The maintenance of the PRIMECLUSTER system is divided as described below, depending on whether maintenance is performed while the job is stopped:
Maintenance that is performed while the entire cluster system is stopped.
Maintenance that is performed while the job continues to operate without state transition of the cluster application (failover or degeneration).
Hot maintenance is executed mainly after disk replacement of the disk array unit.
Also, the PCI Hot Plug function of SPARC Enterprise M series can be used for hot maintenance.
Maintenance that is performed while the maintenance target node is detached from the cluster by state transition of the cluster application (failover or degeneration), while the job is allowed to continue operating.
Of these, the type to be performed depends on the location and contents of the failure. Determine the maintenance that is to be performed after consulting with field engineers.
