def __init__(self, name, inEventQueue, outQueue, config_dict=None, number=0, checkpoint=None):
'''The constructor.
'''
EventAnalyzer.__init__(self, name, inEventQueue, outQueue, config_dict, number, checkpoint)
self.severity = "W"
self.recommendation = "Diagnose the problem that caused the threshold to be reached or exceeded. "
self.alert_id = 'THRESH01'
# Get the list of ras events that have threshold counts
self.msgIDs, self.msgidService, self.msgidCount, self.msgidPeriod = get_eventList()
for msgID in self.msgIDs:
registry.get_logger().debug('msgId = ' + msgID)
# define query for count with no period specified
eventTable = self.appendSchema('tbgqeventlog')
self.count_query = "select count(*) from " + eventTable + " where msg_id = ? and location LOC and serialnumber SN and event_time <= ?"
self.count_query2 = "select sum(bigint(count)) from " + eventTable + " where msg_id = ? and location LOC and serialnumber SN and event_time <= ?"
# define query for count exceeded with period specified
self.period_query = "select count(*) from " + eventTable + " where msg_id = ? and location LOC and serialnumber SN and event_time <= ? and event_time > (timestamp('MYTIME') - PERIOD)"
self.period_query2 = "select sum(bigint(count)) from " + eventTable + " where msg_id = ? and location LOC and serialnumber SN and event_time <= ? and event_time > (timestamp('MYTIME') - PERIOD)"
# the number of consecutive periods that the threshold has to exceed
self.msgidConsecutivePeriods = dict()
self.msgidConsecutivePeriods['0008002F'] = 7
self.msgidConsecutivePeriods['00080030'] = 3
return
def __init__(self, name, inQueue, outQueue, config_dict=None, number=0):
'''The constructor.'''
get_logger().debug('Creating GEAR event analyzer named {0}'.format(name))
self.engine = engine_factory(name, config_dict, event_input=True, number=number, send_alert=self.send_alert)
EventAnalyzer.__init__(self, name, inQueue, outQueue, config_dict, number, checkpoint=self.engine.checkpoint)
return
def __init__(self, name, inQueues, outQueue, config_dict=None, number=0):
'''The constructor.'''
get_logger().debug('metadata = %s', config_dict['events_metadata'])
# TODO: Should really parse by comma into list
#self.event_meta = EventsMetadata([config_dict['events_metadata']])
# Dummy Analysis info
self.ai1 = AnalysisInfo('location_test')
self.ai1.add_event_info('Example0', 1, 2, 3)
self.ai1.add_event_info('Example1', 3, 3, 5)
self.ai1.add_event_info('Example2', 3, None, None)
self.ai1.add_event_info('Example3', 3, None, None)
self.ai1.add_event_info('Example4', 3, None, None)
self.ai1.add_event_info('Example5', 3, None, None)
# TODO: Handle alert metadata
if 'mode' in config_dict:
self.mode = config_dict['mode']
else:
self.mode = POOL_MODE_OCCURRED
if 'initial_pool_duration' in config_dict:
self.duration = int(config_dict['initial_pool_duration'])
get_logger().debug('Duration override %s', str(self.duration))
else:
self.duration = 20
self.pool = IncidentPool.new_pool(self.mode, self.duration, None, self.close_callback)
# TODO: creation of alerts
self.count = 0 # start with 1
get_logger().debug('Creating SimpleEventAnalyzerAllAlert')
EventAnalyzer.__init__(self, name, inQueues, outQueue, config_dict=config_dict, number=number)
return
def __init__(self,
name,
inEventQueue,
outQueue,
config_dict=None,
number=0,
checkpoint=None):
'''The constructor.
'''
EventAnalyzer.__init__(self, name, inEventQueue, outQueue, config_dict,
number, checkpoint)
self.severity = "F"
self.recommendation = "Diagnose the fatal RAS event that ended the job. "
# ring buffer to remember jobs ending due to fatal errors
size = 1024
self.ring = [None for i in xrange(size)]
# Get the list of ras events with END_JOB control action
self.msgIDs, self.msgidService = get_eventList()
for msgID in self.msgIDs:
registry.get_logger().debug('msgId = ' + msgID)
return
def __init__(self, name, inEventQueue, outQueue, config_dict=None, number=0, checkpoint=None):
'''The constructor.
'''
EventAnalyzer.__init__(self, name, inEventQueue, outQueue, config_dict, number, checkpoint)
# Register to receive all ras events (note that a typical analyzer
# would be interested in a small subset of the events).
msgtypesTable = self.appendSchema('tbgqmsgtypes')
sample_query = "select msg_id from " + msgtypesTable
rows = self.executeQuery(sample_query)
self.msgIDs = list()
for r in rows:
#registry.get_logger().debug('msgId = ' + r[0])
self.msgIDs.append(r[0])
def __init__(self, name, inEventQueue, outQueue, config_dict=None, number=0, checkpoint=None):
'''The constructor.
'''
EventAnalyzer.__init__(self, name, inEventQueue, outQueue, config_dict, number, checkpoint)
self.severity = "F"
self.recommendation = "Diagnose hardware that has been placed in an error state. "
# Get the list of ras events with 'hardware in error' control actions
self.msgIDs, self.msgidService = get_eventList()
for msgID in self.msgIDs:
registry.get_logger().debug('msgId = ' + msgID)
return
def __init__(self, name, inEventQueue, outQueue, config_dict=None, number=0, checkpoint=None):
'''The constructor.
'''
EventAnalyzer.__init__(self, name, inEventQueue, outQueue, config_dict, number, checkpoint)
self.severity = "W"
self.recommendation = '''Schedule service to isolate the BQL issue. Possible causes are environmental, cable, or a board. Multiple BQLs reporting issues on multiple boards in the same midplane may be a side effect of an environmental issue like neighboring boards being powered off unexpectedly. Issues with a single cable may be due to poor seating of the cable into the connector or debris on the cable. The cable and board can be cleaned with an optics cleaning tool. Low voltage on all or most lanes on a single optical module may be an issue with the board at this location (the receiver) or its neighboring board (the transmitter) at the other end of the cable.'''
self.alert_id = 'BQL01'
# Get the exclude list of message IDs
cfg = registry.get_service(registry.SERVICE_CONFIGURATION)
excludeList = ''
try:
excludeList = cfg.get(BGQ_TEAL_BQL_ANALYZER, excludeMsgIDs)
registry.get_logger().debug('Exclude List = ' + excludeList)
except Exception, e:
registry.get_logger().debug(e)
def __init__(self, name, inEventQueue, outQueue, config_dict=None, number=0, checkpoint=None):
'''The constructor.
'''
EventAnalyzer.__init__(self, name, inEventQueue, outQueue, config_dict, number, checkpoint)
self.severity = "F"
self.recommendation = "Diagnose the fatal RAS event that ended the job. "
# ring buffer to remember jobs ending due to fatal errors
size = 1024
self.ring = [None for i in xrange(size)]
# Get the list of ras events with END_JOB control action
self.msgIDs, self.msgidService = get_eventList()
for msgID in self.msgIDs:
registry.get_logger().debug('msgId = ' + msgID)
return
def __init__(self,
name,
inEventQueue,
outQueue,
config_dict=None,
number=0,
checkpoint=None):
'''The constructor.
'''
EventAnalyzer.__init__(self, name, inEventQueue, outQueue, config_dict,
number, checkpoint)
self.severity = "W"
self.recommendation = "Diagnose the problem that caused the threshold to be reached or exceeded. "
self.alert_id = 'THRESH01'
# Get the list of ras events that have threshold counts
self.msgIDs, self.msgidService, self.msgidCount, self.msgidPeriod = get_eventList(
)
for msgID in self.msgIDs:
registry.get_logger().debug('msgId = ' + msgID)
# define query for count with no period specified
eventTable = self.appendSchema('tbgqeventlog')
self.count_query = "select count(*) from " + eventTable + " where msg_id = ? and location LOC and serialnumber SN and event_time <= ?"
self.count_query2 = "select sum(bigint(count)) from " + eventTable + " where msg_id = ? and location LOC and serialnumber SN and event_time <= ?"
# define query for count exceeded with period specified
self.period_query = "select count(*) from " + eventTable + " where msg_id = ? and location LOC and serialnumber SN and event_time <= ? and event_time > (timestamp('MYTIME') - PERIOD)"
self.period_query2 = "select sum(bigint(count)) from " + eventTable + " where msg_id = ? and location LOC and serialnumber SN and event_time <= ? and event_time > (timestamp('MYTIME') - PERIOD)"
# the number of consecutive periods that the threshold has to exceed
self.msgidConsecutivePeriods = dict()
self.msgidConsecutivePeriods['0008002F'] = 7
self.msgidConsecutivePeriods['00080030'] = 3
return
def is_not_processing(self):
''' Return true if the analyzer is processing an event or events '''
return EventAnalyzer.is_not_processing(self) and self.engine.is_not_processing_addition()
def __init__(self, name, inQueue, outQueue, config_dict=None, number=0):
'''The constructor.'''
self.count = 0 # start with 1
get_logger().debug('Creating SimpleEventAnalyzerAllAlert')
EventAnalyzer.__init__(self, name, inQueue, outQueue, config_dict=config_dict, number=number)
return
def __init__(self, name, inQueue, outQueue, config_dict=None, number=0):
self.journal = Journal(name)
EventAnalyzer.__init__(self, name, inQueue, outQueue, config_dict, number)
