Serviceguard for Linux allows you to create high availability clustersof HP ProLiant and HP Integrity servers. A high availability computersystem allows application services to continue in spite of a hardware orsoftware failure. Highly available systems protect users from softwarefailures as well as from failure of a system processing unit (SPU), disk, orlocal area network (LAN) component. In the event that one componentfails, the redundant component takes over. Serviceguard and other highavailability subsystems coordinate the transfer between components.

A Serviceguard cluster is a networked grouping of HP ProLiant and HP Integrity servers (host systems known as nodes) having sufficient redundancy of software and hardware that a single point of failure will not significantly disrupt service. Application services (individual Linux processes) are grouped together in packages; in the event of a single service, node, network, or other resource failure, Serviceguard can automatically transfer control of the package to another node within the cluster, allowing services to remain available with minimal interruption. Figure 1-1 Conceptual representation of a Serviceguard cluster configuration with two nodes.

In the figure, node 1 (one of two SPU’s) is running package A, and node 2 is running package B. Each package has a separate group of disks associated with it, containing data needed by the package’s applications, and a copy of the data. Note that both nodes are physically connected to disk arrays. However, only one node at a time may access the data for a given group of disks. In the figure, node 1 is shown with exclusive access to the top two disks (solid line), and node 2 is shown as connected without access to the top disks (dotted line). Similarly, node 2 is shown with exclusive access to the bottom two disks (solid line), and node 1 is shown as connected without access to the bottom disks (dotted line).

Disk arrays provide redundancy in case of disk failures. In addition, a total of four data buses are shown for the disks that are connected to node 1 and node 2. This configuration provides the maximum redundancy and also gives optimal I/O performance, since each package is using different buses.

Note that the network hardware is cabled to provide redundant LAN interfaces on each node. Serviceguard uses TCP/IP network services for reliable communication among nodes in the cluster, including the transmission of heartbeat messages, signals from each functioning node which are central to the operation of the cluster. TCP/IP services  also are used for other types of inter-node communication. (The heartbeat is explained in more detail in the chapter “Understanding Serviceguard Software.”)

Post By Editor (2,827 Posts)

Website: →