Fix duplicate logs in run_simulation

[chongshenng]

Issue

Description

When executing run_simulation in Google Colab, duplicate logs appear in the output.

Current behaviour:

INFO :      Flower ECE: Starting Driver API (gRPC-rere) on 0.0.0.0:9091
INFO:flwr:Flower ECE: Starting Driver API (gRPC-rere) on 0.0.0.0:9091
INFO :      Starting Flower ServerApp, config: num_rounds=3, no round_timeout
INFO:flwr:Starting Flower ServerApp, config: num_rounds=3, no round_timeout
INFO :      Registered 10 nodes
INFO:flwr:Registered 10 nodes
INFO :      
INFO:flwr:
INFO :      Supported backends: ['ray']
INFO:flwr:Supported backends: ['ray']
INFO :      Initialising: RayBackend
INFO:flwr:Initialising: RayBackend
INFO :      Backend config: {'client_resources': None}
INFO:flwr:Backend config: {'client_resources': None}
INFO :      [INIT]
INFO:flwr:[INIT]
INFO :      Using initial global parameters provided by strategy
INFO:flwr:Using initial global parameters provided by strategy
INFO :      Evaluating initial global parameters
INFO:flwr:Evaluating initial global parameters
INFO :      
INFO:flwr:
INFO :      [ROUND 1]
INFO:flwr:[ROUND 1]

This behaviour is unintended.

This behaviour is also not observed when running local notebooks in VSCode or JupyterLab.

Related issues/PRs

N/A

Proposal

Explanation

Set logger.propagate = False when an asyncio event is detected. This allows only the child logger to print the message and not the parent logger. See this SO thread.

With fix:

INFO :      Flower ECE: Starting Driver API (gRPC-rere) on 0.0.0.0:9091
INFO :      Starting Flower ServerApp, config: num_rounds=3, no round_timeout
INFO :      
INFO :      Registered 10 nodes
INFO :      Supported backends: ['ray']
INFO :      [INIT]
INFO :      Using initial global parameters provided by strategy
INFO :      Evaluating initial global parameters
INFO :      
INFO :      [ROUND 1]
INFO :      Initialising: RayBackend
INFO :      Backend config: {'client_resources': None}
2024-04-29 11:54:21,628	WARNING worker.py:1419 -- SIGTERM handler is not set because current thread is not the main thread.
INFO :      DriverServicer.CreateRun
INFO :      configure_fit: strategy sampled 3 clients (out of 10)

Checklist

• Implement proposed change
• [ ] Write tests
• [ ] Update documentation
• Update the changelog entry below
• Make CI checks pass

Changelog entry

Any other comments?

[jafermarq]

This is gold!