A Serviceguard cluster is a networked grouping of HP ProLiant and HPÂ Integrity servers (host systems known as nodes) having sufficientÂ redundancy of software and hardware that a single point of failureÂ will not significantly disrupt service. Application services (individualÂ Linux processes) are grouped together in packages; in the event of aÂ single service, node, network, or other resource failure, Serviceguard canÂ automatically transfer control of the package to another node within theÂ cluster, allowing services to remain available with minimal interruption.Â Figure 1-1 Conceptual representation of a Serviceguard clusterÂ configuration with two nodes.
In the figure, node 1 (one of two SPU’s) is running package A, and node 2Â is running package B. Each package has a separate group of disksÂ associated with it, containing data needed by the package’s applications,Â and a copy of the data. Note that both nodes are physically connected toÂ disk arrays. However, only one node at a time may access the data for aÂ given group of disks. In the figure, node 1 is shown with exclusive accessÂ to the top two disks (solid line), and node 2 is shown as connectedÂ without access to the top disks (dotted line). Similarly, node 2 is shownÂ with exclusive access to the bottom two disks (solid line), and node 1 isÂ shown as connected without access to the bottom disks (dotted line).
Disk arrays provide redundancy in case of disk failures. In addition, aÂ total of four data buses are shown for the disks that are connected toÂ node 1 and node 2. This configuration provides the maximumÂ redundancy and also gives optimal I/O performance, since each packageÂ is using different buses.
Note that the network hardware is cabled to provide redundant LANÂ interfaces on each node. Serviceguard uses TCP/IP network services forÂ reliable communication among nodes in the cluster, including theÂ transmission of heartbeat messages, signals from each functioningÂ node which are central to the operation of the cluster. TCP/IP services Â also are used for other types of inter-node communication. (TheÂ heartbeat is explained in more detail in the chapter â€œUnderstandingÂ Serviceguard Software.â€)