def mainloop(job_config, live, histou, manager):
while True:
try:
l_start = time.time()
try:
host_services = live.get_host_services()
except Exception as e:
logging.getLogger(__name__).error(
"Could not connect to livestatus: " + str(e))
time.sleep(10)
continue
if log_peformance():
logging.getLogger(__name__).debug(
"Livestatusquery took: %dms" % delta_ms(l_start))
for host, hv in host_services.items():
config_names = []
for service, sv in hv.items():
config_names.append({
"host": host,
"service": service,
"command": sv['command'],
"perf_labels": sv['perf_labels']
})
host_service_config = None
l_start = time.time()
host_service_config = histou.get_config(
hosts_services=config_names)
if host_service_config:
if log_peformance():
logging.getLogger(__name__).debug(
"Histouquery took: %dms" % delta_ms(l_start))
i = 0
for service, sv in hv.items():
if host_service_config[i]:
if host not in job_config:
job_config[host] = {}
sv['config'] = host_service_config[i]
job_config[host][service] = sv
i += 1
manager.update_config(job_config)
return job_config
except (URLError, ValueError) as e:
logging.getLogger(__name__).error(
"Could not connect to histou, to receive the config: " +
str(e))
time.sleep(10)
except Exception as e:
logging.getLogger(__name__).error(str(e), exc_info=True)
time.sleep(10)
def forecast_mode(time_for_calculations=10):
if Config.data['livestatus']['protocol'] == "tcp":
l = livestatus.Connector(ip=(Config.data['livestatus']['address'],
Config.data['livestatus']['port']))
elif Config.data['livestatus']['protocol'] == "unix":
l = livestatus.Connector(soc=(Config.data['livestatus']['address']))
else:
sys.exit("Livestatus protocol is unkown: " +
Config.data['livestatus']['protocol'] +
". Allowed are tcp/unix")
h = Histou(protocol=Config.data['histou']['prot'],
address=Config.data['histou']['address'])
m = JobManager()
config_host_service = {}
if Config.data['main']['daemon']:
try:
while True:
start = time.time()
config_host_service = mainloop(job_config=config_host_service,
live=l,
histou=h,
manager=m)
if log_peformance():
logging.getLogger(__name__).debug(
"Updating config took: %dms" % delta_ms(start))
time.sleep(Config.data['main']['update_rate'])
except Exception as e:
logging.getLogger(__name__).fatal(str(e), exc_info=True)
else:
logging.getLogger(__name__).info(
"Running just once, and wait %ds to finish calculations" %
time_for_calculations)
start = time.time()
mainloop(job_config=config_host_service, live=l, histou=h, manager=m)
if log_peformance():
logging.getLogger(__name__).debug("One cycle took: %dms" %
delta_ms(start))
time.sleep(time_for_calculations)
m.stop()
def start_calculation(self, conf):
start = time.time()
lookback_data = InfluxDBReader.request_past(host=self.__host,
service=self.__service,
# command=self.__command,
performance_label=conf[LABEL],
lookback=conf[LOOKBACK_RANGE])
if not lookback_data:
return
if log_peformance():
logging.getLogger(__name__).debug(
"Fetching data of %s %s %s: %s took %dms" % (
self.__host, self.__service, conf[LABEL], self.get_method(conf), delta_ms(start))
)
start = time.time()
my_class = MethodCollector.classes[self.get_method(conf)]
if 'calc_forecast' in dir(my_class):
forecast_data = my_class. \
calc_forecast(options=conf[METHOD_OPTIONS],
forecast_start=self.calc_start_date(lookback_data[len(lookback_data) - 1][0],
conf[FORECAST_INTERVAL]),
forecast_range=conf[FORECAST_RANGE],
forecast_interval=conf[FORECAST_INTERVAL],
lookback_range=conf[LOOKBACK_RANGE],
lookback_data=lookback_data)
if log_peformance():
logging.getLogger(__name__).debug(
"Calculation data of %s %s %s: %s took %dms" % (
self.__host, self.__service, conf[LABEL], self.get_method(conf), delta_ms(start))
)
start = time.time()
if forecast_data:
InfluxDBWriter.write_forecast(data=forecast_data,
host=self.__host,
service=self.__service,
# command=self.__command,
performance_label=conf[LABEL])
if log_peformance():
logging.getLogger(__name__).debug(
"Writing data of %s %s %s: %s took %dms" % (
self.__host, self.__service, conf[LABEL], self.get_method(conf), delta_ms(start))
)
else:
logging.getLogger(__name__).debug(
"Calculation did not return any data: %s %s %s: %s" % (
self.__host, self.__service, conf[LABEL], self.get_method(conf))
)
elif 'search_anomaly' in dir(my_class):
anomaly_data = my_class.search_anomaly(
options=conf[METHOD_OPTIONS],
lookback_range=conf[LOOKBACK_RANGE],
lookback_data=lookback_data)
if log_peformance():
logging.getLogger(__name__).debug(
"Calculation data of %s %s %s: %s took %dms" % (
self.__host, self.__service, conf[LABEL], self.get_method(conf), delta_ms(start))
)
if anomaly_data:
InfluxDBWriter.write_anomaly(data=anomaly_data,
host=self.__host,
service=self.__service,
# command=self.__command,
performance_label=conf[LABEL])
if log_peformance():
logging.getLogger(__name__).debug(
"Writing data of %s %s %s: %s took %dms" % (
self.__host, self.__service, conf[LABEL], self.get_method(conf), delta_ms(start))
)
else:
logging.getLogger(__name__).debug(
"Calculation did not return any data: %s %s %s: %s" % (
self.__host, self.__service, conf[LABEL], self.get_method(conf))
)