"bleak.exc.BleakError: This service is already present in this BleakGATTServiceCollection!" when reconnecting a device #376
Comments
|
Do you have two services with the same uuid on the peripheral? In that case this is related to #362. |
|
No I have ran only one service with |
|
Can you give a full working code example to reproduce the problem? |
|
I'm seeing something perhaps related, when I've got a device that drops connectivity (powered off) and then try to reconnect programmatically. If you think this is a different issue, more than willing to split it off. I'm still diagnosing and trying to discover "why". No minimal, reproducible case yet. bluetoothctl: 5.50 Even if I try an explicit disconnect, the problem still seems to be present. In the disconnect callback if not client.willful_disconnect:
asyncio.get_event_loop().create_task(self._reconnect()) async def _reconnect(self):
"""
Will try immediately, then 1, 2, 3, ..., 10, 10, ... seconds later
"""
await self.disconnect()
class_name = type(self).__name__
if self._logging_reconnect:
logger.info(
f"Will try reconnecting to {class_name} at {self.address} "
f"after waiting {self._reconnect_delay} seconds.")
await asyncio.sleep(self._reconnect_delay)
await self.connect()
if self.is_connected:
self._reconnect_delay = 0
self._logging_reconnect = True
else:
# code to try again in a bitThe wrapper tries to manage clean-up atexit, marginally successfully (mainly around if there is a running loop and things complete before Python shuts it down). async def connect(self, **kwargs) -> bool:
atexit.register(self.sync_disconnect)
retval = await super(BleakClientWrapped, self).connect(**kwargs)
if retval:
self._willful_disconnect = False
return retval
async def disconnect(self) -> bool:
self._willful_disconnect = True
retval = await super(BleakClientWrapped, self).disconnect()
self._willful_disconnect = False
if retval:
logger.debug("Unregistering atexit disconnect "
f"{self.name} at {self.address}")
atexit.unregister(self.sync_disconnect)
return retvalOn powering back on the Skale, it does show up in Without deep knowledge of what is supposed to be happening, it is my guess that a previous service discovery is not being cleared before adding in the re-discovered services. Breakpoint at line 325 in
|
|
To reproduce:
import asyncio
import logging
from bleak import BleakClient
format_string = "%(asctime)s %(levelname)s %(name)s: %(message)s"
logging.basicConfig(level=logging.DEBUG,
format=format_string,
)
logger = logging.getLogger('run')
def disconnect_handler(c: BleakClient):
logger.info("Disconnect handler")
asyncio.get_running_loop().create_task(reconnect(c))
async def reconnect(c: BleakClient):
logger.info("Reconnect")
await c.connect()
if not c.is_connected:
# lazily call recursively, as shouldn't be more than a few
await asyncio.sleep(10)
await reconnect(c)
async def run():
BLE_ID = 'CF:75:75:90:8C:E3'
client = BleakClient(BLE_ID)
client.set_disconnected_callback(disconnect_handler)
await client.connect()
logger.info("Connected")
await asyncio.sleep(90)
logger.info("90 seconds, disconnecting")
await client.disconnect()
if __name__ == '__main__':
loop = asyncio.get_event_loop()
loop.set_debug(True)
loop.run_until_complete(run()) |
|
Ugly fix, but this at least appears to be functional for this specific host and implementation Use "private access" with |
|
For those seeing this issue when reconnecting after a disconnect, I think #621 (comment) will fix the issue (it essentially does this: #376 (comment)). If anyone is seeing this error at a different time, it would be interesting/helpful if you could confirm the sequence of events I describe in #625 (comment). |
dlech
changed the title
Test example for hbldh/bleak#376
|
This should be fixed by #902. |
bluetoothctl -v) in case of Linux: 5.50Description
For very first time the central code for scanning BLE Devices worked and connected with uart device. But after connection I was not able to disconnect it.
What I Did
I had scanned BLE Devices and make connection. It worked and shown
disconnected, scanning
connected
Then I had stopped peripheral service. After service was stopped I had ran scanner class again then i was showing below error.
The text was updated successfully, but these errors were encountered: