How to Update Global Variables in Parent Process When Using multiprocessing in Python?

Global Variable Updates not Reflected in Parent with multiprocessing

In multiprocessing, creating a new process initiates a separate Python instance with its own global state. Modifications made to global variables within child processes are not visible to the parent process.

This issue stems from the fact that data transferred between processes using multiprocessing is typically pickled. Pickling requires objects to be serializable, which may not be possible for certain objects like LORR instances.

To resolve this, consider these approaches:

1. Creating a ctype Object: Store the data in a ctype object and pass the memory reference to the child process. However, this method is not commonly used and can be complex.
2. Subclassing LORR: If LORR is a class and subclassing is possible, implement getstate and setstate methods to enable pickling. However, this requires a deep understanding of LORR's internal structure.
3. Messaging via Queues: Alternatively, you can use queues to communicate between processes. Instead of passing LORR instances, send messages to child processes requesting specific method executions. The data from the methods can then be returned as strings via queues.

The above is the detailed content of How to Update Global Variables in Parent Process When Using multiprocessing in Python?. For more information, please follow other related articles on the PHP Chinese website!