From ff7620fd7f59f6d40d772f7d5a85c4071d33c197 Mon Sep 17 00:00:00 2001
From: Henning Andersen <33268011+henningandersen@users.noreply.github.com>
Date: Thu, 31 Jan 2019 15:40:06 +0100
Subject: [PATCH] Fail start on obsolete indices documentation (#37786)

Added breaking changes documentation for node start up obsolete indices
detection.

Issue #27073
---
 docs/reference/migration/migrate_7_0.asciidoc     |  2 ++
 .../reference/migration/migrate_7_0/node.asciidoc | 15 +++++++++++++++
 2 files changed, 17 insertions(+)
 create mode 100644 docs/reference/migration/migrate_7_0/node.asciidoc

diff --git a/docs/reference/migration/migrate_7_0.asciidoc b/docs/reference/migration/migrate_7_0.asciidoc
index 313fdfdfafb..25c2e3eef44 100644
--- a/docs/reference/migration/migrate_7_0.asciidoc
+++ b/docs/reference/migration/migrate_7_0.asciidoc
@@ -26,6 +26,7 @@ See also <<release-highlights>> and <<es-release-notes>>.
 * <<breaking_70_restclient_changes>>
 * <<breaking_70_low_level_restclient_changes>>
 * <<breaking_70_logging_changes>>
+* <<breaking_70_node_changes>>
 
 [float]
 === Indices created before 7.0
@@ -60,3 +61,4 @@ include::migrate_7_0/snapshotstats.asciidoc[]
 include::migrate_7_0/restclient.asciidoc[]
 include::migrate_7_0/low_level_restclient.asciidoc[]
 include::migrate_7_0/logging.asciidoc[]
+include::migrate_7_0/node.asciidoc[]
diff --git a/docs/reference/migration/migrate_7_0/node.asciidoc b/docs/reference/migration/migrate_7_0/node.asciidoc
new file mode 100644
index 00000000000..3b8a9d84e76
--- /dev/null
+++ b/docs/reference/migration/migrate_7_0/node.asciidoc
@@ -0,0 +1,15 @@
+[float]
+[[breaking_70_node_changes]]
+=== Node start up
+
+[float]
+==== Nodes with left-behind data or metadata refuse to start
+Repurposing an existing node by changing node.master or node.data to false can leave lingering on-disk metadata and
+data around, which will not be accessible by the node's new role. Beside storing non-accessible data, this can lead
+to situations where dangling indices are imported even though the node might not be able to host any shards, leading
+to a red cluster health. To avoid this,
+
+* nodes with on-disk shard data and node.data set to false will refuse to start
+* nodes with on-disk index/shard data and both node.master and node.data set to false will refuse to start
+
+Beware that such role changes done prior to the 7.0 upgrade could prevent node start up in 7.0.