def initialize(self):
self.graph = self.application.my_settings.get("graph")
self.session = self.application.my_settings.get("db_session")()
self._request_start_time = datetime.utcnow()
stats.incr("requests")
stats.incr("requests_{}".format(self.__class__.__name__))
def initialize(self):
self.graph = self.application.my_settings.get("graph")
self.session = self.application.my_settings.get("db_session")()
self._request_start_time = datetime.utcnow()
stats.incr("requests")
stats.incr("requests_{}".format(self.__class__.__name__))
def __call__(self, *args, **kwargs):
try:
self._call(*args, **kwargs)
# Prevent the application from crashing when callback raises
# an exception.
except Exception as err:
stats.incr("callback-failure", 1)
logging.exception("Callback Failed: %s", err)
def __call__(self, *args, **kwargs):
try:
self._call(*args, **kwargs)
# Prevent the application from crashing when callback raises
# an exception.
except Exception as err:
stats.incr("callback-failure", 1)
logging.exception("Callback Failed: %s", err)
def initialize(self):
self.session = self.application.my_settings.get("db_session")()
self.graph = Graph()
if self.get_argument("_profile", False):
self.perf_collector = Collector()
self.perf_trace_uuid = str(uuid4())
self.perf_collector.start()
else:
self.perf_collector = None
self.perf_trace_uuid = None
self._request_start_time = datetime.utcnow()
stats.incr("requests")
stats.incr("requests_{}".format(self.__class__.__name__))
def initialize(self):
self.session = self.application.my_settings.get("db_session")()
self.graph = Graph()
if self.get_argument("_profile", False):
self.perf_collector = Collector()
self.perf_trace_uuid = str(uuid4())
self.perf_collector.start()
else:
self.perf_collector = None
self.perf_trace_uuid = None
self._request_start_time = datetime.utcnow()
stats.incr("requests")
stats.incr("requests_{}".format(self.__class__.__name__))
def on_finish(self):
# log request duration
duration = datetime.utcnow() - self._request_start_time
duration_ms = int(duration.total_seconds() * 1000)
stats.incr("duration_ms", duration_ms)
stats.incr("duration_ms_{}".format(self.__class__.__name__), duration_ms)
# log response status code
response_status = self.get_status()
stats.incr("response_status_{}".format(response_status))
stats.incr("response_status_{}_{}".format(self.__class__.__name__, response_status))
def on_finish(self):
# log request duration
duration = datetime.utcnow() - self._request_start_time
duration_ms = int(duration.total_seconds() * 1000)
stats.incr("duration_ms", duration_ms)
stats.incr("duration_ms_{}".format(self.__class__.__name__),
duration_ms)
# log response status code
response_status = self.get_status()
stats.incr("response_status_{}".format(response_status))
stats.incr("response_status_{}_{}".format(self.__class__.__name__,
response_status))
def _send_mail(self, handler, trap, is_duplicate):
if is_duplicate and not handler["mail_on_duplicate"]:
return
mail = handler["mail"]
if not mail:
return
recipients = handler["mail"].get("recipients")
if not recipients:
return
subject = handler["mail"]["subject"] % {
"trap_oid": trap.oid,
"trap_name": ObjectId(trap.oid).name,
"ipaddress": trap.host,
"hostname": self.resolver.hostname_or_ip(trap.host),
}
ctxt = dict(trap=trap, dest_host=self.hostname)
try:
stats.incr("mail_sent_attempted", 1)
send_trap_email(recipients, "trapperkeeper",
subject, self.template_env, ctxt)
stats.incr("mail_sent_successful", 1)
except socket.error as err:
stats.incr("mail_sent_failed", 1)
logging.warning("Failed to send e-mail for trap: %s", err)
def _send_mail(self, handler, trap, is_duplicate):
if is_duplicate and not handler["mail_on_duplicate"]:
return
mail = handler["mail"]
if not mail:
return
recipients = handler["mail"].get("recipients")
if not recipients:
return
subject = handler["mail"]["subject"] % {
"trap_oid": trap.oid,
"trap_name": ObjectId(trap.oid).name,
"ipaddress": trap.host,
"hostname": self.resolver.hostname_or_ip(trap.host),
}
ctxt = dict(trap=trap, dest_host=self.hostname)
try:
stats.incr("mail_sent_attempted", 1)
send_trap_email(recipients, "trapperkeeper",
subject, self.template_env, ctxt)
stats.incr("mail_sent_successful", 1)
except socket.error as err:
stats.incr("mail_sent_failed", 1)
logging.warning("Failed to send e-mail for trap: %s", err)
def on_finish(self):
if self.perf_collector:
self.perf_collector.stop()
perf_profile.record_trace(self.session, self.perf_collector, self.perf_trace_uuid)
self.session.close()
# log request duration
duration = datetime.utcnow() - self._request_start_time
duration_ms = int(duration.total_seconds() * 1000)
stats.incr("duration_ms", duration_ms)
stats.incr("duration_ms_{}".format(self.__class__.__name__), duration_ms)
# log response status code
response_status = self.get_status()
stats.incr("response_status_{}".format(response_status))
stats.incr("response_status_{}_{}".format(self.__class__.__name__, response_status))
def on_finish(self):
if self.perf_collector:
self.perf_collector.stop()
perf_profile.record_trace(self.session, self.perf_collector,
self.perf_trace_uuid)
self.session.close()
# log request duration
duration = datetime.utcnow() - self._request_start_time
duration_ms = int(duration.total_seconds() * 1000)
stats.incr("duration_ms", duration_ms)
stats.incr("duration_ms_{}".format(self.__class__.__name__),
duration_ms)
# log response status code
response_status = self.get_status()
stats.incr("response_status_{}".format(response_status))
stats.incr("response_status_{}_{}".format(self.__class__.__name__,
response_status))
def initialize(self):
self.graph = self.application.my_settings.get("graph")
stats.incr("requests")
def _call(self, transport_dispatcher, transport_domain, transport_address, whole_msg):
if not whole_msg:
return
msg_version = int(api.decodeMessageVersion(whole_msg))
if msg_version in api.protoModules:
proto_module = api.protoModules[msg_version]
else:
stats.incr("unsupported-notification", 1)
logging.error("Unsupported SNMP version %s", msg_version)
return
host = transport_address[0]
version = SNMP_VERSIONS[msg_version]
try:
req_msg, whole_msg = decoder.decode(whole_msg, asn1Spec=proto_module.Message(),)
except (ProtocolError, ValueConstraintError) as err:
stats.incr("unsupported-notification", 1)
logging.warning("Failed to receive trap (%s) from %s: %s", version, host, err)
return
req_pdu = proto_module.apiMessage.getPDU(req_msg)
community = proto_module.apiMessage.getCommunity(req_msg)
if self.community and community != self.community:
stats.incr("unauthenticated-notification", 1)
logging.warning("Received trap from %s with invalid community: %s... discarding", host, community)
return
if not req_pdu.isSameTypeWith(proto_module.TrapPDU()):
stats.incr("unsupported-notification", 1)
logging.warning("Received non-trap notification from %s", host)
return
if msg_version not in (api.protoVersion1, api.protoVersion2c):
stats.incr("unsupported-notification", 1)
logging.warning("Received trap not in v1 or v2c")
return
trap = Notification.from_pdu(host, proto_module, version, req_pdu)
if trap is None:
stats.incr("unsupported-notification", 1)
logging.warning("Invalid trap from %s: %s", host, req_pdu)
return
dde = DdeNotification(trap, self.config.handlers[trap.oid])
dde_run(dde)
handler = dde.handler
trap.severity = handler["severity"]
trap.manager = self.hostname
if handler.get("expiration", None):
expires = parse_time_string(handler["expiration"])
expires = timedelta(**expires)
trap.expires = trap.sent + expires
stats.incr("traps_received", 1)
objid = ObjectId(trap.oid)
if handler.get("blackhole", False):
stats.incr("traps_blackholed", 1)
logging.debug("Blackholed %s from %s", objid.name, host)
return
logging.info("Trap Received (%s) from %s", objid.name, host)
stats.incr("traps_accepted", 1)
duplicate = False
try:
stats.incr("db_write_attempted", 1)
self.conn.add(trap)
self.conn.commit()
stats.incr("db_write_successful", 1)
except OperationalError as err:
self.conn.rollback()
logging.warning("Failed to commit: %s", err)
stats.incr("db_write_failed", 1)
# TODO(gary) reread config and reconnect to database
except InvalidRequestError as err:
# If we get into this state we should rollback any pending changes.
stats.incr("db_write_failed", 1)
self.conn.rollback()
logging.warning("Bad state, rolling back transaction: %s", err)
except IntegrityError as err:
stats.incr("db_write_duplicate", 1)
duplicate = True
self.conn.rollback()
logging.info("Duplicate Trap (%s) from %s. Likely inserted by another manager.", objid.name, host)
logging.debug(err)
self._send_mail(handler, trap, duplicate)
def _call(self, transport_dispatcher, transport_domain, transport_address, whole_msg):
if not whole_msg:
return
msg_version = int(api.decodeMessageVersion(whole_msg))
if msg_version in api.protoModules:
proto_module = api.protoModules[msg_version]
else:
stats.incr("unsupported-notification", 1)
logging.error("Unsupported SNMP version %s", msg_version)
return
host = transport_address[0]
version = SNMP_VERSIONS[msg_version]
try:
req_msg, whole_msg = decoder.decode(whole_msg, asn1Spec=proto_module.Message(),)
except (ProtocolError, ValueConstraintError) as err:
stats.incr("unsupported-notification", 1)
logging.warning("Failed to receive trap (%s) from %s: %s", version, host, err)
return
req_pdu = proto_module.apiMessage.getPDU(req_msg)
# TODO(gary): Require matching community string or bail.
# community = proto_module.apiMessage.getCommunity(req_msg)
if not req_pdu.isSameTypeWith(proto_module.TrapPDU()):
stats.incr("unsupported-notification", 1)
logging.warning("Received non-trap notification from %s", host)
return
if msg_version not in (api.protoVersion1, api.protoVersion2c):
stats.incr("unsupported-notification", 1)
logging.warning("Received trap not in v1 or v2c")
return
trap = Notification.from_pdu(host, proto_module, version, req_pdu)
if trap is None:
stats.incr("unsupported-notification", 1)
logging.warning("Invalid trap from %s: %s", host, req_pdu)
return
dde = DdeNotification(trap, self.config.handlers[trap.oid])
dde_run(dde)
handler = dde.handler
trap.severity = handler["severity"]
trap.manager = self.hostname
if handler.get("expiration", None):
expires = parse_time_string(handler["expiration"])
expires = timedelta(**expires)
trap.expires = trap.sent + expires
stats.incr("traps_received", 1)
objid = ObjectId(trap.oid)
if handler.get("blackhole", False):
stats.incr("traps_blackholed", 1)
logging.debug("Blackholed %s from %s", objid.name, host)
return
logging.info("Trap Received (%s) from %s", objid.name, host)
stats.incr("traps_accepted", 1)
duplicate = False
try:
stats.incr("db_write_attempted", 1)
self.conn.add(trap)
self.conn.commit()
stats.incr("db_write_successful", 1)
except OperationalError as err:
self.conn.rollback()
logging.warning("Failed to commit: %s", err)
stats.incr("db_write_failed", 1)
# TODO(gary) reread config and reconnect to database
except InvalidRequestError as err:
# If we get into this state we should rollback any pending changes.
stats.incr("db_write_failed", 1)
self.conn.rollback()
logging.warning("Bad state, rolling back transaction: %s", err)
except IntegrityError as err:
stats.incr("db_write_duplicate", 1)
duplicate = True
self.conn.rollback()
logging.info("Duplicate Trap (%s) from %s. Likely inserted by another manager.", objid.name, host)
logging.debug(err)
self._send_mail(handler, trap, duplicate)
def initialize(self):
self.session = self.application.my_settings.get("db_session")()
self.graph = Graph()
stats.incr("requests")
def initialize(self):
self.graph = self.application.my_settings.get("graph")
stats.incr("requests")
def initialize(self):
self.session = self.application.my_settings.get("db_session")()
stats.incr("requests")
