Information about a volume in the DFSMSrmm control data set is stored in several records. A change in a volume's status can require updating several records. Since an update of multiple records takes a certain amount of time, it is possible that external events, such as a power outage, could interrupt the updating process. An interruption can leave some records that reflect the volume's old status and some its new and the control data set is then considered corrupt. It might be impossible to subsequently change the volume's status. This situation is referred to as a multi-record update failure.

DFSMSrmm detects a multi-record update failure at DFSMSrmm address-space start-up time when the control data set is not shared between systems. When a control data set is shared between multiple systems, the multi-record update failure can be detected at any time. A hardware failure on any of the systems could be detected by another system at the next I/O to the control data set. In this case, a SYSTEM RESET of the failing system releases the control data set reserve, and the other systems sharing that control data set, find they had experienced a multi-record update failure.