Page MenuHomeClusterLabs Projects
Files F3151494

No OneTemporary
Actions

View Options

diff --git a/html/quickstart-ubuntu.html b/html/quickstart-ubuntu.html
index f32d32f..1dbdf39 100755
--- a/html/quickstart-ubuntu.html
+++ b/html/quickstart-ubuntu.html
@@ -1,177 +1,166 @@
<!DOCTYPE html>
<!--[if lt IE 7]> <html class="no-js lt-ie9 lt-ie8 lt-ie7"> <![endif]-->
<!--[if IE 7]> <html class="no-js lt-ie9 lt-ie8"> <![endif]-->
<!--[if IE 8]> <html class="no-js lt-ie9"> <![endif]-->
<!--[if gt IE 8]><!--> <html class="no-js"> <!--<![endif]-->
<head>
<!--#include virtual="/header.html" -->
<title>Cluster Labs - The Home of Pacemaker</title>
<meta name="description" content="">
</head>
<body>
<!--#include virtual="/banner.html" -->
<section id="main">
<!--#include virtual="/quickstart-common.html" -->
<h1>Ubuntu</h1>
<p>
- Ubuntu has shipped the Red Hat Cluster Suite (RHCS) stack
- for many years, so we will use that here because it is a
- variation that upstream tests regularly.
+ Ubuntu appears to have switched to Corosync 2 for it's LTS releases.
</p>
<p>
We use aptitude to install pacemaker and some other
necessary packages we will need:
</p>
<p class="command">
-[ALL] # aptitude install pacemaker cman fence-agents
+[ALL] # aptitude install pacemaker corosync fence-agents
</p>
<h2>Configure Cluster Membership and Messaging</h2>
<p>
- Since the ccs tool from RHEL does not exist on Ubuntu, we
- well create the CMAN configuration file on both machines
+ Since the pcs tool from RHEL does not exist on Ubuntu, we
+ well create the corosync configuration file on both machines
manually:
</p>
<p class="command">
-[ALL] # cat <<EOF > /etc/cluster/cluster.conf
-&lt;?xml version="1.0"?>
-&lt;cluster config_version="1" name="pacemaker1">
- &lt;logging debug="off"/>
- &lt;clusternodes>
- &lt;clusternode name="node1" nodeid="1">
- &lt;fence>
- &lt;method name="pcmk-redirect">
- &lt;device name="pcmk" port="node1"/>
- &lt;/method>
- &lt;/fence>
- &lt;/clusternode>
- &lt;clusternode name="node2" nodeid="2">
- &lt;fence>
- &lt;method name="pcmk-redirect">
- &lt;device name="pcmk" port="node2"/>
- &lt;/method>
- &lt;/fence>
- &lt;/clusternode>
- &lt;/clusternodes>
- &lt;fencedevices>
- &lt;fencedevice name="pcmk" agent="fence_pcmk"/>
- &lt;/fencedevices>
-&lt;/cluster>
+[ALL] # cat <<EOF > /etc/corosync/corosync.conf
+totem {
+ version: 2
+ secauth: off
+ cluster_name: pacemaker1
+ transport: udpu
+}
+
+nodelist {
+ node {
+ ring0_addr: node1
+ nodeid: 101
+ }
+ node {
+ ring0_addr: node2
+ nodeid: 102
+ }
+}
+
+quorum {
+ provider: corosync_votequorum
+ two_node: 1
+ wait_for_all: 1
+ last_man_standing: 1
+ auto_tie_breaker: 0
+}
+
EOF
</p>
- <p>
- CMAN was originally written for rgmanager and assumes the
- cluster should not start until the node has
- <a href="http://en.wikipedia.org/wiki/Quorum">quorum</a>,
- so before we try to start the cluster, we need to disable
- this behavior:
- </p>
- <p class="command">
-[ALL] # echo "CMAN_QUORUM_TIMEOUT=0" >> /etc/default/cman
- </p>
-
<h2>Start the Cluster</h2>
<p>
On each machine, run:
</p>
<p class="command">
-[ALL] # service cman start
[ALL] # service pacemaker start
</p>
<h2>Set Cluster Options</h2>
<p>
With so many devices and possible topologies, it is nearly
impossible to include Fencing in a document like this.
For now we will disable it.
</p>
<p class="command">
[ONE] # crm configure property stonith-enabled=false
</p>
<p>
One of the most common ways to deploy Pacemaker is in a
2-node configuration. However quorum as a concept makes
no sense in this scenario (because you only have it when
more than half the nodes are available), so we'll disable
it too.
</p>
<p class="command">
[ONE] # crm configure property no-quorum-policy=ignore
</p>
<p>
For demonstration purposes, we will force the cluster to
move services after a single failure:
</p>
<p class="command">
[ONE] # crm configure property migration-threshold=1
</p>
<h2>Add a Resource</h2>
<p>
Lets add a cluster service, we'll choose one doesn't
require any configuration and works everywhere to make
things easy. Here's the command:
</p>
<p class="command">
[ONE] # crm configure primitive my_first_svc ocf:pacemaker:Dummy op monitor interval=120s
</p>
<p>
"<strong>my_first_svc</strong>" is the name the service
will be known as.
</p>
<p>
"<strong>ocf:pacemaker:Dummy</strong>" tells Pacemaker
which script to use
(<a href="https://github.com/ClusterLabs/pacemaker/blob/master/extra/resources/Dummy">Dummy</a>
- an agent that's useful as a template and for guides like
this one), which namespace it is in (pacemaker) and what
standard it conforms to
(<a href="/doc/en-US/Pacemaker/1.1-crmsh/html/Pacemaker_Explained/s-resource-supported.html#_open_cluster_framework">OCF</a>).
</p>
<p>
"<strong>op monitor interval=120s</strong>" tells Pacemaker to
check the health of this service every 2 minutes by
calling the agent's <strong>monitor</strong> action.
</p>
<p>
You should now be able to see the service running using:
</p>
<p class="command">
[ONE] # crm_mon -1
</p>
<h2>Simulate a Service Failure</h2>
<p>
We can simulate an error by telling the service stop
directly (without telling the cluster):
</p>
<p class="command">
[ONE] # crm_resource --resource my_first_svc --force-stop
</p>
<p>
If you now run <strong>crm_mon</strong> in interactive
mode (the default), you should see (within the monitor
interval - 2 minutes) the cluster notice
that <strong>my_first_svc</strong> failed and move it to
another node.
</p>
<h2>Next Steps</h2>
<p>
<ul>
<li>Configure <a href="/doc/en-US/Pacemaker/1.1-pcs/html/Clusters_from_Scratch/ch09.html">Fencing</a></li>
<li>Add more services - see <a href="/doc/en-US/Pacemaker/1.1-crmsh/html/Clusters_from_Scratch/ch06.html">Clusters from Scratch</a> for examples of how to add IP address, Apache and DRBD to a cluster</li>
<li>Learn how to make services <a href="/doc/en-US/Pacemaker/1.1-crmsh/html/Clusters_from_Scratch/_specifying_a_preferred_location.html">prefer a specific host</a></li>
<li>Learn how to make services <a href="/doc/en-US/Pacemaker/1.1-crmsh/html/Clusters_from_Scratch/_ensuring_resources_run_on_the_same_host.html">run on the same host</a></li>
<li>Learn how to make services <a href="/doc/en-US/Pacemaker/1.1-crmsh/html/Clusters_from_Scratch/_controlling_resource_start_stop_ordering.html">start and stop</a> in a specific order</li>
<li>Find out what else Pacemaker can do - see <a href="/doc/en-US/Pacemaker/1.1-crmsh/html/Pacemaker_Explained/index.html">Pacemaker Explained</a> for an comprehensive list of concepts and options</li>
</ul>
</p>
</section>
<!--#include virtual="/footer.html" -->
</body>
</html>

File Metadata

Mime Type
text/x-diff
Expires
Mon, Feb 24, 4:54 AM (1 d, 15 h ago)
Storage Engine
blob
Storage Format
Raw Data
Storage Handle
1463964
Default Alt Text
(7 KB)

Event Timeline