Chapter 1Introduction to the Foundation Services
This chapter introduces the Foundation Services. For a brief description
of each of the Foundation Services, the installation tools, and the cluster administration
tools, see the following sections:
What Are the Foundation Services?
The Foundation Services are a suite of reliable software services that run
on the SPARC Solaris operating system. The Foundation Services enable you to deploy applications
in a continuous availability environment. The Foundation Services can be used to create
a highly available, dynamically scalable cluster of distributed nodes, or
to augment existing highly available frameworks.
The following figure illustrates a basic Foundation Services cluster.
Figure 1-1 Basic Foundation Services Cluster
 The concepts of cluster, master node, vice-master node, diskless node,
and dataless node are described in Cluster Model.
The Foundation Services have been designed to support the following:
Hardware
replacement or upgrade, and diagnostics without incurring system outage
Redundant services
Redundant dual Ethernet links
Redundant platform services such as Reliable NFS and the Reliable
Boot Service
High-Level View of the Foundation Services
The following figure shows a high-level view of the Foundation Services architecture.
Figure 1-2 High-Level View of the Foundation Services Architecture
 The Foundation Services software offers the following services: A reliable IP transport mechanism provided by the Carrier
Grade Transport Protocol (CGTP). CGTP limits the consequences of single network
failure by duplicating the communication links. For further information, see Chapter 6, Carrier Grade Transport Protocol.
Reliable NFS to ensure that data
is accessible to clients, even in the event of hardware or software failure.
Reliable NFS uses mounted file systems, IP mirroring of disk-based data, and
IP address failover of the master role. For further information, see Chapter 7, File Sharing and Data Replication.
A Cluster Membership
Manager to provide a global view of the cluster. The Cluster Membership Manager
determines which nodes are members of the cluster. It assigns the roles and
attributes of nodes, detects the failure of nodes, and notifies clients of
changes to the cluster. A heartbeat mechanism is used to detect node failure.
For further information, see Chapter 8, Cluster Membership Manager.
A Reliable Boot Service and the Solaris Dynamic
Host Configuration Protocol service to ensure the boot of diskless nodes regardless
of software or hardware failures. For further information, see Chapter 9, Reliable Boot Service.
A Node State Manager with scripts that provide access for external networks
to the node holding the master role. For further information, see Chapter 5, External Addressing.
A Daemon Monitor to survey Foundation Services daemons, many Solaris operating system
daemons, and some companion product daemons. If any of the monitored daemons
fail, the Daemon Monitor initiates a recovery response. The Daemon Monitor
is itself monitored by the Node Management Agent. For further information,
see Chapter 10, Daemon Monitor.
A Node Management Agent to monitor cluster statistics. The Node Management
Agent can initiate a switch to the backup node, change some error recovery
responses, and listen for notifications of some cluster events.
The Node Management Agent is compliant with the Java
Management Extensions (JMX) and based on the Java Dynamic
Management Kit. For further information, see Chapter 11, Node Management Agent.
A Watchdog Timer for low-level system
monitoring of the Foundation Services. For further information, see Chapter 12, Watchdog Timer.
Each of these services is described in detail in Part II, Description of the Foundation Services.
Foundation Services Tools
The Foundation Services
include a suite of tools to facilitate installation and cluster administration.
There tools and their purpose are as follows:
nhadm | Perform
administration tasks on a cluster
| nhcmmqualif | Qualify the current node as master
| nhcmmrole | get
the role of the current node
| nhcmmstat | display
information about peer nodes, trigger a switchover, or force the qualification
of a master-eligible node
| nhcrfsadm | command
line tool for Reliable NFS administration
| nhenablesync | trigger disk synchronization
| nhinstall | install
and configure the Foundation Services
| nhpmdadm | process
monitor daemon administration tool
|
For information about using nhinstall,
see the Netra High Availability Suite Foundation Services 2.1 6/03 Custom Installation Guide.
In addition to the Foundation Services tools, many Solaris tools can be used for
administration of a cluster. For information about using these tools, see
the Netra High Availability Suite Foundation Services 2.1 6/03 Cluster Administration Guide.
|