diff --git a/docs/reference/upgrade/close-ml.asciidoc b/docs/reference/upgrade/close-ml.asciidoc
index c4efddca759c9..ca2d312f2a82b 100644
--- a/docs/reference/upgrade/close-ml.asciidoc
+++ b/docs/reference/upgrade/close-ml.asciidoc
@@ -1,5 +1,16 @@
 [testenv="platinum"]
 
+////////////
+Take us out of upgrade mode after running any snippets on this page.
+
+[source,js]
+--------------------------------------------------
+POST _ml/set_upgrade_mode?enabled=false
+--------------------------------------------------
+// CONSOLE
+// TEARDOWN
+////////////
+
 If your {ml} indices were created earlier than the previous major version, they
 must be reindexed. In those circumstances, there must be no machine learning
 jobs running during the upgrade. 
@@ -29,4 +40,4 @@ POST _ml/set_upgrade_mode?enabled=true
 This method does not persist the absolute latest model state, rather it uses the
 last model state that was automatically saved. By halting the tasks, you avoid
 incurring the cost of managing active jobs during the upgrade and it's quicker
-than stopping {dfeeds} and closing jobs. 
\ No newline at end of file
+than stopping {dfeeds} and closing jobs.