From b9874cd751e2987f9553afe8322e773afe4f46aa Mon Sep 17 00:00:00 2001
From: Ian Barwick <ian@2ndquadrant.com>
Date: Mon, 6 Jul 2020 09:53:23 +0900
Subject: [PATCH] doc: clarify "node rejoin" usage

Emphasize that conninfo must be provided for a running node.
---
 doc/repmgr-node-rejoin.xml | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/doc/repmgr-node-rejoin.xml b/doc/repmgr-node-rejoin.xml
index 69427a4f..dbcb5d6b 100644
--- a/doc/repmgr-node-rejoin.xml
+++ b/doc/repmgr-node-rejoin.xml
@@ -43,7 +43,12 @@
       <programlisting>
       repmgr node rejoin -d '$conninfo'</programlisting>
 
-      where <literal>$conninfo</literal> is the conninfo string of any reachable node in the cluster.
+      where <literal>$conninfo</literal> is the PostgreSQL <literal>conninfo</literal> string of the
+      <emphasis>current</emphasis> primary node (or that of any reachable node in the cluster, but
+      <emphasis>not</emphasis> the local node). This is so that &repmgr; can fetch up-to-date information
+      about the current state of the cluster.
+    </para>
+    <para>
       <filename>repmgr.conf</filename> for the stopped node *must* be supplied explicitly if not
       otherwise available.
     </para>