Fr

ee

The high performance and stability of CentOS Linux are

the key factors that make CentOS Linux the right Linux
distribution to build high availability solutions on. This book
introduces you to high availability before briefly walking
you through the cluster stack and its layers. The book is
then divided into two parts, part A and part B, based on
CentOS versions 6 and 7 respectively. Each part begins
with the installation and configuration of the Corosync
cluster messaging software with CMAN or with the
Pacemaker cluster resource management software.
You will also be introduced to cluster service configuration
and cluster service management before you configure
fencing or STONITH on CentOS 6 and CentOS 7. By the
end of this book, you will have the skills required to
independently design, implement, and maintain a CentOS
high availability multinode cluster environment.

Who this book is written for

Prepare the CentOS 6 and CentOS 7 operating

systems for cluster software installation
Install, configure, and test cluster stacks on
CentOS 6 (Corosync, CMAN, and RGManager)
and CentOS 7 (Corosync and Pacemaker)

Start, stop, add, and remove cluster nodes

from a cluster on CentOS 6 and CentOS 7

P U B L I S H I N G

pl

C o m m u n i t y

E x p e r i e n c e

D i s t i l l e d

Place a cluster on standby mode to perform

maintenance work on the cluster node
Configure fencing on CentOS 6 and STONITH
on CentOS 7
Successfully test cluster failovers

$ 29.99 US
19.99 UK

community experience distilled

Sa
m

Configure failover domains, cluster resources,

cluster services, and cluster resource groups

Mitja Resman

This book is targeted at system engineers and system

administrators who want to upgrade their knowledge and
skills in high availability and want to learn practically how
to achieve high availability with CentOS Linux. You are
expected to have good CentOS Linux knowledge and basic
networking experience.

What you will learn from this book

CentOS High Availability

CentOS High Availability

CentOS High Availability

Leverage the power of high availability clusters on CentOS Linux,
the enterprise-class, open source operating system

Prices do not include

local sales tax or VAT
where applicable

Visit www.PacktPub.com for books, eBooks,

code, downloads, and PacktLib.

Mitja Resman

In this package, you will find:

The author biography

A preview chapter from the book, Chapter 5 'Playing with Cluster Nodes
on CentOS 6'
A synopsis of the books content
More information on CentOS High Availability

About the Author

Mitja Resman hails from a small, beautiful country called Slovenia in south-central
Europe. He is a fan of Linux and also an open source enthusiast. He is a Red Hat
Certified Engineer and Linux Professional Institute professional. Working as a system
administrator, Mitja gained years of professional experience with open source software
and Linux system administration on local and international projects. The Swiss Army
Knife syndrome makes him an expert in the field of VMware virtualization, Microsoft
system administration, and more recently, Android system administration.
Mitja has a strong desire to learn, develop, and share knowledge with others. This is the
reason he started a blog called
. It provides CentOS Linux guides and
how-to articles covering all sorts of topics appropriate for both beginners and advanced
users. Mitja is also a devoted father and husband. His daughter and wife are the ones
who take his mind off the geek stuff and make him appreciate life and look forward
to the things to come.

CentOS High Availability

The book will guide you through the process of installing, configuring, and administering
a multinode computer cluster on CentOS version 6 and 7 using two different cluster
suites with respect to the CentOS version.

What This Book Covers

Chapter 1, Getting Started with High Availability, provides facts about high availability.
Chapter 2, Meet the Cluster Stack on CentOS, shows you how cluster stack software
works as a whole to provide high availability.
Chapter 3, Cluster Stack Software on CentOS 6, covers installation and configuration of
cluster stack software on CentOS 6.
Chapter 4, Resource Manager on CentOS 6, teaches you how to manage your cluster
resources and services with the RGManager resource manager on CentOS 6.
Chapter 5, Playing with Cluster Nodes on CentOS 6, explains how to manage and
administer cluster nodes on CentOS 6.
Chapter 6, Fencing on CentOS 6, provides the details and recommendations on the
quorum disk on CentOS 6.
Chapter 7, Testing Failover on CentOS 6, discovers cluster node fencing on CentOS 6.
Chapter 8, Two-node Cluster Considerations on CentOS 6, covers some final cluster
failover tests on CentOS 6.
Chapter 9, Cluster Stack Software on CentOS 7, shows you how to install and configure
cluster stack software on CentOS 7.
Chapter 10, Resource Manager on CentOS 7, teaches you how to manage your cluster
resources and services with the Pacemaker resource manager on CentOS 7.
Chapter 11, Playing with Cluster Nodes on CentOS 7, explains how to manage and
administer cluster nodes on CentOS 7.
Chapter 12, STONITH on CentOS 7, introduces the details and recommendations on
quorum disk on CentOS 7.
Chapter 13, Testing Failover on CentOS 7, covers cluster node fencing, the so-called
STONITH, on CentOS 7.
Chapter 14, Two-node Cluster Considerations on CentOS 7, provides some final cluster
failover tests on CentOS 7.

Playing with Cluster Nodes

on CentOS 6
In the previous chapter, you learned how to add, manage, and remove failover
domains, cluster resources, and cluster services with RGManager on CentOS 6.
Since a computer cluster is a live system and is constantly developing and growing
with additional resources and services, it is also very important for you, as a cluster
administrator, to know how to add or remove a cluster node from the cluster if the
situation requires you to.
This chapter focuses on familiarizing you with cluster node management on CentOS
6. You will learn how to add and remove cluster nodes from an existing and running
computer cluster.

Adding a new cluster node

Adding a new cluster node to an existing cluster configuration does not require any
cluster service downtime. By this point, you should be familiar with the process of
installing and configuring cluster stack software on a new CentOS 6 cluster node, so
the steps for adding a new cluster node are not covered in detail.
To add a new cluster node to the existing cluster configuration, follow these steps:
1. Operating system preparation:

Configure network interfaces and networking

Configure NTP time synchronization

Configure DNS resolving

Check the network connectivity between the cluster nodes

[ 57 ]

Playing with Cluster Nodes on CentOS 6

For a detailed procedure on operating system preparation on CentOS 6,

refer to Chapter 3, Cluster Stack Software on CentOS 6.

2. Install and configure Corosync:

Install Corosync on the new cluster node and configure iptables

rules to allow Corosync cluster communication.

Transfer the corosync.conf configuration file and the authkey

authentication key from one of the preexisting cluster nodes to the
new cluster node. In the following screenshot, you can see the scp
command used to transfer the corosync.conf configuration file
and the authkey authentication key to the node-4.geekpeek.net
cluster node:

Edit the Corosync configuration file on the new cluster node and
change the bindnetaddr parameter accordingly. You can download
an example corosync.conf configuration file from the corosyncconf-1.txt file in the code bundle.

The bindnetaddr parameter must be changed to the node-4.

geekpeek.net cluster node's IP address.
For a detailed discussion of Corosync installation and configuration
on CentOS 6, refer to Chapter 3, Cluster Stack Software on CentOS 6.

3. Install and configure CMAN and Ricci:

Install CMAN and Ricci and configure the iptables rules to allow
Ricci cluster communication.

Start Ricci on the new cluster node. In the following screenshot, you
can see the command used to start Ricci on the node-4 cluster node:

[ 58 ]

Chapter 5

Edit the CMAN configuration file on one of the pre-existing cluster

nodes, and add a new cluster node to the <clusternodes> section
and the <failoverdomains> section, if required. Do not forget to
increase the config_version parameter number. You can download
an example CMAN cluster.conf configuration file from the cmanconf-2.txt file in the code bundle.

A new cluster node called node-4 is now added to the cluster

configuration within the <clusternodes> and <failoverdomains>
sections.

Verify the cluster configuration and distribute the new cluster

configuration to all other cluster nodes as follows. In the following
screenshot, you can see the command used to verify the cluster
configuration and propagate it to other cluster nodes:

The CMAN cluster.conf configuration file is now successfully

distributed at the new node-4 cluster node due to the running
Ricci service.

Start the CMAN service on the new cluster node and check the cluster
node's status with the cman_tool nodes command. In the following
screenshot, you can see the output of the cman_tool nodes command:

[ 59 ]

Playing with Cluster Nodes on CentOS 6

Here, you can also see that the new node-4 cluster node has
successfully joined the cluster and is listed with its status as M.
4. Install and configure RGManager:

Install RGManager and configure the iptables rules to allow

RGManager traffic.

Start RGmanager on the new cluster node, as shown in the

following screenshot:

Check the cluster status and confirm that all cluster nodes are listed
as expected. In the following screenshot, you can see the output of
the clustat command, confirming that the node-4 cluster node has
successfully joined the cluster:

[ 60 ]

Chapter 5

Removing a cluster node

Removing a cluster node from an existing cluster configuration is easy, and is done
by turning it off and editing the CMAN configuration file on existing cluster nodes.
You can also just stop the cluster stack software services on the node you want to
remove from a cluster. Removing a cluster node from the cluster configuration does
not require cluster service downtime, unless the cluster service is running on the
cluster node you would like to remove.
Before you start the procedure of removing a cluster node from the
cluster configuration, verify that no cluster resources and services are
running on the node.

The steps required to successfully remove a cluster node from a running cluster
configuration are as follows:
1. Stop the cluster software:

Stop Ricci

Stop RGmanager

Stop CMAN
Stop the cluster software on the cluster node you will remove
from the cluster configuration.

2. Edit the CMAN configuration file:

Remove the cluster node from the <clusternodes> section

Remove the cluster node from the <failoverdomains> section,

if required

Increase the config_version parameter number

Do not edit the CMAN cluster.conf configuration file on the
cluster node you are going to remove from the cluster. Edit the
CMAN configuration file on one of cluster nodes that will stay active.
You can download an example CMAN cluster.conf configuration
file from the cman-conf-3.txt file in the code bundle.

Now the config_version parameter has been increased and the

node-4 cluster node has been removed from the <clusternodes>
and <failoverdomains> sections
[ 61 ]

Playing with Cluster Nodes on CentOS 6

3. Distribute the CMAN configuration file:

Validate the cluster configuration and distribute the CMAN

configuration file to other, still active cluster nodes. In the following
screenshot, you can see the command used to verify the cluster
configuration and distribute the cluster.conf configuration file to
other cluster nodes:

4. Check the Cluster Status parameter to confirm that the cluster node was
removed from the cluster configuration. In the following screenshot, you
can see that the node-4 cluster node has been successfully removed from
the cluster:

The node-4 cluster node is no longer listed and has been successfully
removed from the cluster configuration

[ 62 ]

Chapter 5

Summary
In this chapter, you learned how to add a cluster node to a running cluster
configuration, and how to remove a cluster node from it without any cluster service
disruption or downtime. In the following chapter, you will explore the cluster node
fencing mechanism and how to configure cluster node fencing.

[ 63 ]

Get more information CentOS High Availability

Where to buy this book

You can buy CentOS High Availability from the Packt Publishing website.
Alternatively, you can buy the book from Amazon, BN.com, Computer Manuals and most internet
book retailers.
Click here for ordering and shipping details.

www.PacktPub.com

Stay Connected: