From 2d6f965404198649d3824fa659b5cb5164e2a86f Mon Sep 17 00:00:00 2001
From: Lawrence Mitchell <lmitchell@nvidia.com>
Date: Thu, 6 Jun 2024 11:09:40 +0000
Subject: [PATCH] Document how to use cudf.pandas in tandem with
 multiprocessing

We need to arrange that cudf.pandas.install() is run on the workers,
this requires that we programmatically install the metapath loader in
our script. Unfortunately, passing an initializer function to the pool
startup is not sufficient if any part of the script transitively loads
pandas at the top level.

- Closes #15246
---
 docs/cudf/source/cudf_pandas/usage.md | 30 +++++++++++++++++++++++++++
 1 file changed, 30 insertions(+)

diff --git a/docs/cudf/source/cudf_pandas/usage.md b/docs/cudf/source/cudf_pandas/usage.md
index b174c606d66..376784439aa 100644
--- a/docs/cudf/source/cudf_pandas/usage.md
+++ b/docs/cudf/source/cudf_pandas/usage.md
@@ -26,6 +26,36 @@ From the command line, run your Python scripts with `-m cudf.pandas`:
 python -m cudf.pandas script.py
 ```
 
+### Usage in tandem with
+[`multiprocessing`](https://docs.python.org/3/library/multiprocessing.html)
+or
+[`concurrent.futures`](https://docs.python.org/3/library/concurrent.futures.html)
+process pools
+
+To use a pool of workers (for example
+[`multiprocessing.Pool`](https://docs.python.org/3/library/multiprocessing.html#multiprocessing.pool.Pool)
+or
+[`concurrent.futures.ProcessPoolExecutor`](https://docs.python.org/3/library/concurrent.futures.html#concurrent.futures.ProcessPoolExecutor))
+in your script with `cudf.pandas`, the `cudf.pandas` module must be
+loaded on the worker processes, as well as by the controlling script.
+The most foolproof way to do this is to programmatically install
+`cudf.pandas` at the top of your script, before anything else.
+For example
+
+```python
+# This is equivalent to python -m cudf.pandas, but will run on the
+# workers too. These two lines must run before pandas is imported,
+# either directly or transitively.
+import cudf.pandas
+cudf.pandas.install()
+
+from multiprocessing import Pool
+
+with Pool(4) as pool:
+    # use pool here
+    ...
+```
+
 ## Understanding performance - the `cudf.pandas` profiler
 
 `cudf.pandas` will attempt to use the GPU whenever possible and fall