def manage_alarm(name, network, level, action, severity, alarm_id, timestamp):
""" Manage raise and clear of port and interface alarms """
ts = datetime.datetime.fromtimestamp(
float(timestamp)).strftime('%Y-%m-%d %H:%M:%S')
collectd.debug("%s %s %s %s alarm for %s:%s [%s] %s" % (PLUGIN,
severity, level, alarm_id, network, name, action, ts))
if action == ALARM_ACTION_CLEAR:
alarm_state = fm_constants.FM_ALARM_STATE_CLEAR
reason = ''
repair = ''
else:
# reason ad repair strings are only needed on alarm assertion
alarm_state = fm_constants.FM_ALARM_STATE_SET
reason = "'" + network.upper() + "' " + level
repair = 'Check cabling and far-end port configuration ' \
'and status on adjacent equipment.'
# build the alarm eid and name string
if level == LEVEL_PORT:
eid = 'host=' + obj.hostname + "." + level + '=' + name
reason += " failed"
else:
eid = 'host=' + obj.hostname + "." + level + '=' + network
if severity == fm_constants.FM_ALARM_SEVERITY_MAJOR:
reason += " degraded"
else:
reason += " failed"
if alarm_state == fm_constants.FM_ALARM_STATE_CLEAR:
if api.clear_fault(alarm_id, eid) is False:
collectd.error("%s %s:%s clear_fault failed" %
(PLUGIN, alarm_id, eid))
return True
else:
return False
else:
fault = fm_api.Fault(
uuid="",
alarm_id=alarm_id,
alarm_state=alarm_state,
entity_type_id=fm_constants.FM_ENTITY_TYPE_HOST,
entity_instance_id=eid,
severity=severity,
reason_text=reason,
alarm_type=fm_constants.FM_ALARM_TYPE_7,
probable_cause=fm_constants.ALARM_PROBABLE_CAUSE_UNKNOWN,
proposed_repair_action=repair,
service_affecting=True,
timestamp=ts,
suppression=True)
alarm_uuid = api.set_fault(fault)
if pc.is_uuid_like(alarm_uuid) is False:
collectd.error("%s %s:%s set_fault failed:%s" %
(PLUGIN, alarm_id, eid, alarm_uuid))
return True
else:
return False
def raise_alarm():
"""Raise Remote Logging Server Alarm"""
repair = 'Ensure Remote Log Server IP is reachable from '
repair += 'Controller through OAM interface; otherwise '
repair += 'contact next level of support.'
reason = 'Controller cannot establish connection with '
reason += 'remote logging server.'
try:
fault = fm_api.Fault(
alarm_id=PLUGIN_ALARMID,
alarm_state=fm_constants.FM_ALARM_STATE_SET,
entity_type_id=fm_constants.FM_ENTITY_TYPE_HOST,
entity_instance_id=obj.base_eid,
severity=fm_constants.FM_ALARM_SEVERITY_MINOR,
reason_text=reason,
alarm_type=fm_constants.FM_ALARM_TYPE_1,
probable_cause=fm_constants.ALARM_PROBABLE_CAUSE_6,
proposed_repair_action=repair,
service_affecting=False,
suppression=False)
alarm_uuid = api.set_fault(fault)
if pc.is_uuid_like(alarm_uuid) is False:
collectd.error("%s 'set_fault' failed ; %s:%s ; %s" %
(PLUGIN, PLUGIN_ALARMID,
obj.base_eid, alarm_uuid))
else:
collectd.info("%s %s:%s alarm raised" %
(PLUGIN, PLUGIN_ALARMID, obj.base_eid))
obj.alarmed = True
except Exception as ex:
collectd.error("%s 'set_fault' exception ; %s:%s ; %s " %
(PLUGIN, PLUGIN_ALARMID, obj.base_eid, ex))
def raise_alarm(alarm_cause, interface=None, data=0):
"""Assert a cause based PTP alarm"""
collectd.debug("%s Raising Alarm %d" % (PLUGIN, alarm_cause))
alarm = get_alarm_object(alarm_cause, interface)
if alarm is None:
# log created for None case in the get_alarm_object util
return True
# copy the reason as it might be updated for the OOT,
# most typical, case.
reason = alarm.reason
# Handle some special cases
#
if alarm_cause == ALARM_CAUSE__OOT:
# If this is an out of tolerance alarm then add the
# out of tolerance reading to the reason string before
# asserting the alarm.
#
# Keep the alarm updated with the latest sample reading
# and severity even if its already asserted.
if abs(float(data)) > 100000000000:
reason += 'more than 100 seconds'
elif abs(float(data)) > 10000000000:
reason += 'more than 10 seconds'
elif abs(float(data)) > 1000000000:
reason += 'more than 1 second'
elif abs(float(data)) > 1000000:
reason += str(abs(int(data)) / 1000000)
reason += ' millisecs'
elif abs(float(data)) > 1000:
reason += str(abs(int(data)) / 1000)
reason += ' microsecs'
else:
reason += str(float(data))
reason += ' ' + PLUGIN_TYPE_INSTANCE
elif alarm.raised is True:
# If alarm already raised then exit.
#
# All other alarms are a Major so there is no need to
# track a change in severity and update accordingly.
return True
elif alarm_cause == ALARM_CAUSE__PROCESS:
reason = 'Provisioned ' + PTP + ' \'' + obj.mode
reason += '\' time stamping mode seems to be unsupported by this host'
try:
fault = fm_api.Fault(
alarm_id=PLUGIN_ALARMID,
alarm_state=fm_constants.FM_ALARM_STATE_SET,
entity_type_id=fm_constants.FM_ENTITY_TYPE_HOST,
entity_instance_id=alarm.eid,
severity=alarm.severity,
reason_text=reason,
alarm_type=obj.alarm_type,
probable_cause=alarm.cause,
proposed_repair_action=alarm.repair,
service_affecting=False, # obj.service_affecting,
suppression=True) # obj.suppression)
alarm_uuid = api.set_fault(fault)
if pc.is_uuid_like(alarm_uuid) is False:
# Don't _add_unreachable_server list if the fm call failed.
# That way it will be retried at a later time.
collectd.error("%s 'set_fault' failed ; %s:%s ; %s" %
(PLUGIN, PLUGIN_ALARMID, alarm.eid, alarm_uuid))
return False
else:
collectd.info("%s %s:%s:%s alarm raised" %
(PLUGIN, PLUGIN_ALARMID, alarm.eid, alarm.severity))
alarm.raised = True
return True
except Exception as ex:
collectd.error("%s 'set_fault' exception ; %s:%s:%s ; %s" %
(PLUGIN,
PLUGIN_ALARMID,
alarm.eid,
alarm.severity,
ex))
return False
def manage_alarm(name, level, action, severity, alarm_id, timestamp):
"""Manage raise and clear port and interface alarms"""
ts = timestamp.strftime('%Y-%m-%d %H:%M:%S.%f')
if action == ALARM_ACTION_CLEAR:
alarm_state = fm_constants.FM_ALARM_STATE_CLEAR
reason = ''
repair = ''
else:
# reason ad repair strings are only needed on alarm assertion
alarm_state = fm_constants.FM_ALARM_STATE_SET
reason = "'" + name + "' " + level
repair = 'Check cabling and far-end port configuration ' \
'and status on adjacent equipment.'
# build the alarm eid and name string
if level == LEVEL_INTERFACE:
eid = 'host=' + obj.hostname + "." + level + '=' + name
reason += " failed"
else:
eid = 'host=' + obj.hostname + "." + level + '=' + name
if severity == fm_constants.FM_ALARM_SEVERITY_MAJOR:
reason += " degraded"
else:
reason += " failed"
if alarm_state == fm_constants.FM_ALARM_STATE_CLEAR:
try:
if api.clear_fault(alarm_id, eid) is False:
collectd.info("%s %s:%s alarm already cleared" %
(PLUGIN, alarm_id, eid))
else:
collectd.info("%s %s:%s alarm cleared" %
(PLUGIN, alarm_id, eid))
return True
except Exception as ex:
collectd.error("%s 'clear_fault' failed ; %s:%s ; %s" %
(PLUGIN, alarm_id, eid, ex))
return False
else:
fault = fm_api.Fault(
uuid="",
alarm_id=alarm_id,
alarm_state=alarm_state,
entity_type_id=fm_constants.FM_ENTITY_TYPE_HOST,
entity_instance_id=eid,
severity=severity,
reason_text=reason,
alarm_type=fm_constants.FM_ALARM_TYPE_7,
probable_cause=fm_constants.ALARM_PROBABLE_CAUSE_UNKNOWN,
proposed_repair_action=repair,
service_affecting=True,
timestamp=ts,
suppression=True)
try:
alarm_uuid = api.set_fault(fault)
except Exception as ex:
collectd.error("%s 'set_fault' exception ; %s:%s ; %s" %
(PLUGIN, alarm_id, eid, ex))
return False
if pc.is_uuid_like(alarm_uuid) is False:
collectd.error("%s 'set_fault' failed ; %s:%s ; %s" %
(PLUGIN, alarm_id, eid, alarm_uuid))
return False
else:
return True