Extended remote copy works together with cached storage controls to efficiently manage existing system resources for disaster recovery. The system data mover issues I/Os to drain record updates from the cache that are designated to be copied by XRC. Each logical XRC session may have one or more storage control sessions for each primary storage control. A logical XRC session manages all volumes that are associated with those storage control readers.
High performance can be achieved by ensuring that the storage subsystems have adequate resources, including cache, NVS, and path connections to handle the peak workload that is associated with remote copy sessions (and coexisting concurrent copy operations). The SDM software requires adequate processor MIPs, multiple processors, and real storage to accomplish its function in an efficient and timely fashion. The SDM, without adequate resources, may not be able to drain the storage control cache rapidly enough under peak workload conditions. Based on the workload, if this continues for an extended period of time, the cache could become overcommitted, and ultimately this could affect the performance of your primary systems.
Given adequate resources, the system data mover functions smoothly even as the workload stress on it grows. Constrained resources will become apparent as they create bottlenecks that impede the performance of the system data mover.