To provide high availability of RMS resources, RMS takes advantage of the following redundancy scheme:
Multiple nodes, each configured to assume the resource load of any other RMS node
Duplication of stored data using mirror disks, hardware RAID, and remote mirroring
Multiple-path access to storage media
Multiple communications channels dedicated to internode communications
Multiple nodes
An RMS configuration is made up of multiple nodes, each containing identical operating systems and RMS software. The maximum number of nodes in an RMS configuration is theoretically unlimited.
See
Refer to "PRIMECLUSTER Software Release Guide" and "PRIMECLUSTER Installation Guide" for the appropriate number of nodes.
Shared storage
All the nodes in the RMS configuration must have access to whatever data is shared. Typically, all the nodes have the ability to access shared disks over a SAN. However, other access methods, such as Network Attached Storage are possible.
RMS network
RMS uses TCP/IP protocols for communication between the nodes in the configuration. RMS uses Cluster Interconnect Protocol (CIP) on the redundant cluster interconnect that configures PRIMECLUSTER so that RMS can monitor RMS on other nodes in the RMS configuration.
See
See "3.1.2 Interconnect protocol" for more information.
