- make sure dispose() is mentioned as a strategy for

dealing with os.fork().

Change-Id: I6d85f330327e79c353e191096d71528d8300f22f
(cherry picked from commit a02a75588d87fcbec84b4689ee8ed462c631838c)

diff --git a/doc/build/core/connections.rst b/doc/build/core/connections.rst
index 2c646bc9a043c94deff71af5a38063993866e303..49fb4552a538b943be09d2f171a5eac51456d144 100644 (file)
--- a/doc/build/core/connections.rst
+++ b/doc/build/core/connections.rst
@@ -34,7 +34,9 @@ per-object or per-function call.
 
 For a multiple-process application that uses the ``os.fork`` system call, or
 for example the Python ``multiprocessing`` module, it's usually required that a
-separate :class:`.Engine` be used for each child process. This is because the
+separate :class:`.Engine` be used for each child process, or
+that the :meth:`.Engine.dispose` method is called at the start of
+the process to flush existing connections. This is because the
 :class:`.Engine` maintains a reference to a connection pool that ultimately
 references DBAPI connections - these tend to not be portable across process
 boundaries. An :class:`.Engine` that is configured not to use pooling (which