AN!Cluster Tutorial 3 (using crm) - Revision history

Digimer: /* Installing crmsh */

2013-09-30T23:16:41Z

Installing crmsh

Digimer: /* Configuring IPMI Fencing */

2013-02-20T00:06:24Z

Configuring IPMI Fencing

← Older revision		Revision as of 00:06, 20 February 2013
Line 1,063:		Line 1,063:

	<source lang="text">		<source lang="text">
	configure primitive fence_an-c03n01 stonith::fence_ipmilan params pcmk_host_list="an-c03n01" ipaddr="an-c03n01.ipmi" login="admin" passwd="admin" op monitor interval="60"		configure primitive fence_an-c03n01 stonith::fence_ipmilan params pcmk_host_list="an-c03n01" ipaddr="an-c03n01.ipmi" login="admin" passwd="admin" pcmk_host_check="static-list" op monitor interval="60"
	configure primitive fence_an-c03n02 stonith::fence_ipmilan params pcmk_host_list="an-c03n02" ipaddr="an-c03n02.ipmi" login="admin" passwd="admin" op monitor interval="60"		configure primitive fence_an-c03n02 stonith::fence_ipmilan params pcmk_host_list="an-c03n02" ipaddr="an-c03n02.ipmi" login="admin" passwd="admin" pcmk_host_check="static-list" op monitor interval="60"
	</source>		</source>

Digimer: /* Configuring Fencing aka STONITH */

2013-02-19T23:46:48Z

Configuring Fencing aka STONITH

@@ Line 843: / Line 843: @@
    </status>
 </cib>
 </source>

Digimer: /* Configuring Fencing aka STONITH */

2013-02-19T23:37:37Z

Configuring Fencing aka STONITH

← Older revision		Revision as of 23:37, 19 February 2013
Line 1,050:		Line 1,050:
	</source>		</source>

	Now configure the fence agent with the arguments needed to kill <span class="code">an-c03n01</span>;		Now configure pacemaker to call the fence agent with the arguments needed to kill <span class="code">an-c03n01</span> and <span class="code">an-c03n02</span>;

	<source lang="text">		<source lang="text">
	primitive fence_an-c03n01 stonith:~~external/ipmi~~ params ~~hostname~~="an-c03n01" ipaddr="an-c03n01.ipmi" ~~userid~~="admin" passwd="admin" op monitor interval="60"		configure primitive fence_an-c03n01 stonith::fence_ipmilan params pcmk_host_list="an-c03n01" ipaddr="an-c03n01.ipmi" login="admin" passwd="admin" op monitor interval="60"
			configure primitive fence_an-c03n02 stonith::fence_ipmilan params pcmk_host_list="an-c03n02" ipaddr="an-c03n02.ipmi" login="admin" passwd="admin" op monitor interval="60"
	</source>		</source>

			Next, we have to enable [[stonith]];

			<source lang="text">
			configure property stonith-enabled="true"
			</source>

			Lastly, push the <span class="code">stonith</span> shadow [[CIB]] over to the active one.

			<source lang="text">
			cib commit stonith
			</source>
			<source lang="text">
			INFO: commited 'stonith' shadow CIB to the cluster
			</source>

	~~Lastly,~~ we ~~have~~ to ~~enable [[stonith]];~~		Now we can exit back to the shell.

	<source lang="text">		<source lang="text">
	~~configure property stonith-enabled~~="~~true~~"		quit
			</source>
			<source lang="text">
			bye
	</source>		</source>

Digimer: /* Configuring Fencing aka STONITH */

2013-02-19T22:08:02Z

Configuring Fencing aka STONITH

Show changes

Digimer: /* Configuring Pacemaker */

2013-02-19T19:00:57Z

Configuring Pacemaker

Show changes

Digimer: /* Configuring Pacemaker */

2013-02-19T18:25:08Z

Configuring Pacemaker

← Older revision		Revision as of 18:25, 19 February 2013
Line 645:		Line 645:

	The remainder of this tutorial will revolve around configuring pacemaker.		The remainder of this tutorial will revolve around configuring pacemaker.

			== Initializing Pacemaker ==

			First, start <span class="code">pacemakerd</span> on both nodes.

			<source lang="bash">
			systemctl start pacemaker.service
			</source>

			If you watch syslog, you will see messages like:

			<source lang="text">
			Feb 19 13:17:16 an-c03n01 systemd[1]: Starting Pacemaker High Availability Cluster Manager...
			Feb 19 13:17:16 an-c03n01 systemd[1]: Started Pacemaker High Availability Cluster Manager.
			Feb 19 13:17:16 an-c03n01 pacemakerd[1387]: Could not establish pacemakerd connection: Connection refused (111)
			Feb 19 13:17:16 an-c03n01 pacemakerd[1387]: notice: crm_add_logfile: Additional logging available in /var/log/cluster/corosync.log
			Feb 19 13:17:16 an-c03n01 pacemakerd[1387]: notice: main: Starting Pacemaker 1.1.9-0.1318.a7966fb.git.fc18 (Build: a7966fb): generated-manpages agent-manpages ncurses libqb-logging libqb-ipc upstart systemd nagios corosync-native
			Feb 19 13:17:16 an-c03n01 pacemakerd[1387]: notice: corosync_node_name: Unable to get node name for nodeid 0
			Feb 19 13:17:16 an-c03n01 pacemakerd[1387]: notice: get_local_node_name: Defaulting to uname -n for the local corosync node name
			Feb 19 13:17:16 an-c03n01 pacemakerd[1387]: notice: update_node_processes: 0x1602f40 Node 186520586 now known as an-c03n01.alteeve.ca, was:
			Feb 19 13:17:16 an-c03n01 pengine[1393]: notice: crm_add_logfile: Additional logging available in /var/log/cluster/corosync.log
			Feb 19 13:17:16 an-c03n01 attrd[1392]: notice: crm_add_logfile: Additional logging available in /var/log/cluster/corosync.log
			Feb 19 13:17:16 an-c03n01 attrd[1392]: notice: crm_cluster_connect: Connecting to cluster infrastructure: corosync
			Feb 19 13:17:16 an-c03n01 stonith-ng[1390]: notice: crm_add_logfile: Additional logging available in /var/log/cluster/corosync.log
			Feb 19 13:17:16 an-c03n01 stonith-ng[1390]: notice: crm_cluster_connect: Connecting to cluster infrastructure: corosync
			Feb 19 13:17:16 an-c03n01 cib[1388]: notice: crm_add_logfile: Additional logging available in /var/log/cluster/corosync.log
			Feb 19 13:17:16 an-c03n01 cib[1388]: notice: main: Using new config location: /var/lib/pacemaker/cib
			Feb 19 13:17:16 an-c03n01 cib[1388]: warning: retrieveCib: Cluster configuration not found: /var/lib/pacemaker/cib/cib.xml
			Feb 19 13:17:16 an-c03n01 cib[1388]: warning: readCibXmlFile: Primary configuration corrupt or unusable, trying backup...
			Feb 19 13:17:16 an-c03n01 cib[1388]: warning: readCibXmlFile: Continuing with an empty configuration.
			Feb 19 13:17:16 an-c03n01 pacemakerd[1387]: notice: update_node_processes: 0x1908a10 Node 203297802 now known as an-c03n02.alteeve.ca, was:
			Feb 19 13:17:16 an-c03n01 attrd[1392]: notice: corosync_node_name: Unable to get node name for nodeid 186520586
			Feb 19 13:17:16 an-c03n01 attrd[1392]: notice: get_local_node_name: Defaulting to uname -n for the local corosync node name
			Feb 19 13:17:16 an-c03n01 stonith-ng[1390]: notice: corosync_node_name: Unable to get node name for nodeid 186520586
			Feb 19 13:17:16 an-c03n01 attrd[1392]: notice: main: Starting mainloop...
			Feb 19 13:17:16 an-c03n01 stonith-ng[1390]: notice: get_local_node_name: Defaulting to uname -n for the local corosync node name
			Feb 19 13:17:16 an-c03n01 lrmd[1391]: notice: crm_add_logfile: Additional logging available in /var/log/cluster/corosync.log
			Feb 19 13:17:16 an-c03n01 crmd[1394]: notice: crm_add_logfile: Additional logging available in /var/log/cluster/corosync.log
			Feb 19 13:17:16 an-c03n01 crmd[1394]: notice: main: CRM Git Version: a7966fb
			Feb 19 13:17:16 an-c03n01 cib[1388]: notice: crm_cluster_connect: Connecting to cluster infrastructure: corosync
			Feb 19 13:17:16 an-c03n01 cib[1388]: notice: corosync_node_name: Unable to get node name for nodeid 186520586
			Feb 19 13:17:16 an-c03n01 cib[1388]: notice: get_local_node_name: Defaulting to uname -n for the local corosync node name
			Feb 19 13:17:17 an-c03n01 stonith-ng[1390]: notice: setup_cib: Watching for stonith topology changes
			Feb 19 13:17:17 an-c03n01 crmd[1394]: notice: crm_cluster_connect: Connecting to cluster infrastructure: corosync
			Feb 19 13:17:17 an-c03n01 crmd[1394]: notice: corosync_node_name: Unable to get node name for nodeid 186520586
			Feb 19 13:17:17 an-c03n01 crmd[1394]: notice: get_local_node_name: Defaulting to uname -n for the local corosync node name
			Feb 19 13:17:17 an-c03n01 crmd[1394]: notice: init_quorum_connection: Quorum acquired
			Feb 19 13:17:17 an-c03n01 crmd[1394]: notice: crm_update_peer_state: pcmk_quorum_notification: Node an-c03n01.alteeve.ca[186520586] - state is now member
			Feb 19 13:17:17 an-c03n01 crmd[1394]: notice: corosync_node_name: Unable to get node name for nodeid 203297802
			Feb 19 13:17:17 an-c03n01 crmd[1394]: notice: crm_update_peer_state: pcmk_quorum_notification: Node (null)[203297802] - state is now member
			Feb 19 13:17:17 an-c03n01 crmd[1394]: notice: do_started: The local CRM is operational
			</source>

			You can see that <span class="code">pacemakerd</span> is running using <span class="code">ps</span>.

			<source lang="bash">
			ps axf
			</source>
			<source lang="text">
			...
			1336 ? Ssl 0:10 corosync
			1387 ? Ssl 0:00 /usr/sbin/pacemakerd -f
			1388 ? Ssl 0:00 \_ /usr/libexec/pacemaker/cib
			1390 ? Ss 0:00 \_ /usr/libexec/pacemaker/stonithd
			1391 ? Ss 0:00 \_ /usr/libexec/pacemaker/lrmd
			1392 ? Ss 0:00 \_ /usr/libexec/pacemaker/attrd
			1393 ? Ss 0:00 \_ /usr/libexec/pacemaker/pengine
			1394 ? Ss 0:00 \_ /usr/libexec/pacemaker/crmd
			</source>

			You can see the initial <span class="code">cib.xml</span> using <span class="code">cibadmin</span>.

			<source lang="bash">
			cibadmin --query --local
			</source>
			<source lang="xml">
			<cib epoch="4" num_updates="6" admin_epoch="0" validate-with="pacemaker-1.2" crm_feature_set="3.0.7" cib-last-written="Tue Feb 19 13:17:38 2013" update-origin="an-c03n02.alteeve.ca" update-client="crmd" have-quorum="1" dc-uuid="203297802">
			<configuration>
			<crm_config>
			<cluster_property_set id="cib-bootstrap-options">
			<nvpair id="cib-bootstrap-options-dc-version" name="dc-version" value="1.1.9-0.1318.a7966fb.git.fc18-a7966fb"/>
			<nvpair id="cib-bootstrap-options-cluster-infrastructure" name="cluster-infrastructure" value="corosync"/>
			</cluster_property_set>
			</crm_config>
			<nodes>
			<node id="203297802" uname="an-c03n02.alteeve.ca"/>
			<node id="186520586" uname="an-c03n01.alteeve.ca"/>
			</nodes>
			<resources/>
			<constraints/>
			</configuration>
			<status>
			<node_state id="203297802" uname="an-c03n02.alteeve.ca" in_ccm="true" crmd="online" crm-debug-origin="do_state_transition" join="member" expected="member">
			<lrm id="203297802">
			<lrm_resources/>
			</lrm>
			<transient_attributes id="203297802">
			<instance_attributes id="status-203297802">
			<nvpair id="status-203297802-probe_complete" name="probe_complete" value="true"/>
			</instance_attributes>
			</transient_attributes>
			</node_state>
			<node_state id="186520586" uname="an-c03n01.alteeve.ca" in_ccm="true" crmd="online" crm-debug-origin="do_state_transition" join="member" expected="member">
			<lrm id="186520586">
			<lrm_resources/>
			</lrm>
			<transient_attributes id="186520586">
			<instance_attributes id="status-186520586">
			<nvpair id="status-186520586-probe_complete" name="probe_complete" value="true"/>
			</instance_attributes>
			</transient_attributes>
			</node_state>
			</status>
			</cib>
			</source>



	<span class="code"></span>		<span class="code"></span>

Digimer: /* Configuring Corosync */

2013-02-17T21:10:43Z

Configuring Corosync

@@ Line 625: / Line 625: @@
 runtime.totem.pg.mrp.srp.members.35525642.join_count (u32) = 1
 runtime.totem.pg.mrp.srp.members.35525642.status (str) = joined
 </source>

Digimer: /* Configuring Corosync */

2013-02-17T21:08:42Z

Configuring Corosync

@@ Line 544: / Line 544: @@
 total size is 2897  speedup is 0.96
 </source>
 = Configuring Pacemaker =

Digimer: /* Configuring Corosync */

2013-02-17T20:36:25Z

Configuring Corosync

← Older revision		Revision as of 20:36, 17 February 2013
Line 437:		Line 437:
	Now let's edit it.		Now let's edit it.

	The main three values to check are;		The main three values to check/change are;
	* <span class="code">mcastaddr</span>; Set the multicast address to use in this cluster. We will use <span class="code">239.255.1.1</span>, which is the one already set in the example config.		* <span class="code">mcastaddr</span>; Set the multicast address to use in this cluster. We will use <span class="code">239.255.1.1</span>, which is the one already set in the example config.
	* <span class="code">mcastport</span>; Set the multicast port to use in this cluster. We will use <span class="code">5405</span>, which is the one already set in the example config.		* <span class="code">mcastport</span>; Set the multicast port to use in this cluster. We will use <span class="code">5405</span>, which is the one already set in the example config.
	* <span class="code">bindnetaddr</span>; this tells corosync which network to use. Our [[BCN]] is <span class="code">10.20.0.0</span>/<span class="code">255.255.0.0</span>, so we will set this to <span class="code">10.20.0.0</span>. This is the only value we need to change from the default configuration.		* <span class="code">bindnetaddr</span>; this tells corosync which network to use. Our [[BCN]] is <span class="code">10.20.0.0</span>/<span class="code">255.255.0.0</span>, so we will set this to <span class="code">10.20.0.0</span>. This is the only value we need to change from the default configuration.

			We also need to enable the [[quorum]] provider <span class="code">[[corosync_votequorum]]</span> and tell it how many votes to expect (one vote per node, two nodes).

	<source lang="bash">		<source lang="bash">
Line 525:		Line 527:
	# Enable and configure quorum subsystem (default: off)		# Enable and configure quorum subsystem (default: off)
	# see also corosync.conf.5 and votequorum.5		# see also corosync.conf.5 and votequorum.5
	#provider: corosync_votequorum		provider: corosync_votequorum
			expected_votes: 2
	}		}
	</source>		</source>

			Copy the <span class="code">corosync.conf</span> file to the other node.

			<source lang="bash">
			rsync -av /etc/corosync/corosync.conf root@an-c03n02:/etc/corosync/
			</source>
			<source lang="text">
			sending incremental file list
			corosync.conf

			sent 2978 bytes received 31 bytes 6018.00 bytes/sec
			total size is 2897 speedup is 0.96
			</source>

			= Configuring Pacemaker =

			The remainder of this tutorial will revolve around configuring pacemaker.

	<span class="code"></span>		<span class="code"></span>