RMS is a software monitor designed to provide system-level High Availability (HA) to applications. A minimum of two nodes with shared access to common storage are administered from a console or cluster operation management PC. As the PRIMECLUSTER HA monitor, RMS uses detectors to monitor the state of components and resources for applications. If a resource or application fails, the event is notified by the detector and reported to RMS. RMS then takes the appropriate action as described in the sections that follow.
The figure below shows the components of a basic RMS cluster on Solaris.
Figure 4.1 RMS cluster on Solaris
For RMS, high availability means maximum availability of applications rather than uninterrupted availability of individual nodes.
RMS accomplishes high availability in two ways: redundancy and switchover.
