Use _stop_event consistently across all threads as the way to signal that it should stop. Add thread joins with timeouts to allow the program to exit cleanly

providers/alert/http_alert_provider.py

@@ -27,11 +27,15 @@ class HTTPAlertProvider(AlertProvider):
         # Fire off the polling thread. It will poll immediately on startup, then sleep for poll_interval between
         # subsequent polls, so start() returns immediately and the application can continue starting.
         logger.info(f"Set up query of {self.name} alert API every {self._poll_interval!s} seconds.")
-        self._thread = Thread(target=self._run, name=f"HTTPAlertProvider-{self.name}")
+        self._thread = Thread(target=self._run, name=f"HTTPAlertProvider-{self.name}", daemon=True)
         self._thread.start()
 
     def stop(self):
         self._stop_event.set()
+        if self._thread:
+            self._thread.join(timeout=35)
+            if self._thread.is_alive():
+                logger.warning(f"{self.name} alert worker thread did not exit on time and will be killed.")
 
     def _run(self):
         while True:

providers/callsigndata/file_download_callsign_data_provider.py

@@ -32,11 +32,15 @@ class FileDownloadCallsignDataProvider(CallsignDataProvider):
         # Fire off the polling thread. It will poll immediately on startup, then sleep for poll_interval between
         # subsequent polls, so start() returns immediately and the application can continue starting.
         logger.info(f"Set up query of {self.name} callsign reference data every {self._poll_interval!s} days.")
-        self._thread = Thread(target=self._run, name=f"FileDownloadCallsignDataProvider-{self.name}")
+        self._thread = Thread(target=self._run, name=f"FileDownloadCallsignDataProvider-{self.name}", daemon=True)
         self._thread.start()
 
     def stop(self):
         self._stop_event.set()
+        if self._thread:
+            self._thread.join(timeout=35)
+            if self._thread.is_alive():
+                logger.warning(f"{self.name} callsign data worker thread did not exit on time and will be killed.")
 
     def _run(self):
         while True:

providers/sigrefdata/file_download_sig_ref_data_provider.py

@@ -1,6 +1,6 @@
 import logging
 from datetime import datetime
-from threading import Event, Thread
+from threading import Thread
 
 import pytz
 from requests.exceptions import ConnectionError, ConnectTimeout, ReadTimeout
@@ -21,19 +21,21 @@ class FileDownloadSIGRefDataProvider(SIGRefDataProvider):
         self._url = url
         self._poll_interval = poll_interval
         self._thread = None
-        self._stop_event = Event()
         self._url_data_cache = URLDataCache(f"sigrefdata_{sig_name}")
 
     def start(self):
         # Fire off the polling thread. It will poll immediately on startup, then sleep for poll_interval between
         # subsequent polls, so start() returns immediately and the application can continue starting.
         logger.info(f"Set up query of {self.sig_name} SIG ref data every {self._poll_interval!s} days.")
-        self._thread = Thread(target=self._run, name=f"FileDownloadSIGRefDataProvider-{self.sig_name}")
+        self._thread = Thread(target=self._run, name=f"FileDownloadSIGRefDataProvider-{self.sig_name}", daemon=True)
         self._thread.start()
 
     def stop(self):
         super().stop()
-        self._stop_event.set()
+        if self._thread:
+            self._thread.join(timeout=35)
+            if self._thread.is_alive():
+                logger.warning(f"{self.sig_name} SIG ref data worker thread did not exit on time and will be killed.")
 
     def _run(self):
         while True:

providers/sigrefdata/sig_ref_data_provider.py

@@ -1,5 +1,6 @@
 import logging
 from datetime import datetime
+from threading import Event
 
 import pytz
 
@@ -19,7 +20,7 @@ class SIGRefDataProvider:
         self.last_update_time = datetime.min.replace(tzinfo=pytz.UTC)
         self.status = "Not Started" if self.enabled else "Disabled"
         self.reference_count = 0
-        self._stop = False
+        self._stop_event = Event()
 
     def start(self):
         """Start the provider. This should return immediately after spawning threads to access the remote resources"""
@@ -30,7 +31,7 @@ class SIGRefDataProvider:
         """Stop any threads and prepare for application shutdown. Subclasses should implement this method and call
         super()."""
 
-        self._stop = True
+        self._stop_event.set()
 
     def _add_data(self, new_data):
         """Add all the provided reference data objects to the data store."""
@@ -43,7 +44,7 @@ class SIGRefDataProvider:
                 # For the big data sources, loading will take a few minutes. If we want to shut down the software neatly
                 # within the first few minutes of startup, we need a way to abort this expensive process of filling up the
                 # disk cache.
-                if self._stop:
+                if self._stop_event.is_set():
                     break
 
         self.reference_count = len(new_data)

providers/solarconditions/giroionosonde.py

@@ -67,11 +67,15 @@ class GIROIonosonde(SolarConditionsProvider):
 
     def start(self):
         logger.info(f"Set up query of GIRO ionosonde data API every {POLL_INTERVAL} seconds.")
-        self._thread = Thread(target=self._run, name="GIROIonosondeDataProvider")
+        self._thread = Thread(target=self._run, name="GIROIonosondeDataProvider", daemon=True)
         self._thread.start()
 
     def stop(self):
         self._stop_event.set()
+        if self._thread:
+            self._thread.join(timeout=35)
+            if self._thread.is_alive():
+                logger.warning("GIRO ionosonde worker thread did not exit on time and will be killed.")
 
     def _run(self):
         # Real interval at which we poll is the "once per hour" divided by the number of stations, so each one gets

providers/solarconditions/http_solar_conditions_provider.py

@@ -25,11 +25,15 @@ class HTTPSolarConditionsProvider(SolarConditionsProvider):
 
     def start(self):
         logger.info(f"Set up query of {self.name} solar conditions API every {self._poll_interval!s} seconds.")
-        self._thread = Thread(target=self._run, name=f"HTTPSolarConditionsProvider-{self.name}")
+        self._thread = Thread(target=self._run, name=f"HTTPSolarConditionsProvider-{self.name}", daemon=True)
         self._thread.start()
 
     def stop(self):
         self._stop_event.set()
+        if self._thread:
+            self._thread.join(timeout=35)
+            if self._thread.is_alive():
+                logger.warning(f"{self.name} solar conditions worker thread did not exit on time and will be killed.")
 
     def _run(self):
         while True:

providers/solarconditions/kc2gprop.py

@@ -32,11 +32,15 @@ class KC2GProp(SolarConditionsProvider):
 
     def start(self):
         logger.info(f"Set up query of KC2G ionosonde data API every {POLL_INTERVAL} seconds.")
-        self._thread = Thread(target=self._run, name="KC2GPropProvider")
+        self._thread = Thread(target=self._run, name="KC2GPropProvider", daemon=True)
         self._thread.start()
 
     def stop(self):
         self._stop_event.set()
+        if self._thread:
+            self._thread.join(timeout=35)
+            if self._thread.is_alive():
+                logger.warning("KC2G ionosonde worker thread did not exit on time and will be killed.")
 
     def _run(self):
         while True:

providers/spot/aprsis.py

@@ -17,17 +17,16 @@ class APRSIS(SpotProvider):
 
     def __init__(self, provider_config):
         super().__init__("APRS-IS", provider_config)
-        self._thread = Thread(target=self._run, name="APRSISSpotProvider")
-        self._thread.daemon = True
+        self._thread = None
         self._aprsis = None
-        self._running = True
         self._stop_event = Event()
 
     def start(self):
+        self._thread = Thread(target=self._run, name="APRSISSpotProvider", daemon=True)
         self._thread.start()
 
     def _run(self):
-        while self._running:
+        while not self._stop_event.is_set():
             try:
                 self._aprsis = aprslib.IS(SERVER_OWNER_CALLSIGN)
                 self.status = "Connecting"
@@ -37,20 +36,22 @@ class APRSIS(SpotProvider):
                 self._aprsis.consumer(self._handle, immortal=True)
 
             except Exception:
-                if self._running:
+                if not self._stop_event.is_set():
                     self.status = "Error"
                     logger.exception("Exception in APRS-IS provider")
 
-            if self._running:
+            if not self._stop_event.is_set():
                 self._stop_event.wait(timeout=5)
 
     def stop(self):
-        self._running = False
         self.status = "Shutting down"
         self._stop_event.set()
         if self._aprsis:
             self._aprsis.close()
-        self._thread.join()
+        if self._thread:
+            self._thread.join(timeout=15)
+            if self._thread.is_alive():
+                logger.warning("APRS-IS worker thread did not exit on time and will be killed.")
 
     def _handle(self, data):
         try:

providers/spot/dxcluster.py

@@ -1,8 +1,7 @@
 import logging
 import re
 from datetime import datetime
-from threading import Thread
-from time import sleep
+from threading import Event, Thread
 
 import pytz
 import telnetlib3
@@ -41,23 +40,26 @@ class DXCluster(SpotProvider):
             self._LINE_PATTERN_ALLOW_RBN if self._allow_rbn_spots else self._LINE_PATTERN_EXCLUDE_RBN
         )
         self._telnet = None
-        self._thread = Thread(target=self._handle, name=f"DXClusterSpotProvider-{self.name}")
-        self._thread.daemon = True
-        self._running = True
+        self._thread = None
+        self._stop_event = Event()
 
     def start(self):
+        self._thread = Thread(target=self._handle, name=f"DXClusterSpotProvider-{self.name}", daemon=True)
         self._thread.start()
 
     def stop(self):
-        self._running = False
+        self._stop_event.set()
         if self._telnet:
             self._telnet.close()
-        self._thread.join()
+        if self._thread:
+            self._thread.join(timeout=15)
+            if self._thread.is_alive():
+                logger.warning(f"DX Cluster {self._hostname} worker thread did not exit on time and will be killed.")
 
     def _handle(self):
-        while self._running:
+        while not self._stop_event.is_set():
             connected = False
-            while not connected and self._running:
+            while not connected and not self._stop_event.is_set():
                 try:
                     self.status = "Connecting"
                     logger.info(f"DX Cluster {self._hostname} connecting...")
@@ -69,14 +71,14 @@ class DXCluster(SpotProvider):
                 except ConnectionRefusedError:
                     self.status = "Error"
                     logger.warning(f"Connection refused to DX cluster {self._hostname}")
-                    sleep(300)
+                    self._stop_event.wait(timeout=300)
                 except Exception:
                     self.status = "Error"
                     logger.exception(f"Exception while connecting to DX Cluster Provider ({self._hostname}).")
-                    sleep(5)
+                    self._stop_event.wait(timeout=5)
 
             self.status = "Waiting for Data"
-            while connected and self._running:
+            while connected and not self._stop_event.is_set():
                 try:
                     # Check new telnet info against regular expression
                     telnet_output = self._telnet.read_until("\n".encode("latin-1"))
@@ -106,19 +108,19 @@ class DXCluster(SpotProvider):
 
                 except EOFError:
                     connected = False
-                    if self._running:
+                    if not self._stop_event.is_set():
                         self.status = "Restarting"
                         logger.warning(f"Disconnected from DX Cluster {self._hostname}. Reconnecting...")
-                        sleep(5)
+                        self._stop_event.wait(timeout=5)
                     else:
                         logger.info(f"DX Cluster {self._hostname} shutting down...")
                         self.status = "Shutting down"
                 except Exception:
                     connected = False
-                    if self._running:
+                    if not self._stop_event.is_set():
                         self.status = "Error"
                         logger.exception(f"Exception in DX Cluster Provider ({self._hostname})")
-                        sleep(5)
+                        self._stop_event.wait(timeout=5)
                     else:
                         logger.info(f"DX Cluster {self._hostname} shutting down...")
                         self.status = "Shutting down"

providers/spot/http_spot_provider.py

@@ -28,12 +28,16 @@ class HTTPSpotProvider(SpotProvider):
         # Fire off the polling thread. It will poll immediately on startup, then sleep for poll_interval between
         # subsequent polls, so start() returns immediately and the application can continue starting.
         logger.info(f"Set up query of {self.name} spot API every {self._poll_interval!s} seconds.")
-        self._thread = Thread(target=self._run, name=f"HTTPSpotProvider-{self.name}")
+        self._thread = Thread(target=self._run, name=f"HTTPSpotProvider-{self.name}", daemon=True)
         self._thread.start()
 
     def stop(self):
         self._stop_event.set()
         self._wakeup_event.set()
+        if self._thread:
+            self._thread.join(timeout=35)
+            if self._thread.is_alive():
+                logger.warning(f"{self.name} spot worker thread did not exit on time and will be killed.")
 
     def force_poll(self):
         """Trigger an immediate poll without waiting for the normal interval."""

providers/spot/rbn.py

@@ -1,8 +1,7 @@
 import logging
 import re
 from datetime import datetime
-from threading import Thread
-from time import sleep
+from threading import Event, Thread
 
 import pytz
 import telnetlib3
@@ -30,23 +29,26 @@ class RBN(SpotProvider):
         super().__init__(name, provider_config)
         self._port = provider_config["port"]
         self._telnet = None
-        self._thread = Thread(target=self._handle, name=f"RBNSpotProvider-{self.name}")
-        self._thread.daemon = True
-        self._running = True
+        self._thread = None
+        self._stop_event = Event()
 
     def start(self):
+        self._thread = Thread(target=self._handle, name=f"RBNSpotProvider-{self.name}", daemon=True)
         self._thread.start()
 
     def stop(self):
-        self._running = False
+        self._stop_event.set()
         if self._telnet:
             self._telnet.close()
-        self._thread.join()
+        if self._thread:
+            self._thread.join(timeout=15)
+            if self._thread.is_alive():
+                logger.warning(f"RBN (port {self._port!s}) worker thread did not exit on time and will be killed.")
 
     def _handle(self):
-        while self._running:
+        while not self._stop_event.is_set():
             connected = False
-            while not connected and self._running:
+            while not connected and not self._stop_event.is_set():
                 try:
                     self.status = "Connecting"
                     logger.info(f"RBN port {self._port!s} connecting...")
@@ -58,10 +60,10 @@ class RBN(SpotProvider):
                 except Exception:
                     self.status = "Error"
                     logger.exception(f"Exception while connecting to RBN (port {self._port!s}).")
-                    sleep(5)
+                    self._stop_event.wait(timeout=5)
 
             self.status = "Waiting for Data"
-            while connected and self._running:
+            while connected and not self._stop_event.is_set():
                 try:
                     # Check new telnet info against regular expression
                     telnet_output = self._telnet.read_until("\n".encode("latin-1"))
@@ -91,19 +93,19 @@ class RBN(SpotProvider):
 
                 except EOFError:
                     connected = False
-                    if self._running:
+                    if not self._stop_event.is_set():
                         self.status = "Restarting"
                         logger.warning(f"Disconnected from RBN provider (port {self._port!s}). Reconnecting...")
-                        sleep(5)
+                        self._stop_event.wait(timeout=5)
                     else:
                         logger.info(f"RBN provider (port {self._port!s}) shutting down...")
                         self.status = "Shutting down"
                 except Exception:
                     connected = False
-                    if self._running:
+                    if not self._stop_event.is_set():
                         self.status = "Error"
                         logger.exception(f"Exception in RBN provider (port {self._port!s})")
-                        sleep(5)
+                        self._stop_event.wait(timeout=5)
                     else:
                         logger.info(f"RBN provider (port {self._port!s}) shutting down...")
                         self.status = "Shutting down"

providers/spot/websocket_spot_provider.py

@@ -1,7 +1,6 @@
 import logging
 from datetime import datetime
-from threading import Thread
-from time import sleep
+from threading import Event, Thread
 
 import pytz
 from websocket import create_connection
@@ -20,22 +19,24 @@ class WebsocketSpotProvider(SpotProvider):
         self._url = url
         self._ws = None
         self._thread = None
-        self._stopped = False
+        self._stop_event = Event()
         self._last_event_id = None
 
     def start(self):
         logger.info(f"Set up websocket connection to {self.name} spot API.")
-        self._stopped = False
+        self._stop_event.clear()
         self._thread = Thread(target=self._run, name=f"WebsocketSpotProvider-{self.name}")
         self._thread.daemon = True
         self._thread.start()
 
     def stop(self):
-        self._stopped = True
+        self._stop_event.set()
         if self._ws:
             self._ws.close()
         if self._thread:
-            self._thread.join()
+            self._thread.join(timeout=15)
+            if self._thread.is_alive():
+                logger.warning(f"{self.name} websocket worker thread did not exit on time and will be killed.")
 
     def _on_open(self):
         self.status = "Waiting for Data"
@@ -44,7 +45,7 @@ class WebsocketSpotProvider(SpotProvider):
         self.status = "Connecting"
 
     def _run(self):
-        while not self._stopped:
+        while not self._stop_event.is_set():
             try:
                 logger.debug(f"Connecting to {self.name} spot API...")
                 self.status = "Connecting"
@@ -53,7 +54,7 @@ class WebsocketSpotProvider(SpotProvider):
 
                 # Keep reading from this same connection until it drops or we're asked to stop, rather than
                 # reconnecting for every message.
-                while not self._stopped:
+                while not self._stop_event.is_set():
                     data = self._ws.recv()
                     if not data:
                         break
@@ -82,8 +83,8 @@ class WebsocketSpotProvider(SpotProvider):
                         # No problem, we were getting rid of this object anyway.
                         pass
                     self._ws = None
-            if not self._stopped:
-                sleep(5)  # Wait before trying to reconnect
+            if not self._stop_event.is_set():
+                self._stop_event.wait(timeout=5)  # Wait before trying to reconnect
 
     def _ws_message_to_spot(self, b):
         """Convert a WS message received from the API into a spot. The exact message data (in bytes) is provided here so the

providers/staticdata/file_download_static_data_provider.py

@@ -29,11 +29,15 @@ class FileDownloadStaticDataProvider(StaticDataProvider):
         # Fire off the polling thread. It will poll immediately on startup, then sleep for poll_interval between
         # subsequent polls, so start() returns immediately and the application can continue starting.
         logger.info(f"Set up query of {self.name} static reference data every {self._poll_interval!s} days.")
-        self._thread = Thread(target=self._run, name=f"FileDownloadStaticDataProvider-{self.name}")
+        self._thread = Thread(target=self._run, name=f"FileDownloadStaticDataProvider-{self.name}", daemon=True)
         self._thread.start()
 
     def stop(self):
         self._stop_event.set()
+        if self._thread:
+            self._thread.join(timeout=35)
+            if self._thread.is_alive():
+                logger.warning(f"{self.name} static data worker thread did not exit on time and will be killed.")
 
     def _run(self):
         while True: