Chapter 2Introduction to the SMCT Installation and Deployment Process
The SMCT installation and deployment
is a three-stage process. At each stage, you configure the cluster environment,
including the software, to generate flash archives. The software distributions
and configuration files are used by the SMCT commands to generate the flash
archives and software loads.
For details on the process of installing and deploying the software
by using the SMCT, see the following sections:
Overview of the SMCT Installation and Deployment Process
To install and deploy the software by using the SMCT, you
need a build server, a prototype machine, and an installation server. One system
can be used both as an installation server and as a build server. A master-eligible
or dataless node from the cluster can be used as a prototype machine. For
more information on the hardware required, see the Netra High Availability Suite Foundation Services 2.1 6/03 Hardware Guide.
The following diagram illustrates the configuration stages of the installation
and deployment process. Each configuration stage is described in detail in
the following sections.
Figure 2-1 Overview of the SMCT Installation and Deployment Process
 Configuration Stage 1: Creating a Generic Flash Archive
In stage 1 of the installation and deployment process, you configure
your cluster environment and generate a generic flash archive.
A generic flash archive is not adapted to any particular cluster
and is not deployable.
The following diagram illustrates stage 1.
Figure 2-2 Overview of Configuration Stage 1
 Stage 1 Configuration Tasks
To complete stage 1, you perform the following configuration tasks.
You need the indicated software and configuration files to create the software
repository, software load respository, and the SMCT environment.
You need the software distributions that are to be installed
on your cluster:
Foundation Services packages and
patches, consisting of the runtime packages and patches supplied with the Foundation Services.
Solaris packages and patches, which are in the Solaris distribution
that is included as part of the Solaris operating system.
Optionally, user application packages and patches. These packages
and patches are processed as Solaris add-on products.
Optionally, third-party packages and patches. These packages
and patches are processed as Solaris add-on products.
For information about installing the distribution software, see Installing Software on the Installation Hardware.
You configure the following cluster
configuration files to define your cluster topology and hardware
environment:
Cluster model configuration file
The cluster.conf configuration file contains a logical view of the cluster
in terms of nodes, node groups, domain, and services. Each node
in the cluster belongs to a node group. A node group
is a group of nodes of the same type, with the same configuration. A node
group can contain master-eligible nodes, diskless nodes, or dataless nodes,
but not a combination of diskless nodes and dataless nodes.
For more information, refer to the cluster.conf(4) man page.
Machine model configuration file
The machine.conf
configuration file describes the cluster hardware and hardware components.
This file also contains definitions of disk partitions and file systems for
master-eligible nodes and dataless nodes. For more information, refer to the machine.conf(4)
man page.
For information about how to configure these files, see Chapter 5, Configuration Stage 1: Creating a Generic Flash Archive.
For information about defining node groups, see Defining Nodes and Node Groups.
You also configure
the Solaris JumpStart utility to define the Solaris JumpStart environment
for the prototype machine.
Finally, you configure the software
configuration files. These files enable you to define the packages
and patches for additional applications that you might want to install on
node groups in a cluster. For more information on software configuration files,
refer to the software.conf(4) man page. See also Creating a Software Configuration File for a Node Group.
After you define your cluster and software in the
configuration files, you use the following SMCT commands to create a software load that contains the Foundation Services, the Solaris operating system, and
optionally, user applications:
The software load contains the software
for each node group. The node
group software is transferred to the prototype machine from the build server
by using the Solaris JumpStart mechanism. The Solaris environment, the Foundation Services
packages, and any user applications are installed on the prototype machine,
but they are not started.
|