Author: buildbot
Date: Tue Nov 11 19:13:31 2014
New Revision: 928822
Log:
Staging update by buildbot for slider
Modified:
websites/staging/slider/trunk/content/ (props changed)
websites/staging/slider/trunk/content/docs/high_availability.html
Propchange: websites/staging/slider/trunk/content/
------------------------------------------------------------------------------
--- cms:source-revision (original)
+++ cms:source-revision Tue Nov 11 19:13:31 2014
@@ -1 +1 @@
-1638219
+1638285
Modified: websites/staging/slider/trunk/content/docs/high_availability.html
==============================================================================
--- websites/staging/slider/trunk/content/docs/high_availability.html (original)
+++ websites/staging/slider/trunk/content/docs/high_availability.html Tue Nov
11 19:13:31 2014
@@ -156,10 +156,102 @@ Latest release: <strong>0.40-incubating<
<h1 class="title"></h1>
- <h1
id="high-availability-setup-rm-ha-nn-ha-work-preserving-rm-and-work-preserving-nm">High
Availability: Setup RM HA, NN HA, Work Preserving RM and Work Preserving
NM</h1>
-<p>Setting up High Availability ensures uninterrupted service provided by long
running applications installed by Slider
-in the event of Resource Manager or Name Node failure. This document provides
an overview of High Availability
-of YARN's ResourceManager, and details how to configure and use this feature.
</p>
+ <h1
id="high-availability-setup-resource-manager-ha-name-node-ha-work-preserving-resource-manager-restart-and-work-preserving-node-manager-restart">High
Availability: Setup Resource Manager HA, Name Node HA, work preserving
Resource Manager restart and work preserving Node Manager restart.</h1>
+<p>Setting up High Availability ensures uninterrupted service provided
+by long running applications installed by Slider in the event of any or all of
Resource Manager, Name Node
+and Node Manager failure. This document provides
+details on how to configure YARN's RM HA.</p>
+<p>Following are the properties required to be set as YARN properties in order
to setup:</p>
+<h3 id="resource-manager-ha">Resource Manager HA</h3>
+<div class="codehilite"><pre> <span class="nt"><property></span>
+ <span
class="nt"><name></span>yarn.resourcemanager.ha.enabled<span
class="nt"></name></span>
+ <span class="nt"><value></span>true<span
class="nt"></value></span>
+ <span class="nt"></property></span>
+
+ <span class="nt"><property></span>
+ <span
class="nt"><name></span>yarn.resourcemanager.ha.rm-ids<span
class="nt"></name></span>
+ <span class="nt"><value></span>rm1,rm2<span
class="nt"></value></span>
+ <span class="nt"></property></span>
+
+ <span class="nt"><property></span>
+ <span
class="nt"><name></span>yarn.resourcemanager.hostname.rm1<span
class="nt"></name></span>
+ <span class="nt"><value></span>192.168.1.9<span
class="nt"></value></span>
+ <span class="nt"></property></span>
+
+ <span class="nt"><property></span>
+ <span
class="nt"><name></span>yarn.resourcemanager.hostname.rm2<span
class="nt"></name></span>
+ <span class="nt"><value></span>192.168.1.10<span
class="nt"></value></span>
+ <span class="nt"></property></span>
+
+ <span class="nt"><property></span>
+ <span
class="nt"><name></span>yarn.resourcemanager.recovery.enabled<span
class="nt"></name></span>
+ <span class="nt"><value></span>true<span
class="nt"></value></span>
+ <span class="nt"></property></span>
+
+ <span class="nt"><property></span>
+ <span
class="nt"><name></span>yarn.resourcemanager.store.class<span
class="nt"></name></span>
+ <span
class="nt"><value></span>org.apache.hadoop.yarn.server.resourcemanager.recovery.ZKRMStateStore<span
class="nt"></value></span>
+ <span class="nt"></property></span>
+
+ <span class="nt"><property></span>
+ <span
class="nt"><name></span>yarn.resourcemanager.zk-address<span
class="nt"></name></span>
+ <span
class="nt"><value></span>192.168.1.9:2181,192.168.1.10:2181<span
class="nt"></value></span>
+ <span class="nt"><description></span>For multiple zk services,
separate them with comma<span class="nt"></description></span>
+ <span class="nt"></property></span>
+
+ <span class="nt"><property></span>
+ <span
class="nt"><name></span>yarn.resourcemanager.cluster-id<span
class="nt"></name></span>
+ <span class="nt"><value></span>yarn-cluster<span
class="nt"></value></span>
+ <span class="nt"></property></span>
+</pre></div>
+
+
+<h3 id="name-node-ha">Name Node HA</h3>
+<div class="codehilite"><pre> <span class="n">TODO</span>
+</pre></div>
+
+
+<h3 id="work-preserving-rm-restart">Work Preserving RM Restart</h3>
+<div class="codehilite"><pre> <span class="nt"><property></span>
+ <span class="nt"><description></span>Enable RM to recover
state after starting. If true, then yarn.resourcemanager.store.class must be
specified<span class="nt"></description></span>
+ <span
class="nt"><name></span>yarn.resourcemanager.recovery.enabled<span
class="nt"></name></span>
+ <span class="nt"><value></span>true<span
class="nt"></value></span>
+ <span class="nt"></property></span>
+
+ <span class="nt"><property></span>
+ <span class="nt"><description></span>Enable RM work preserving
recovery. This configuration is private to YARN for experimenting the feature.
NOTE: this config has to be set on both RM and ALL NMs. <span
class="nt"></description></span>
+ <span
class="nt"><name></span>yarn.resourcemanager.workÂpreservingÂrecovery.enabled<span
class="nt"></name></span>
+ <span class="nt"><value></span>true<span
class="nt"></value></span>
+ <span class="nt"></property></span>
+
+ <span class="nt"><property></span>
+ <span class="nt"><description></span>The class to use as the
persistent store.<span class="nt"></description></span>
+ <span
class="nt"><name></span>yarn.resourcemanager.store.class<span
class="nt"></name></span>
+ <span
class="cp"><!ÂÂvalue></span>org.apache.hadoop.yarn.server.resourcemanager.recovery.FileSystemRMStateStore<span
class="err"><</span>/valueÂ-->
+ <span
class="nt"><value></span>org.apache.hadoop.yarn.server.resourcemanager.recovery.ZKRMStateStore<span
class="nt"></value></span>
+ <span class="nt"></property></span>
+
+ <span class="nt"><property></span>
+ <span class="nt"><description></span>Host:Port of the
ZooKeeper server where RM state will be stored. This must be supplied when
using org.apache.hadoop.yarn.server.resourcemanager.recovery.ZKRMStateStore as
the value for yarn.resourcemanager.store.class<span
class="nt"></description></span>
+ <span
class="nt"><name></span>yarn.resourcemanager.zkÂaddress<span
class="nt"></name></span>
+ <span class="nt"><value></span>127.0.0.1:2181<span
class="nt"></value></span>
+ <span class="nt"></property></span>
+</pre></div>
+
+
+<h3 id="work-preserving-nm-restart">Work Preserving NM Restart</h3>
+<div class="codehilite"><pre> <span class="nt"><property></span>
+ <span class="nt"><description></span>Enable the node manager
to recover after starting<span class="nt"></description></span>
+ <span
class="nt"><name></span>yarn.nodemanager.recovery.enabled<span
class="nt"></name></span>
+ <span class="nt"><value></span>false<span
class="nt"></value></span>
+ <span class="nt"></property></span>
+
+ <span class="nt"><property></span>
+ <span class="nt"><description></span>The local filesystem
directory in which the node manager will store state when recovery is
enabled.<span class="nt"></description></span>
+ <span
class="nt"><name></span>yarn.nodemanager.recovery.dir<span
class="nt"></name></span>
+ <span class="nt"><value></span><span class="cp">${</span><span
class="n">hadoop</span><span class="o">.</span><span class="n">tmp</span><span
class="o">.</span><span class="n">dir</span><span
class="cp">}</span>/yarn-nm-recovery<span class="nt"></value></span>
+ <span class="nt"></property></span>
+</pre></div>
</div>
<div id="footer">
