Skip to content

Commit

Permalink
Fix bsc#992742 recommend per-node maintenance mode
Browse files Browse the repository at this point in the history
Title: "Should section 4.3 recommend per-node maintenance mode instead of
whole-cluster maintenance mode?"

Use "crm node" for putting a node into maintenance mode instead of the
whole cluster.

* Use crm node maintenance NODE = set node maintenance mode ON
* Use crm node ready NODE = set node maintenance mode OFF
  • Loading branch information
tomschr committed Aug 16, 2016
1 parent 61164fc commit 462918d
Showing 1 changed file with 7 additions and 6 deletions.
13 changes: 7 additions & 6 deletions xml/ha_migration.xml
Expand Up @@ -626,8 +626,9 @@
<para>
Before starting an update for a node, either <emphasis>stop</emphasis>
the cluster stack <emphasis>on that node</emphasis> or put the
<emphasis>cluster into maintenance mode</emphasis> (see
<xref linkend="step.update.check"/> for details).
<emphasis>cluster into maintenance mode</emphasis>, depending on if the

This comment has been minimized.

Copy link
@gao-yan

gao-yan Aug 16, 2016

Contributor

Probably just "... node into maintenance mode".

cluster stack is affected or not. See <xref linkend="step.update.check"/>
for details.
</para>
<para>
If the cluster resource manager on a node is active during the software
Expand Down Expand Up @@ -661,10 +662,10 @@
<listitem>
<para>
If none of the situations above apply, you do not need to stop the
cluster stack. In that case, put the cluster into maintenance mode
cluster stack. In that case, put the node into maintenance mode
before starting the software update:
</para>
<screen>&prompt.root;<command>crm</command> configure property maintenance-mode=true</screen>
<screen>&prompt.root;<command>crm</command> node maintenance <replaceable>NODE_NAME</replaceable></screen>
<para>
For more details on maintenance mode, see
<xref linkend="sec.ha.config.basics.maint.mode"/>.
Expand All @@ -691,10 +692,10 @@
</listitem>
<listitem>
<para>
or remove the cluster maintenance flag to bring the cluster back to
or remove the maintenance flag to bring the node back to
normal mode:
</para>
<screen>&prompt.root;<command>crm</command> configure property maintenance-mode=false</screen>
<screen>&prompt.root;<command>crm</command> node ready <replaceable>NODE_NAME</replaceable></screen>
</listitem>
</itemizedlist>
</step>
Expand Down

0 comments on commit 462918d

Please sign in to comment.