RuntimeError: threads can only be started once

This error occurs when you try to start a Python thread more than once using the threading module. A thread can only be started once, and attempting to restart an already started thread results in a RuntimeError.

1. Understanding the Cause

Python’s threading.Thread objects cannot be restarted once they have been started or completed execution.

Incorrect Code (Trying to Restart a Thread)

import threading

def print_numbers():
    for i in range(5):
        print(i)

t = threading.Thread(target=print_numbers)
t.start()  # First start - Works fine
t.start()  # Second start - Causes RuntimeError

Output:

0
1
2
3
4
RuntimeError: threads can only be started once

2. How to Fix the Issue?

Solution 1: Create a New Thread Instance Instead of Restarting

Once a thread completes execution, you must create a new thread object if you want to run it again.

import threading

def print_numbers():
    for i in range(5):
        print(i)

t = threading.Thread(target=print_numbers)
t.start()
t.join()  # Wait for the thread to finish

# Create a new thread instance before restarting
t = threading.Thread(target=print_numbers)
t.start()

Solution 2: Use a Loop Instead of Restarting a Thread

If you want the function to run multiple times, use a loop inside a single thread instead of restarting it.

import threading
import time

def print_numbers():
    while True:
        for i in range(5):
            print(i)
        time.sleep(2)  # Wait before repeating

t = threading.Thread(target=print_numbers)
t.start()

This avoids restarting the thread, keeping it running in a loop.

Solution 3: Use a Thread Pool Instead

If you need to run a function multiple times with threading, use concurrent.futures.ThreadPoolExecutor, which automatically manages threads.

from concurrent.futures import ThreadPoolExecutor

def print_numbers():
    for i in range(5):
        print(i)

with ThreadPoolExecutor(max_workers=2) as executor:
    executor.submit(print_numbers)
    executor.submit(print_numbers)  # Runs in a separate thread

This method avoids managing threads manually.