From a3bbd7eeabee7c6b229e95e0e04af5b430ea32db Mon Sep 17 00:00:00 2001 From: Eric Eastwood Date: Tue, 13 May 2025 10:40:49 -0500 Subject: Explain why we `flush_buffer()` for Python `print(...)` output (#18420) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Spawning from using this code elsewhere and not knowing why it's there. Based on this article and @reivilibre's experience mentioning `PYTHONUNBUFFERED=1`, > #### programming languages where the default “print” statement buffers > > Also, here are a few programming language where the default print statement will buffer output when writing to a pipe, and some ways to disable buffering if you want: > > - Python (disable with `python -u`, or `PYTHONUNBUFFERED=1`, or `sys.stdout.reconfigure(line_buffering=False)`, or `print(x, flush=True)`) > > _-- https://jvns.ca/blog/2024/11/29/why-pipes-get-stuck-buffering/#programming-languages-where-the-default-print-statement-buffers_ --- docker/configure_workers_and_start.py | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'docker/configure_workers_and_start.py') diff --git a/docker/configure_workers_and_start.py b/docker/configure_workers_and_start.py index df34d51f77..102a88fad1 100755 --- a/docker/configure_workers_and_start.py +++ b/docker/configure_workers_and_start.py @@ -352,6 +352,11 @@ def error(txt: str) -> NoReturn: def flush_buffers() -> None: + """ + Python's `print()` buffers output by default, typically waiting until ~8KB + accumulates. This method can be used to flush the buffers so we can see the output + of any print statements so far. + """ sys.stdout.flush() sys.stderr.flush() -- cgit 1.5.1