2 node clustering with corosync/pacemaker

kirukan · 09-01-2012, 10:33 AM

This post may help others, who are trying to setup 2 nodes cluster for Virtual-IP, Service fail-over.

1. Install basic packages from repository

Code:

#yum install pacemaker corosync

2. Generate corosync key and place the key to all the cluster nodes under /etc/corosync with 400 permission

Code:

#corosync-keygen

3. Place the corosync.conf in /etc/corosync

Code:

 
==========================================
compatibility: whitetank
totem {
        version: 2
        token: 3000
        token_retransmits_before_loss_const: 10
        join: 60
        consensus: 3600
        vsftype: none
        max_messages: 20
        secauth: off
        threads: 0
        clear_node_high_bit: yes
        rrp_mode: none
        interface {
                 ringnumber: 0
                 bindnetaddr: network_address (This ip range need to be change according to your network setup)
                 mcastaddr: 226.94.1.1 (Change the IP based on #netstat -g result)
                 mcastport: 5405
                 ttl: 1
                 }
         }

logging {
        fileline: off
        to_stderr: no
        to_logfile: yes
        to_syslog: yes
        syslog_facility: daemon
        logfile: /var/log/cluster/corosync.log
        debug: off
        timestamp: on
        logger_subsys {
                subsys: AMF
                debug: off
                tags: enter|leave|trace1|trace2|trace3|trace4|trace6
                }
        }

 amf {
     mode: disabled
     }
=======================================

4. Place the pcmk under the following path /etc/corosync/service.d/

Code:

service{
        # Load the Pacemaker Cluster Resource Manager
        name: pacemaker
        ver: 1
}

The above is pretty much enough to configure corosync, and there are no separate config file for pacemaker & openais(everything is under corosync.conf)

5. Now start the services
/etc/init.d/corosync start (corosync must start before pacemaker)
/etc/init.d/pacemaker start

6. Verify the configuration

Code:

#crm_verify -L -V

if any errors related to stonith & quorum excute the following to avoid errors

7. Ignore the stonith

Code:

#crm configure property stonith-enabled=false

8.Ignore quorum

Code:

#crm configure property no-quorum-policy=ignore

9. For ocf scripts, Linux-ha resource-agents need to be install

Code:

#yum install resource-agents

can use some other resource agents too..
To use LSB Resource Agents
http://linux-ha.org/wiki/LSB_Resource_Agents

10. The following is the way to add and manage resources

Code:

crm configure

11. Add virtual IPv4

Code:

primitive VIPv4 ocf:heartbeat:IPaddr2 \
params ip="xx.xx.xx.xx" \
nic="bond0" \
cidr_netmask="28" \
op start interval="0s" timeout="60s" \
op monitor interval="5s" timeout="20s" \
op stop interval="0s" timeout="60s"

12. Add virtual IPv6

Code:

primitive VIPv6 ocf:heartbeat:IPv6addr \
params ipv6addr="xxxx:xxxx:xx:xx:xx:xx" \
nic="bond0" \
cidr_netmask="64" \
op start interval="0s" timeout="60s" \
op monitor interval="5s" timeout="20s" \
op stop interval="0s" timeout="60s"

13. Add Apache service

Code:

primitive HTTPD ocf:heartbeat:httpd \
params configfile="/etc/httpd/conf/httpd.conf" \
port="80" \
op start interval="0s" timeout="60s" \
op monitor interval="5s" timeout="20s" \
op stop interval="0s" timeout="60s"

14. Add group

Code:

#group webserver VIPv4 VIPv6 HTTPD

15. To prevent situation when resource apache migrate to node2 and resource IP stays at node1(It happens when apache at node1 hung but network stack works well) we need to make colocation

Code:

#crm configure colocation SPEED_TEST inf: HTTPD VIPv4 VIPv6

16. To make pacemaker start up apache only after IP is set up. In other words describe start up order run:

Code:

#crm configure order START_ORDER inf: VIPv4 VIPv6 HTTPD

17. To check cluster status

Code:

crm status

Note:- If anything wrong please correct...