feat: add fetch_status_task to poll external service with concurrency control

Author: Lumabots

examples/background_task.py

@@ -1,3 +1,4 @@
+import asyncio
 from datetime import time, timezone
 
 import discord
@@ -10,7 +11,6 @@ def __init__(self, *args, **kwargs):
 
         # An attribute we can access from our task
         self.counter = 0
-
         # Start the tasks to run in the background
         self.my_background_task.start()
         self.time_task.start()
@@ -37,6 +37,36 @@ async def time_task(self):
     async def before_my_task(self):
         await self.wait_until_ready()  # Wait until the bot logs in
 
+    # Schedule every 10s; each run takes ~15s. With overlap=2, at most 2 runs
+    # execute concurrently so we don't build an ever-growing backlog.
+    @tasks.loop(seconds=10, overlap=2)
+    async def fetch_status_task(self):
+        """
+        Practical overlap use-case:
+
+        Poll an external service and post a short summary. Each poll may take
+        ~15s due to network latency or rate limits, but we want fresh data
+        every 10s. Allowing a small amount of overlap avoids drifting schedules
+        without opening the floodgates to unlimited concurrency.
+        """
+        # Book-keeping so we can show concurrency in logs/messages
+        run_no = self.fetch_status_task.current_loop
+
+        try:
+            print(f"[status] start run #{run_no}")
+
+            # Simulate slow I/O (e.g., HTTP requests, DB queries, file I/O)
+            await asyncio.sleep(15)
+
+            channel = self.get_channel(1234567)  # Replace with your channel ID
+            msg = f"[status] run #{run_no} complete"
+            if channel:
+                await channel.send(msg)
+            else:
+                print(msg)
+        finally:
+            print(f"[status] end run #{run_no}")
+
 
 client = MyClient()
 client.run("TOKEN")