PRIMECLUSTER Installation and Administration Guide 4.1 (for Solaris(TM) Operating System)

12.1 PRIMECLUSTER Operation Modes

This section explains the cluster operation mode of PRIMECLUSTER, based on SynfinityCluster.
Note that the "cluster service" of SynfinityCluster is relevant to the "cluster application" of PRIMECLUSTER.

Also, "network takeover" of SynfinityCluster relates to the "public LAN takeover" of PRIMECLUSTER.

For the terminologies, see Appendix "Terminology".

Cluster operation mode

Cluster operation mode	Description
1:1 standby	A single standby cluster application runs on a system that consists of two nodes. One is an operating node, and the other is a standby node.
N:1 standby	"N" standby cluster applications run on a system consisting of "N+1" nodes. One application runs on each operating node. Two or more applications stand by on one node.
Mutual standby	One cluster application runs, and the other cluster application stands by on each node. Generally, a cluster system consists of 2 nodes. This is referred to as two-node mutual standby.
Cascade	One operating node has two or more standby nodes. Even when one node stops, the cluster application assures its redundancy by using other node. This operation mode is effective for maintaining the availability of a cluster application during maintenance.
Priority transfer	Topology applied from N:1 standby Effective in maintaining availability during maintenance because the restored node becomes a standby node through the transfer of the standby node, even after the standby node stops and the cluster application can adopt a redundant configuration.
Scalable	Topology that is best suited for parallel job execution because one or more cluster applications operate together.
High-availability scalable	Topology in which standby operation is performed for each cluster application that constitutes a scalable configuration.

Cluster operation mode

Cluster operation mode	Description
Standby operation mode	General names for the four operation modes are 1:1 standby, N:1 standby, mutual standby, and cascade.
Scalable operation mode	One scalable cluster application consists of two or more operating nodes. If one node stops in the event of a failure, ongoing operations are continued on other operating nodes. Each operating node can be configured to be in the standby operation mode.

Public LAN takeover function

Public LAN takeover function	Description
IP address takeover	If a failover occurs during standby operation, one or more IP addresses are taken over to other node. In such a case, two or more network interface cards can be used.
MAC takeover	If a failover occurs during standby operation, the MAC address of the network interface card is taken over to other node. This function enables ongoing operations on a client that uses MAC addresses.
Node name takeover	If a failover occurs during standby operation, the node name is taken over to other node. This function enables access to the logical IP address corresponding to the node name.

Combination of the takeover network

Status of support		Type of takeover
SynfinityCluster	PRIMECLUSTER	IP	MAC	Node name
Y	Y	Y	N	N
Y	Y	Y	N	Y
Y	Y	Y	Y	Y
Y	N	N	Y	N
Y	Y	Y	Y	N

* MAC takeover must be used at the same time as IP address takeover with PRIMECLUSTER.

Physical IP address takeover and MAC address takeover deactivate a physical IP address at failover. To avoid disrupting ongoing operations of Web-Based Admin View or Shutdown Facility, set up physical IP address takeover or MAC address takeover for a network interface that is not being used by that Web-Based Admin View or Shutdown Facility.