Running a prefect server from a executable and freeze_support is not working and unable to access prefect server from localhost:4200 on browser

[kondasandeep13]

main.py

from multiprocessing import Process, freeze_support, set_start_method
import os
os.environ["PREFECT_API_URL"] = "http://127.0.0.1:4200/api"
os.environ["no_proxy"] = "127.0.0.1:4200"

def run_prefect_server():
    """Run the Prefect server."""
    print("Starting Prefect server...")

    try:
        server.start(host='127.0.0.1', port=4200, keep_alive_timeout=5,
                     log_level='WARNING', scheduler=True, analytics=True,
                     late_runs=True, ui=True, background=False)
    except Exception as e:
        print(f"Error starting Prefect server: {e}")

if __name__ == "__main__":

    set_start_method("spawn", force=True)
    freeze_support()
    prefect_process = Process(target=run_prefect_server)
    prefect_process.start()
    # Wait for all processes to finish
    prefect_process.join()

Error

Process ID: 33416
Script is executing...
Starting Prefect server...

 ___ ___ ___ ___ ___ ___ _____
| _ \ _ \ __| __| __/ __|_   _|
|  _/   / _|| _|| _| (__  | |
|_| |_|_\___|_| |___\___| |_|

Configure Prefect to communicate with the server with:

    prefect config set PREFECT_API_URL=http://127.0.0.1:4200/api

View the API reference documentation at http://127.0.0.1:4200/docs

Check out the dashboard at http://127.0.0.1:4200



16:29:31.401 | INFO    | prefect.cli.server - Opening server process with command: 'C:\Users\e40017818\Desktop\final\project\dist\main.exe' -m uvicorn --app-dir 'C:\Users\E40017~1\AppData\Local\Temp\_MEI310402' --factory prefect.server.api.server:create_app --host 127.0.0.1 --port 4200 --timeout-keep-alive 5


Process ID: 33152
Script is executing...
Starting Prefect server...

 ___ ___ ___ ___ ___ ___ _____
| _ \ _ \ __| __| __/ __|_   _|
|  _/   / _|| _|| _| (__  | |
|_| |_|_\___|_| |___\___| |_|

Configure Prefect to communicate with the server with:

    prefect config set PREFECT_API_URL=http://127.0.0.1:4200/api

View the API reference documentation at http://127.0.0.1:4200/docs

Check out the dashboard at http://127.0.0.1:4200

16:29:38.351 |

INFO    | prefect.cli.server - Opening server process with command: 'C:\Users\e40017818\Desktop\final\project\dist\main.exe' -m uvicorn --app-dir 'C:\Users\E40017~1\AppData\Local\Temp\_MEI310402' --factory prefect.server.api.server:create_app --host 127.0.0.1 --port 4200 --timeout-keep-alive 5

.......................
Here, I see the prefect server code is executed multiple times, and the console is having all the above lines multiple times

Help would be appreciable.

Thanks

[rokm]

Does server.start() spawn its own subprocess using sys.executable?

Try adding print("Program arguments:", sys.argv) at the very top of your script (and maybe also print("Program (main) arguments:", sys.argv) after the freeze_support call). This should dump arguments as passed to each instance of the program - the first print would also catch the ones from multiprocessing, while the second one should catch only the one from your main process (and potentially any other sub-processes that were spawned via sys.executable outside of multiprocessing).

[rokm]

Ah, nevermind, here's the answer:

Opening server process with command: 'C:\Users\e40017818\Desktop\final\project\dist\main.exe' -m uvicorn --app-dir 'C:\Users\E40017~1\AppData\Local\Temp\_MEI310402' --factory prefect.server.api.server:create_app --host 127.0.0.1 --port 4200 --timeout-keep-alive 5

So server.start() does spawn another instance using sys.executable, which then executes your main program code again.
(If this were unfrozen script, sys.executable would be your python interpreter executable, which would then run the uvicorn module's __main__ script with provided arguments. But in the frozen application, sys.executable is your program executable).

You need to catch those specific command-line arguments (probably if sys.argv[1:3] == ['-m', 'uvicorn'] should suffice), and redirect the program flow into separate function. There, import uvicorn, and run uvicorn.main() (you'll need to look into how uvicorn.__main__ would deal with the rest of the passed command-line arguments, though).

(That's similar to what our multiprocessing run-time hook does with freeze_support...)