Python Daemon Thread Memory Leak?

Question

If I create a daemon thread in Python (3.6+) and that daemon thread is finished execution, are its resources automatically freed up, or is it a memory leak? I looked around couldn't really find the answer...I assume it is freed up but paranoid and hoping someone can clarify.

I made a basic example:

import threading
import time

class Processor():
    
    def get_data_in_thread(self, id):
        for i in range(3):
            time.sleep(1)
            print(f'{id} count: {i}')

        #at this point is this thread's resources automatically garbage collected? Or is this a memory leak?

    def get_data(self, id):
        t = threading.Thread(target=self.get_data_in_thread, args=(id,))
        t.daemon = True
        t.start()


my_processor = Processor()
my_processor.get_data(17)

time.sleep(30)

When the work inside the daemon thread is finished in get_data_in_thread does the memory allocated within that thread free up automatically or is there a specific command I can use to self terminate its own daemon thread? Like del() or similar?

Thanks!

It would make sense for the GIL to take care of it; you can read more about it [here](https://wiki.python.org/moin/GlobalInterpreterLock#:~:text=In%20CPython%2C%20the%20global%20interpreter,management%20is%20not%20thread%2Dsafe.). — Algebra8, Aug 11 '20 at 15:39
You can find an answer to your question in [this](https://stackoverflow.com/questions/190010/daemon-threads-explanation) stack overflow question — Algebra8, Aug 11 '20 at 16:04
Does this answer your question? [Do threads in python need to be joined to avoid leakage?](https://stackoverflow.com/questions/38275944/do-threads-in-python-need-to-be-joined-to-avoid-leakage) — MisterMiyagi, Aug 11 '20 at 16:17

score 1 · Accepted Answer · answered Aug 11 '20 at 15:38

When a thread is destroyed, just like any other object, if it's no longer referenced (refcount goes to 0), it gets collected immediately. So do all the resources the thread referenced to, in your case the internal function variables, as they aren't referenced anywhere else.

Anmol Singh Jaggi · Answer 2 · 2020-08-11T16:14:46.667

0

You have to call thread.join() to completely mark it eligible for garbage collection.
And if you want to get all its memory deallocated immediately after its joined, then execute gc.collect().

Reference

edited Aug 11 '20 at 16:14

answered Aug 11 '20 at 16:08

Anmol Singh Jaggi

8,376
4
36
77

The question is with regards to daemon threads, you don't need to call `thread.join()` on daemon threads. – Algebra8 Aug 11 '20 at 16:31
You don't need to call ``thread.join()`` on *any* thread. It merely exists to wait for a thread to end – there is no need to do that if you don't want to wait for the thread. – MisterMiyagi Aug 11 '20 at 16:34
Algebra8 is right. I ran [this](https://stackoverflow.com/a/63299113/1925388) code after making the threads daemon and there was no need of calling join. However, what I said is true for normal threads @MisterMiyagi – Anmol Singh Jaggi Aug 11 '20 at 18:18

Python Daemon Thread Memory Leak?

2 Answers2