Handle connection resets in the status server more gracefully

Connection resets due to network instability can lead to the status
server not catching a test status, an asyncio error like

full.log│2024-03-08 03:08:44,053 asyncio base_events      L1744 ERROR| Task exception was never retrieved                                                                                                                                  │
full.log│future: <Task finished name='Task-2038' coro=<StatusServer.cb() done, defined at /usr/lib/python3.10/site-packages/avocado/core/status/server.py:51> exception=ConnectionResetError(104, 'Connection reset by peer')>             ├
full.log│Traceback (most recent call last):                                                                                                                                                                                                │
full.log│  File "/usr/lib/python3.10/site-packages/avocado/core/status/server.py", line 53, in cb                                                                                                                                          │
full.log│    raw_message = await reader.readline()                                                                                                                                                                                         │
full.log│  File "/usr/lib64/python3.10/asyncio/streams.py", line 525, in readline                                                                                                                                                          │
full.log│    line = await self.readuntil(sep)                                                                                                                                                                                              │
full.log│  File "/usr/lib64/python3.10/asyncio/streams.py", line 617, in readuntil                                                                                                                                                         │
full.log│    await self._wait_for_data('readuntil')                                                                                                                                                                                        │
full.log│  File "/usr/lib64/python3.10/asyncio/streams.py", line 502, in _wait_for_data                                                                                                                                                    │
full.log│    await self._waiter                                                                                                                                                                                                            │
full.log│  File "/usr/lib64/python3.10/asyncio/selector_events.py", line 854, in _read_ready__data_received                                                                                                                                │
full.log│    data = self._sock.recv(self.max_size)                                                                                                                                                                                         │
full.log│ConnectionResetError: [Errno 104] Connection reset by peer

and worst yet to test tasks hanging indefinitely without the job
ever completing properly. This was mostly observed in cases of
LXC and remote spawner isolation where the isolated task process
completes but the task on the side of the task machine remains
unfinished.

Signed-off-by: Plamen Dimitrov <[email protected]>

avocado/core/status/server.py

@@ -50,7 +50,10 @@ def close(self):
 
     async def cb(self, reader, _):
         while True:
-            raw_message = await reader.readline()
+            try:
+                raw_message = await reader.readline()
+            except ConnectionResetError:
+                continue
             if not raw_message:
                 return
             self._repo.process_raw_message(raw_message)