Bugfixes and better error handling on zabbix alerter.

elastalert/zabbix.py

@@ -1,46 +1,47 @@
-from alerts import Alerter  # , BasicMatchString
-import logging
-from pyzabbix.api import ZabbixAPI
-from pyzabbix import ZabbixSender, ZabbixMetric
 from datetime import datetime
 
+from pyzabbix import ZabbixSender, ZabbixMetric, ZabbixAPI
+
+from .alerts import Alerter
+from .util import elastalert_logger, EAException
+
 
 class ZabbixClient(ZabbixAPI):
 
-    def __init__(self, url='http://localhost', use_authenticate=False, user='Admin', password='zabbix', sender_host='localhost',
-                 sender_port=10051):
+    def __init__(self, url='http://localhost', use_authenticate=False, user='Admin', password='zabbix',
+                 sender_host='localhost', sender_port=10051):
         self.url = url
         self.use_authenticate = use_authenticate
         self.sender_host = sender_host
         self.sender_port = sender_port
         self.metrics_chunk_size = 200
         self.aggregated_metrics = []
-        self.logger = logging.getLogger(self.__class__.__name__)
-        super(ZabbixClient, self).__init__(url=self.url, use_authenticate=self.use_authenticate, user=user, password=password)
+
+        super(ZabbixClient, self).__init__(url=self.url,
+                                           use_authenticate=self.use_authenticate,
+                                           user=user,
+                                           password=password)
 
     def send_metric(self, hostname, key, data):
         zm = ZabbixMetric(hostname, key, data)
         if self.send_aggregated_metrics:
-
             self.aggregated_metrics.append(zm)
             if len(self.aggregated_metrics) > self.metrics_chunk_size:
-                self.logger.info("Sending: %s metrics" % (len(self.aggregated_metrics)))
+                elastalert_logger.info("Sending: %s metrics" % (len(self.aggregated_metrics)))
                 try:
-                    ZabbixSender(zabbix_server=self.sender_host, zabbix_port=self.sender_port).send(self.aggregated_metrics)
+                    ZabbixSender(zabbix_server=self.sender_host, zabbix_port=self.sender_port) \
+                        .send(self.aggregated_metrics)
                     self.aggregated_metrics = []
                 except Exception as e:
-                    self.logger.exception(e)
-                    pass
+                    elastalert_logger.exception(e)
         else:
             try:
-                ZabbixSender(zabbix_server=self.sender_host, zabbix_port=self.sender_port).send(zm)
+                ZabbixSender(zabbix_server=self.sender_host, zabbix_port=self.sender_port).send([zm])
             except Exception as e:
-                self.logger.exception(e)
-                pass
+                elastalert_logger.exception(e)
 
 
 class ZabbixAlerter(Alerter):
-
     # By setting required_options to a set of strings
     # You can ensure that the rule config file specifies all
     # of the options. Otherwise, ElastAlert will throw an exception
@@ -54,6 +55,9 @@ def __init__(self, *args):
         self.zbx_sender_port = self.rule.get('zbx_sender_port', 10051)
         self.zbx_host = self.rule.get('zbx_host')
         self.zbx_key = self.rule.get('zbx_key')
+        self.timestamp_field = self.rule.get('timestamp_field', '@timestamp')
+        self.timestamp_type = self.rule.get('timestamp_type', 'iso')
+        self.timestamp_strptime = self.rule.get('timestamp_strptime', '%Y-%m-%dT%H:%M:%S.%fZ')
 
     # Alert is called
     def alert(self, matches):
@@ -63,10 +67,26 @@ def alert(self, matches):
         # the aggregation option set
         zm = []
         for match in matches:
-            ts_epoch = int(datetime.strptime(match['@timestamp'], "%Y-%m-%dT%H:%M:%S.%fZ").strftime('%s'))
-            zm.append(ZabbixMetric(host=self.zbx_host, key=self.zbx_key, value=1, clock=ts_epoch))
+            if ':' not in match[self.timestamp_field] or '-' not in match[self.timestamp_field]:
+                ts_epoch = int(match[self.timestamp_field])
+            else:
+                try:
+                    ts_epoch = int(datetime.strptime(match[self.timestamp_field], self.timestamp_strptime)
+                                   .strftime('%s'))
+                except ValueError:
+                    ts_epoch = int(datetime.strptime(match[self.timestamp_field], '%Y-%m-%dT%H:%M:%SZ')
+                                   .strftime('%s'))
+            zm.append(ZabbixMetric(host=self.zbx_host, key=self.zbx_key, value='1', clock=ts_epoch))
 
-        ZabbixSender(zabbix_server=self.zbx_sender_host, zabbix_port=self.zbx_sender_port).send(zm)
+        try:
+            response = ZabbixSender(zabbix_server=self.zbx_sender_host, zabbix_port=self.zbx_sender_port).send(zm)
+            if response.failed:
+                elastalert_logger.warning("Missing zabbix host '%s' or host's item '%s', alert will be discarded"
+                                          % (self.zbx_host, self.zbx_key))
+            else:
+                elastalert_logger.info("Alert sent to Zabbix")
+        except Exception as e:
+            raise EAException("Error sending alert to Zabbix: %s" % e)
 
     # get_info is called after an alert is sent to get data that is written back
     # to Elasticsearch in the field "alert_info"