Working on resilience...

BAC0/core/devices/Device.py

@@ -796,6 +796,16 @@ def update_description(self, value):
         )
         self.properties.description = self.read_property("description")
 
+    def ping(self):
+        try:
+            if self.read_property("objectName"):
+                return True
+            else:
+                return False
+        except Exception as e:
+            self._log_error("Error in ping")
+            return False
+
     def __repr__(self):
         return "{} / Connected".format(self.properties.name)
 

BAC0/db/influxdb.py

@@ -33,7 +33,7 @@ def __init__(self, params):
             setattr(self, k, v)
         if self.bucket is None:
             raise ValueError("Missing bucket name, please provide one in db_params")
-        self.connect_to_db()    
+        self.connect_to_db()
 
         self.write_api = self.client.write_api(
             write_options=WriteOptions(
@@ -68,12 +68,10 @@ def connect_to_db(self):
         except:
             raise ConnectionError("Error connecting to InfluxDB")
 
-
     @property
     def health(self):
         return self.client.health()
 
-
     def clean_value(self, object_type, val, units_state):
         if "analog" in object_type:
             _string_value = "{:.3f} {}".format(val, units_state)

BAC0/scripts/Lite.py

@@ -182,9 +182,15 @@ def __init__(
                     if db_params["name"].lower() == "influxdb"
                     else None
                 )
-                self._log.info("Connection made to InfluxDB bucket : {}".format(self.database.bucket))
+                self._log.info(
+                    "Connection made to InfluxDB bucket : {}".format(
+                        self.database.bucket
+                    )
+                )
             except ConnectionError:
-                self._log.error("Unable to connect to InfluxDB. Please validate parameters")
+                self._log.error(
+                    "Unable to connect to InfluxDB. Please validate parameters"
+                )
 
     @property
     def known_network_numbers(self):

BAC0/tasks/Poll.py

@@ -19,6 +19,8 @@
 from ..core.utils.notes import note_and_log
 
 # ------------------------------------------------------------------------------
+class MultiplePollingFailures(Exception):
+    pass
 
 
 @note_and_log
@@ -70,6 +72,8 @@ def __init__(self, device, delay=10, name="", prefix="basic_poll"):
 
         :returns: Nothing
         """
+        self.failures = 0
+        self.MAX_FAILURES = 3
         self._device = weakref.ref(device)
         Task.__init__(self, name="{}_{}".format(prefix, name), delay=delay)
         self._counter = 0
@@ -80,6 +84,10 @@ def device(self):
 
     def task(self):
         try:
+            if self.failures >= self.MAX_FAILURES:
+                raise MultiplePollingFailures(
+                    "Polling failed numerous times in a row... let see what we can do"
+                )
             self.device.read_multiple(
                 list(self.device.pollable_points_name), points_per_request=25
             )
@@ -89,17 +97,32 @@ def task(self):
                 if self.device.properties.clear_history_on_save:
                     self.device.clear_histories()
                 self._counter = 0
-        except AttributeError:
+            self.failures = 0
+        except AttributeError as e:
             # This error can be seen when defining a controller on a busy network...
             # When creation fail, polling is created and fail the first time...
             # So kill the task
-            self.stop()
+            self.device._log.error(
+                "Something is wrong while creating the polling task."
+                "Error: {}".format(e)
+            )
+            # self.stop()
+            self.failures += 1
         except ValueError as e:
+            self.failures += 1
             self.device._log.error(
-                "Something is wrong with polling...stopping. Try setting off "
-                "segmentation. Error: {}".format(e)
+                "Something is wrong with polling...stopping. Will pass this time"
+                "Error: {}".format(e)
+            )
+            pass
+
+        except MultiplePollingFailures as e:
+            self.device._log.warning(
+                "Trying to ping device then we'll reset the number of failures and get back with polling"
+                "Error: {}".format(e)
             )
-            self.stop()
+            if self.device.ping():
+                self.failures = 0
 
 
 @note_and_log