
:param action: The action that is being attended, as a parsed URL.
:param alarm_id: The triggered alarm.
:param alarm_name: The name of triggered alarm.
:param severity: The level of triggered alarm
:param previous: The previous state of the alarm.
:param current: The current state of the alarm.
:param reason: The reason the alarm changed its state.
:param reason_data: A dict representation of the reason.
"""
try:
action = netutils.urlsplit(action)
except Exception:
LOG.error(
_LE("Unable to parse action %(action)s for alarm "
"%(alarm_id)s"), {'action': action, 'alarm_id': alarm_id})
return
try:
notifier = notifiers[action.scheme].obj
except KeyError:
scheme = action.scheme
LOG.error(
_LE("Action %(scheme)s for alarm %(alarm_id)s is unknown, "
"cannot notify"),
{'scheme': scheme, 'alarm_id': alarm_id})
return
try:
LOG.debug("Notifying alarm %(id)s with action %(act)s",
{'id': alarm_id, 'act': action})
notifier.notify(action, alarm_id, alarm_name, severity,
previous, current, reason, reason_data)
except Exception:
LOG.exception(_LE("Unable to notify alarm %s"), alarm_id)
def stop(self):
if not self._coordinator:
return
for group in list(self._groups):
self.leave_group(group)
try:
self._coordinator.stop()
except tooz.coordination.ToozError:
LOG.exception(_LE('Error connecting to coordination backend.'))
finally:
self._coordinator = None
def stop(self):
if not self._coordinator:
return
for group in list(self._groups):
self.leave_group(group)
try:
self._coordinator.stop()
except tooz.coordination.ToozError:
LOG.exception(_LE('Error connecting to coordination backend.'))
finally:
self._coordinator = None
def _get_endpoint(self):
if self._zendpoint is None:
try:
ks_client = keystone_client.get_client(self.conf)
z_srv = ks_client.services.find(
type=self.conf.service_types.zaqar)
endpoint_type = self.conf.service_credentials.interface
z_endpoint = ks_client.endpoints.find(service_id=z_srv.id,
interface=endpoint_type)
self._zendpoint = z_endpoint.url
except Exception:
LOG.error(_LE("Aodh was configured to use zaqar:// action,"
" but Zaqar endpoint could not be found in"
" Keystone service catalog."))
return self._zendpoint
def put(self, data):
"""Modify this alarm.
:param data: an alarm within the request body.
"""
# Ensure alarm exists
alarm_in = self._enforce_rbac('change_alarm')
now = timeutils.utcnow()
data.alarm_id = self._id
user, project = rbac.get_limited_to(pecan.request.headers,
pecan.request.enforcer)
if user:
data.user_id = user
elif data.user_id == wtypes.Unset:
data.user_id = alarm_in.user_id
if project:
data.project_id = project
elif data.project_id == wtypes.Unset:
data.project_id = alarm_in.project_id
data.timestamp = now
if alarm_in.state != data.state:
data.state_timestamp = now
else:
data.state_timestamp = alarm_in.state_timestamp
ALARMS_RULES[data.type].plugin.update_hook(data)
old_data = Alarm.from_db_model(alarm_in)
old_alarm = old_data.as_dict(models.Alarm)
data.update_actions(old_data)
updated_alarm = data.as_dict(models.Alarm)
try:
alarm_in = models.Alarm(**updated_alarm)
except Exception:
LOG.exception(_LE("Error while putting alarm: %s"), updated_alarm)
raise base.ClientSideError(_("Alarm incorrect"))
alarm = pecan.request.storage.update_alarm(alarm_in)
change = dict(
(k, v) for k, v in updated_alarm.items()
if v != old_alarm[k] and k not in ['timestamp', 'state_timestamp'])
self._record_change(change, now, on_behalf_of=alarm.project_id)
return Alarm.from_db_model(alarm)
def put(self, data):
"""Modify this alarm.
:param data: an alarm within the request body.
"""
# Ensure alarm exists
alarm_in = self._enforce_rbac('change_alarm')
now = timeutils.utcnow()
data.alarm_id = self._id
user, project = rbac.get_limited_to(pecan.request.headers,
pecan.request.enforcer)
if user:
data.user_id = user
elif data.user_id == wtypes.Unset:
data.user_id = alarm_in.user_id
if project:
data.project_id = project
elif data.project_id == wtypes.Unset:
data.project_id = alarm_in.project_id
data.timestamp = now
if alarm_in.state != data.state:
data.state_timestamp = now
else:
data.state_timestamp = alarm_in.state_timestamp
ALARMS_RULES[data.type].plugin.update_hook(data)
old_data = Alarm.from_db_model(alarm_in)
old_alarm = old_data.as_dict(models.Alarm)
data.update_actions(old_data)
updated_alarm = data.as_dict(models.Alarm)
try:
alarm_in = models.Alarm(**updated_alarm)
except Exception:
LOG.exception(_LE("Error while putting alarm: %s"), updated_alarm)
raise base.ClientSideError(_("Alarm incorrect"))
alarm = pecan.request.storage.update_alarm(alarm_in)
change = dict((k, v) for k, v in updated_alarm.items()
if v != old_alarm[k] and k not in
['timestamp', 'state_timestamp'])
self._record_change(change, now, on_behalf_of=alarm.project_id)
return Alarm.from_db_model(alarm)
def _get_endpoint(self):
if self._zendpoint is None:
try:
ks_client = keystone_client.get_client(self.conf)
z_srv = ks_client.services.find(
type=self.conf.service_types.zaqar)
endpoint_type = self.conf.service_credentials.interface
z_endpoint = ks_client.endpoints.find(service_id=z_srv.id,
interface=endpoint_type)
self._zendpoint = z_endpoint.url
except Exception:
LOG.error(
_LE("Aodh was configured to use zaqar:// action,"
" but Zaqar endpoint could not be found in"
" Keystone service catalog."))
return self._zendpoint
def _inner():
try:
join_req = self._coordinator.join_group(group_id)
join_req.get()
LOG.info(_LI('Joined partitioning group %s'), group_id)
except tooz.coordination.MemberAlreadyExist:
return
except tooz.coordination.GroupNotCreated:
create_grp_req = self._coordinator.create_group(group_id)
try:
create_grp_req.get()
except tooz.coordination.GroupAlreadyExist:
pass
raise ErrorJoiningPartitioningGroup()
except tooz.coordination.ToozError:
LOG.exception(_LE('Error joining partitioning group %s,'
' re-trying'), group_id)
raise ErrorJoiningPartitioningGroup()
self._groups.add(group_id)
def _inner():
try:
join_req = self._coordinator.join_group(group_id)
join_req.get()
LOG.info(_LI('Joined partitioning group %s'), group_id)
except tooz.coordination.MemberAlreadyExist:
return
except tooz.coordination.GroupNotCreated:
create_grp_req = self._coordinator.create_group(group_id)
try:
create_grp_req.get()
except tooz.coordination.GroupAlreadyExist:
pass
raise ErrorJoiningPartitioningGroup()
except tooz.coordination.ToozError:
LOG.exception(_LE('Error joining partitioning group %s,'
' re-trying'), group_id)
raise ErrorJoiningPartitioningGroup()
self._groups.add(group_id)
def get_zaqar_client(self):
conf = self.conf.service_credentials
params = {
"auth_opts": {
"backend": "keystone",
"options": {
"os_username": conf.os_username,
"os_password": conf.os_password,
"os_project_name": conf.os_tenant_name,
"os_auth_url": conf.os_auth_url,
"insecure": "",
},
}
}
try:
from zaqarclient.queues import client as zaqar_client
return zaqar_client.Client(self._get_endpoint(), version=1.1, conf=params)
except Exception:
LOG.error(_LE("Failed to connect to Zaqar service "), exc_info=True)
def get_zaqar_client(self):
conf = self.conf.service_credentials
params = {
'auth_opts': {
'backend': 'keystone',
'options': {
'os_username': conf.os_username,
'os_password': conf.os_password,
'os_project_name': conf.os_tenant_name,
'os_auth_url': conf.os_auth_url,
'insecure': ''
}
}
}
try:
from zaqarclient.queues import client as zaqar_client
return zaqar_client.Client(self._get_endpoint(),
version=1.1, conf=params)
except Exception:
LOG.error(_LE("Failed to connect to Zaqar service "),
exc_info=True)
def _process_alarm(notifiers, data):
"""Notify that alarm has been triggered.
:param notifiers: list of possible notifiers
:param data: (dict): alarm data
"""
actions = data.get('actions')
if not actions:
LOG.error(_LE("Unable to notify for an alarm with no action"))
return
for action in actions:
AlarmEndpoint._handle_action(notifiers, action,
data.get('alarm_id'),
data.get('alarm_name'),
data.get('severity'),
data.get('previous'),
data.get('current'),
data.get('reason'),
data.get('reason_data'))
def _process_alarm(notifiers, data):
"""Notify that alarm has been triggered.
:param notifiers: list of possible notifiers
:param data: (dict): alarm data
"""
actions = data.get('actions')
if not actions:
LOG.error(_LE("Unable to notify for an alarm with no action"))
return
for action in actions:
AlarmEndpoint._handle_action(notifiers, action,
data.get('alarm_id'),
data.get('alarm_name'),
data.get('severity'),
data.get('previous'),
data.get('current'),
data.get('reason'),
data.get('reason_data'))
def get_zaqar_client(self):
conf = self.conf.service_credentials
params = {
'auth_opts': {
'backend': 'keystone',
'options': {
'os_username': conf.os_username,
'os_password': conf.os_password,
'os_project_name': conf.os_tenant_name,
'os_auth_url': conf.os_auth_url,
'insecure': ''
}
}
}
try:
from zaqarclient.queues import client as zaqar_client
return zaqar_client.Client(self._get_endpoint(),
version=1.1,
conf=params)
except Exception:
LOG.error(_LE("Failed to connect to Zaqar service "),
exc_info=True)
def extract_my_subset(self, group_id, iterable):
"""Filters an iterable, returning only objects assigned to this agent.
We have a list of objects and get a list of active group members from
`tooz`. We then hash all the objects into buckets and return only
the ones that hashed into *our* bucket.
"""
if not group_id:
return iterable
if group_id not in self._groups:
self.join_group(group_id)
try:
members = self._get_members(group_id)
LOG.debug('Members of group: %s', members)
hr = utils.HashRing(members)
filtered = [v for v in iterable
if hr.get_node(str(v)) == self._my_id]
LOG.debug('My subset: %s', filtered)
return filtered
except tooz.coordination.ToozError:
LOG.exception(_LE('Error getting group membership info from '
'coordination backend.'))
return []
def __init__(self, group_id, members, my_id):
super(MemberNotInGroupError, self).__init__(_LE(
'Group ID: %(group_id)s, Members: %(members)s, Me: %(me)s: '
'Current agent is not part of group and cannot take tasks') %
{'group_id': group_id, 'members': members, 'me': my_id})
def __init__(self):
super(ErrorJoiningPartitioningGroup, self).__init__(_LE(
'Error occurred when joining partitioning group'))
def __call__(self, environ, start_response):
# Request for this state, modified by replace_start_response()
# and used when an error is being reported.
state = {}
def replacement_start_response(status, headers, exc_info=None):
"""Overrides the default response to make errors parsable."""
try:
status_code = int(status.split(' ')[0])
state['status_code'] = status_code
except (ValueError, TypeError): # pragma: nocover
raise Exception(('ErrorDocumentMiddleware received an invalid '
'status %s' % status))
else:
if (state['status_code'] // 100) not in (2, 3):
# Remove some headers so we can replace them later
# when we have the full error message and can
# compute the length.
headers = [(h, v) for (h, v) in headers
if h not in ('Content-Length', 'Content-Type')]
# Save the headers in case we need to modify them.
state['headers'] = headers
return start_response(status, headers, exc_info)
app_iter = self.app(environ, replacement_start_response)
if (state['status_code'] // 100) not in (2, 3):
req = webob.Request(environ)
error = environ.get('translatable_error')
user_locale = self.best_match_language(req.accept_language)
if (req.accept.best_match(['application/json', 'application/xml'
]) == 'application/xml'):
content_type = 'application/xml'
try:
# simple check xml is valid
fault = etree.fromstring(b'\n'.join(app_iter))
# Add the translated error to the xml data
if error is not None:
for fault_string in fault.findall('faultstring'):
fault_string.text = i18n.translate(
error, user_locale)
error_message = etree.tostring(fault)
body = b''.join((b'<error_message>', error_message,
b'</error_message>'))
except etree.XMLSyntaxError as err:
LOG.error(_LE('Error parsing HTTP response: %s'), err)
error_message = state['status_code']
body = '<error_message>%s</error_message>' % error_message
if six.PY3:
body = body.encode('utf-8')
else:
content_type = 'application/json'
app_data = b'\n'.join(app_iter)
if six.PY3:
app_data = app_data.decode('utf-8')
try:
fault = json.loads(app_data)
if error is not None and 'faultstring' in fault:
fault['faultstring'] = i18n.translate(
error, user_locale)
except ValueError as err:
fault = app_data
body = json.dumps({'error_message': fault})
if six.PY3:
body = body.encode('utf-8')
state['headers'].append(('Content-Length', str(len(body))))
state['headers'].append(('Content-Type', content_type))
body = [body]
else:
body = app_iter
return body
def __init__(self):
super(ErrorJoiningPartitioningGroup, self).__init__(
_LE('Error occurred when joining partitioning group'))
def __call__(self, environ, start_response):
# Request for this state, modified by replace_start_response()
# and used when an error is being reported.
state = {}
def replacement_start_response(status, headers, exc_info=None):
"""Overrides the default response to make errors parsable."""
try:
status_code = int(status.split(" ")[0])
state["status_code"] = status_code
except (ValueError, TypeError): # pragma: nocover
raise Exception(("ErrorDocumentMiddleware received an invalid " "status %s" % status))
else:
if (state["status_code"] // 100) not in (2, 3):
# Remove some headers so we can replace them later
# when we have the full error message and can
# compute the length.
headers = [(h, v) for (h, v) in headers if h not in ("Content-Length", "Content-Type")]
# Save the headers in case we need to modify them.
state["headers"] = headers
return start_response(status, headers, exc_info)
app_iter = self.app(environ, replacement_start_response)
if (state["status_code"] // 100) not in (2, 3):
req = webob.Request(environ)
error = environ.get("translatable_error")
user_locale = self.best_match_language(req.accept_language)
if req.accept.best_match(["application/json", "application/xml"]) == "application/xml":
content_type = "application/xml"
try:
# simple check xml is valid
fault = etree.fromstring(b"\n".join(app_iter))
# Add the translated error to the xml data
if error is not None:
for fault_string in fault.findall("faultstring"):
fault_string.text = i18n.translate(error, user_locale)
error_message = etree.tostring(fault)
body = b"".join((b"<error_message>", error_message, b"</error_message>"))
except etree.XMLSyntaxError as err:
LOG.error(_LE("Error parsing HTTP response: %s"), err)
error_message = state["status_code"]
body = "<error_message>%s</error_message>" % error_message
if six.PY3:
body = body.encode("utf-8")
else:
content_type = "application/json"
app_data = b"\n".join(app_iter)
if six.PY3:
app_data = app_data.decode("utf-8")
try:
fault = json.loads(app_data)
if error is not None and "faultstring" in fault:
fault["faultstring"] = i18n.translate(error, user_locale)
except ValueError as err:
fault = app_data
body = json.dumps({"error_message": fault})
if six.PY3:
body = body.encode("utf-8")
state["headers"].append(("Content-Length", str(len(body))))
state["headers"].append(("Content-Type", content_type))
body = [body]
else:
body = app_iter
return body
