class SnmpTrapHandler(object):
def __init__(self):
self.api = None
def run(self):
endpoint = os.environ.get('ALERTA_ENDPOINT', 'http://localhost:8080')
key = os.environ.get('ALERTA_API_KEY', None)
self.api = ApiClient(endpoint=endpoint, key=key)
data = sys.stdin.read()
LOG.info('snmptrapd -> %r', data)
data = unicode(data, 'utf-8', errors='ignore')
LOG.debug('unicoded -> %s', data)
snmptrapAlert = SnmpTrapHandler.parse_snmptrap(data)
if snmptrapAlert:
try:
self.api.send(snmptrapAlert)
except Exception, e:
LOG.warning('Failed to send alert: %s', e)
LOG.debug('Send heartbeat...')
heartbeat = Heartbeat(tags=[__version__])
try:
self.api.send(heartbeat)
except Exception, e:
LOG.warning('Failed to send heartbeat: %s', e)
def run(self):
api = ApiClient(endpoint=OPTIONS['endpoint'], key=OPTIONS['key'])
keep_alive = 0
while not self.should_stop:
for alertid in on_hold.keys():
try:
(alert, hold_time) = on_hold[alertid]
except KeyError:
continue
if time.time() > hold_time:
self.send_email(alert)
try:
del on_hold[alertid]
except KeyError:
continue
if keep_alive >= 10:
tag = OPTIONS['smtp_host'] or 'alerta-mailer'
try:
api.send(Heartbeat(tags=[tag]))
except Exception as e:
time.sleep(5)
continue
keep_alive = 0
keep_alive += 1
time.sleep(2)
def run(self):
api = ApiClient(endpoint=OPTIONS['endpoint'], key=OPTIONS['key'])
keep_alive = 0
while not self.should_stop:
for alertid in on_hold.keys():
try:
(alert, hold_time) = on_hold[alertid]
except KeyError:
continue
if time.time() > hold_time:
self.send_email(alert)
try:
del on_hold[alertid]
except KeyError:
continue
if keep_alive >= 10:
tag = OPTIONS['smtp_host'] or 'alerta-mailer'
try:
api.send(Heartbeat(tags=[tag]))
except Exception as e:
time.sleep(5)
continue
keep_alive = 0
keep_alive += 1
time.sleep(2)
def __init__(self, endpoint, key):
self.api = ApiClient(endpoint=endpoint, key=key)
self.version = ''
self.last_metrics = dict()
self.app_last_time = None
self.rate_metrics = dict()
self.latency_metrics = dict()
self.alerts = []
self.total = 0
self.status = ''
self.groupby = None
self.top10 = []
self.last_time = datetime.utcnow()
# self.prev_time = None
# self.diff_time = 0
# self.running_total = 0
# self.start_time = time.time()
# self.max_rate = 0.0
# self.latency = 0
# self.running_latency = 0
#
# self.dupl_cache = dict()
# self.dupl_total = 0
# self.running_dupl_total = 0
self.running = True
threading.Thread.__init__(self)
class SnmpTrapHandler(object):
def __init__(self):
self.api = None
def run(self):
endpoint = os.environ.get('ALERTA_ENDPOINT', 'http://localhost:8080')
key = os.environ.get('ALERTA_API_KEY', None)
self.api = ApiClient(endpoint=endpoint, key=key)
data = sys.stdin.read()
LOG.info('snmptrapd -> %r', data)
data = unicode(data, 'utf-8', errors='ignore')
LOG.debug('unicoded -> %s', data)
snmptrapAlert = SnmpTrapHandler.parse_snmptrap(data)
if snmptrapAlert:
try:
self.api.send(snmptrapAlert)
except Exception, e:
LOG.warning('Failed to send alert: %s', e)
LOG.debug('Send heartbeat...')
heartbeat = Heartbeat(tags=[__version__])
try:
self.api.send(heartbeat)
except Exception, e:
LOG.warning('Failed to send heartbeat: %s', e)
def main():
api = ApiClient()
listener = Listener()
while True:
listener.send_cmd('READY\n')
headers, body = listener.wait()
event = headers['eventname']
if event.startswith('TICK'):
supervisorAlert = Heartbeat(
origin='supervisord',
tags=[headers['ver'], event]
)
else:
if event.endswith('FATAL'):
severity = 'critical'
elif event.endswith('BACKOFF'):
severity = 'warning'
elif event.endswith('EXITED'):
severity = 'minor'
else:
severity = 'normal'
supervisorAlert = Alert(
resource='%s:%s' % (platform.uname()[1], body['processname']),
environment='Production',
service=['supervisord'],
event=event,
correlate=[
'PROCESS_STATE_STARTING',
'PROCESS_STATE_RUNNING',
'PROCESS_STATE_BACKOFF',
'PROCESS_STATE_STOPPING',
'PROCESS_STATE_EXITED',
'PROCESS_STATE_STOPPED',
'PROCESS_STATE_FATAL',
'PROCESS_STATE_UNKNOWN'
],
value='serial=%s' % headers['serial'],
severity=severity,
origin=headers['server'],
text='State changed from %s to %s.' % (body['from_state'], event),
raw_data='%s\n\n%s' % (json.dumps(headers), json.dumps(body))
)
try:
api.send(supervisorAlert)
except Exception as e:
listener.log_stderr(e)
listener.send_cmd('RESULT 4\nFAIL')
else:
listener.send_cmd('RESULT 2\nOK')
def on_callback_query(self, msg):
query_id, from_id, query_data = telepot.glance(msg, flavor='callback_query')
for message in sent_messages:
if msg['message']['message_id'] == message['message_id']:
sent_messages.remove(message)
self.editor.editMessageReplyMarkup()
self.bot.sendMessage(self.id[0], "Thanks... !")
api = ApiClient(endpoint=config.ALERTA_API_URL, key=config.ALERTA_API_KEY)
text = "status change via API (Telegram) by {}.".format(message['telegram'])
api.update_status(alertid=message['data']['id'], status = query_data, text=text)
break
else:
self.editor.editMessageReplyMarkup()
self.bot.sendMessage(self.id[0], "Oh Sorry, You were too late. You can not ACK this now.")
self.close()
def main():
api = ApiClient()
listener = Listener()
while True:
listener.send_cmd('READY\n')
headers, body = listener.wait()
event = headers['eventname']
if event.startswith('TICK'):
supervisorAlert = Heartbeat(origin='supervisord',
tags=[headers['ver'], event])
else:
if event.endswith('FATAL'):
severity = 'critical'
elif event.endswith('BACKOFF'):
severity = 'warning'
elif event.endswith('EXITED'):
severity = 'minor'
else:
severity = 'normal'
supervisorAlert = Alert(
resource='%s:%s' % (platform.uname()[1], body['processname']),
environment='Production',
service=['supervisord'],
event=event,
correlate=[
'PROCESS_STATE_STARTING', 'PROCESS_STATE_RUNNING',
'PROCESS_STATE_BACKOFF', 'PROCESS_STATE_STOPPING',
'PROCESS_STATE_EXITED', 'PROCESS_STATE_STOPPED',
'PROCESS_STATE_FATAL', 'PROCESS_STATE_UNKNOWN'
],
value='serial=%s' % headers['serial'],
severity=severity,
origin=headers['server'],
text='State changed from %s to %s.' %
(body['from_state'], event),
raw_data='%s\n\n%s' % (json.dumps(headers), json.dumps(body)))
try:
api.send(supervisorAlert)
except Exception as e:
listener.log_stderr(e)
listener.send_cmd('RESULT 4\nFAIL')
else:
listener.send_cmd('RESULT 2\nOK')
def alert(self, matches):
# Matches is a list of match dictionaries.
# It contains more than one match when the alert has
# the aggregation option set
# for match in matches:
api = ApiClient(endpoint=self.endpoint, key=self.api_key)
alert = Alert(resource=self.resource,
event=self.event,
environment=self.environment,
service=self.service,
severity=self.severity,
text=self.text,
value=self.value)
api.send(alert)
def run(self):
endpoint = os.environ.get('ALERTA_ENDPOINT', 'http://localhost:8080')
key = os.environ.get('ALERTA_API_KEY', None)
self.api = ApiClient(endpoint=endpoint, key=key)
data = sys.stdin.read()
LOG.info('snmptrapd -> %r', data)
data = unicode(data, 'utf-8', errors='ignore')
LOG.debug('unicoded -> %s', data)
snmptrapAlert = SnmpTrapHandler.parse_snmptrap(data)
if snmptrapAlert:
try:
self.api.send(snmptrapAlert)
except Exception, e:
LOG.warning('Failed to send alert: %s', e)
def __init__(self):
self.api = ApiClient()
try:
connection = boto.sqs.connect_to_region(
AWS_REGION,
aws_access_key_id=AWS_ACCESS_KEY_ID,
aws_secret_access_key=AWS_SECRET_ACCESS_KEY)
except boto.exception.SQSError as e:
LOG.error('SQS API call failed: %s', e)
sys.exit(1)
try:
self.sqs = connection.create_queue(AWS_SQS_QUEUE)
self.sqs.set_message_class(RawMessage)
except boto.exception.SQSError as e:
LOG.error('SQS queue error: %s', e)
sys.exit(1)
def __init__(self):
self.api = ApiClient()
LOG.info('Starting UDP listener...')
# Set up syslog UDP listener
try:
self.udp = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
self.udp.bind(('', SYSLOG_UDP_PORT))
except socket.error, e:
LOG.error('Syslog UDP error: %s', e)
sys.exit(2)
def run(self):
api = ApiClient(endpoint=OPTIONS['endpoint'], key=OPTIONS['key'])
keep_alive = 0
while not self.should_stop:
for alertid in on_hold.keys():
try:
(alert, hold_time) = on_hold[alertid]
except KeyError:
continue
if time.time() > hold_time:
self.send_email(alert)
try:
del on_hold[alertid]
except KeyError:
continue
def run(self):
self.running = True
self.queue = Queue.Queue()
self.api = self.api = ApiClient(endpoint=settings.ENDPOINT,
key=settings.API_KEY)
# Start worker threads
LOG.debug('Starting %s worker threads...', SERVER_THREADS)
for i in range(SERVER_THREADS):
w = WorkerThread(self.queue, self.api)
try:
w.start()
except Exception, e:
LOG.error('Worker thread #%s did not start: %s', i, e)
continue
LOG.info('Started worker thread: %s', w.getName())
def run(self):
endpoint = os.environ.get('ALERTA_ENDPOINT', 'http://localhost:8080')
key = os.environ.get('ALERTA_API_KEY', None)
self.api = ApiClient(endpoint=endpoint, key=key)
data = sys.stdin.read()
LOG.info('snmptrapd -> %r', data)
data = unicode(data, 'utf-8', errors='ignore')
LOG.debug('unicoded -> %s', data)
snmptrapAlert = SnmpTrapHandler.parse_snmptrap(data)
if snmptrapAlert:
try:
self.api.send(snmptrapAlert)
except Exception, e:
LOG.warning('Failed to send alert: %s', e)
def run(self):
self.running = True
# Create internal queue
self.queue = Queue.Queue()
self.api = ApiClient()
# Initialiase ping targets
ping_list = init_targets()
# Start worker threads
LOG.debug('Starting %s worker threads...', SERVER_THREAD_COUNT)
for i in range(SERVER_THREAD_COUNT):
w = WorkerThread(self.api, self.queue)
try:
w.start()
except Exception, e:
LOG.error('Worker thread #%s did not start: %s', i, e)
continue
LOG.info('Started worker thread: %s', w.getName())
class UpdateThread(threading.Thread):
def __init__(self, endpoint, key):
self.api = ApiClient(endpoint=endpoint, key=key)
self.version = ''
self.last_metrics = dict()
self.app_last_time = None
self.rate_metrics = dict()
self.latency_metrics = dict()
self.alerts = []
self.total = 0
self.status = ''
self.groupby = None
self.top10 = []
self.last_time = datetime.utcnow()
# self.prev_time = None
# self.diff_time = 0
# self.running_total = 0
# self.start_time = time.time()
# self.max_rate = 0.0
# self.latency = 0
# self.running_latency = 0
#
# self.dupl_cache = dict()
# self.dupl_total = 0
# self.running_dupl_total = 0
self.running = True
threading.Thread.__init__(self)
def run(self):
self.running = True
while self.running:
self.status = 'updating...'
self._update()
try:
time.sleep(2)
except (KeyboardInterrupt, SystemExit):
sys.exit(0)
def set_groupby(self, groupby):
self.groupby = groupby
def get_groupby(self):
return self.groupby
def _update(self):
try:
status = self.api.get_status()
except ConnectionError:
return ''
self.version = 'v' + status['version']
app_time = status['time'] / 1000 # epoch ms
metrics = [
metric for metric in status['metrics'] if metric['type'] == 'timer'
]
for m in metrics:
count = m['count']
totalTime = m['totalTime']
if m['name'] in self.last_metrics:
(last_count, last_totalTime) = self.last_metrics[m['name']]
else:
self.last_metrics[m['name']] = (count, totalTime)
self.app_last_time = app_time
continue
if count >= last_count:
num = count - last_count
period = app_time - self.app_last_time
try:
rate = 1.0 * num / period
except ZeroDivisionError:
rate = 0.0
self.rate_metrics[m['name']] = rate
if totalTime >= last_totalTime:
diff = totalTime - last_totalTime
num = count - last_count
try:
latency = 1.0 * diff / num
except ZeroDivisionError:
latency = 0.0
self.latency_metrics[m['name']] = latency
self.last_metrics[m['name']] = (count, totalTime)
self.app_last_time = app_time
try:
response = self.api.get_alerts()
except ConnectionError:
return ''
with lock:
self.alerts = response.get('alerts', [])
self.total = response.get('total', 0)
self.status = '%s - %s' % (response['status'],
response.get('message', 'no errors'))
try:
response = self.api.get_top10(self.groupby)
except ConnectionError:
return ''
with lock:
self.top10 = response.get('top10', [])
j = json.load(sys.stdin)
print "Request:"
print j
url = client.kv.get('alerta/apiurl')[1]['Value']
key = client.kv.get('alerta/apikey')[1]['Value']
max_retries = int(client.kv.get('alerta/max_retries')[1]['Value'])
sleep = int(client.kv.get('alerta/sleep')[1]['Value'])
timeout = int(client.kv.get('alerta/timeout')[1]['Value'])
origin = client.kv.get('alerta/origin')[1]['Value']
alerttype = client.kv.get('alerta/alerttype')[1]['Value']
api = ApiClient(endpoint=url, key=key)
SEVERITY_MAP = {
'critical': 'critical',
'warning': 'warning',
'passing': 'ok',
}
def createalert(data):
try:
environment = client.kv.get('alerta/env/{0}'.format(
data['Node']))[1]['Value']
except:
environment = client.kv.get('alerta/defaultenv')[1]['Value']
alert = Alert(resource=data['Node'],
#!/usr/bin/env python
from alertaclient.api import ApiClient
from alertaclient.alert import Alert
api = ApiClient()
alert = Alert(resource='web-server-01',
event='HttpError',
correlate=['HttpOK'],
group='Web',
environment='Production',
service=['theguardian.com'],
severity='major',
value='Bad Gateway (502)',
text='Web server error.',
tags=['web', 'dc1', 'london'],
attributes={'customer': 'The Guardian'})
print alert
try:
print api.send(alert)
except Exception as e:
print e
class CloudWatch(object):
def __init__(self):
self.api = ApiClient()
try:
connection = boto.sqs.connect_to_region(
AWS_REGION,
aws_access_key_id=AWS_ACCESS_KEY_ID,
aws_secret_access_key=AWS_SECRET_ACCESS_KEY)
except boto.exception.SQSError as e:
LOG.error('SQS API call failed: %s', e)
sys.exit(1)
try:
self.sqs = connection.create_queue(AWS_SQS_QUEUE)
self.sqs.set_message_class(RawMessage)
except boto.exception.SQSError as e:
LOG.error('SQS queue error: %s', e)
sys.exit(1)
def run(self):
while True:
LOG.debug('Waiting for CloudWatch alarms on %s...', AWS_SQS_QUEUE)
try:
notification = self.sqs.read(wait_time_seconds=20)
except boto.exception.SQSError as e:
LOG.warning('Could not read from queue: %s', e)
time.sleep(20)
continue
if notification:
cloudwatchAlert = self.parse_notification(notification)
try:
self.api.send(cloudwatchAlert)
except Exception as e:
LOG.warning('Failed to send alert: %s', e)
self.sqs.delete_message(notification)
LOG.debug('Send heartbeat...')
heartbeat = Heartbeat(tags=[__version__])
try:
self.api.send(heartbeat)
except Exception as e:
LOG.warning('Failed to send heartbeat: %s', e)
def parse_notification(self, notification):
notification = json.loads(notification.get_body())
alarm = json.loads(notification['Message'])
if 'Trigger' not in alarm:
return
# Defaults
resource = '%s:%s' % (alarm['Trigger']['Dimensions'][0]['name'],
alarm['Trigger']['Dimensions'][0]['value'])
event = alarm['AlarmName']
severity = self.cw_state_to_severity(alarm['NewStateValue'])
group = 'CloudWatch'
value = alarm['Trigger']['MetricName']
text = alarm['AlarmDescription']
service = [
AWS_ACCOUNT_ID.get(alarm['AWSAccountId'],
'AWSAccountId:' + alarm['AWSAccountId'])
]
tags = [alarm['Trigger']['Namespace']]
correlate = list()
origin = notification['TopicArn']
timeout = None
create_time = datetime.datetime.strptime(notification['Timestamp'],
'%Y-%m-%dT%H:%M:%S.%fZ')
raw_data = notification['Message']
cloudwatchAlert = Alert(
resource=resource,
event=event,
correlate=correlate,
group=group,
value=value,
severity=severity,
environment='Production',
service=service,
text=text,
event_type='cloudwatchAlarm',
tags=tags,
attributes={
'awsMessageId': notification['MessageId'],
'awsRegion': alarm['Region'],
'thresholdInfo': alarm['NewStateReason']
},
origin=origin,
timeout=timeout,
create_time=create_time,
raw_data=raw_data,
)
return cloudwatchAlert
@staticmethod
def cw_state_to_severity(state):
if state == 'ALARM':
return 'major'
elif state == 'INSUFFICIENT_DATA':
return 'warning'
elif state == 'OK':
return 'normal'
else:
return 'unknown'
def main():
config_file = os.environ.get('ALERTA_CONF_FILE') or OPTIONS['config_file']
config = configparser.RawConfigParser(defaults=OPTIONS)
try:
config.read(os.path.expanduser(config_file))
except Exception:
sys.exit("Problem reading configuration file %s - is this an ini file?" % config_file)
parser = argparse.ArgumentParser(
prog='zabbix-alerta',
usage='zabbix-alerta SENDTO SUMMARY BODY',
description='Zabbix-to-Alerta integration script',
epilog=epilog,
formatter_class=argparse.RawTextHelpFormatter
)
parser.add_argument(
'sendto',
help='config profile or alerta API endpoint and key'
)
parser.add_argument(
'summary',
help='alert summary'
)
parser.add_argument(
'body',
help='alert body (see format below)'
)
args, left = parser.parse_known_args()
# sendto=apiUrl[;key]
if args.sendto.startswith('http://') or args.sendto.startswith('https://'):
want_profile = None
try:
OPTIONS['endpoint'], OPTIONS['key'] = args.sendto.split(';', 1)
except ValueError:
OPTIONS['endpoint'] = args.sendto
# sendto=profile
else:
want_profile = args.sendto or os.environ.get('ALERTA_DEFAULT_PROFILE') or config.defaults().get('profile')
if want_profile and config.has_section('profile %s' % want_profile):
for opt in OPTIONS:
try:
OPTIONS[opt] = config.getboolean('profile %s' % want_profile, opt)
except (ValueError, AttributeError):
OPTIONS[opt] = config.get('profile %s' % want_profile, opt)
else:
for opt in OPTIONS:
try:
OPTIONS[opt] = config.getboolean('DEFAULT', opt)
except (ValueError, AttributeError):
OPTIONS[opt] = config.get('DEFAULT', opt)
parser.set_defaults(**OPTIONS)
args = parser.parse_args()
if args.debug:
LOG.basicConfig(stream=sys.stderr, format=LOG_FORMAT, datefmt=LOG_DATE_FMT, level=LOG.DEBUG)
else:
LOG.basicConfig(filename=LOG_FILE, format=LOG_FORMAT, datefmt=LOG_DATE_FMT, level=LOG.INFO)
LOG.info("[alerta] endpoint=%s key=%s", args.endpoint, args.key)
api = ApiClient(endpoint=args.endpoint, key=args.key, ssl_verify=args.sslverify)
LOG.debug("[alerta] sendto=%s, summary=%s, body=%s", args.sendto, args.summary, args.body)
try:
alert = parse_zabbix(args.summary, args.body)
api.send(alert)
except (SystemExit, KeyboardInterrupt):
LOG.warning("Exiting zabbix-alerta.")
sys.exit(0)
except Exception as e:
LOG.error(e, exc_info=1)
sys.exit(1)
