Abstract: Aspects of the invention include a continuous availability for workload and site switches with no data loss at unlimited distances in an active-active sites configuration. A non-limiting example computer-implemented method includes synchronously replicating commit records of changed workload data, by a processor, from an active site to a bunker site upon a workload data commit. The method asynchronously replicates committed transactions on changed workload data, by the processor, from the bunker site to a recovery site.

Abstract: A method for resynchronizing at least one batch job is provided. The method may include detecting a type of region switch request. The method may further include stopping execution workloads on a primary system based on the switch request. The method may further include suspending software and hardware data replication from the primary system to the secondary system. The method may further utilizing proxy replication engine to determine a point-in-time (PIT) at which the execution is stopped and the suspension. The method may also include switching the replication of the software and hardware data to occur the secondary system to the primary system. The method may further include synchronizing the software and hardware data up to the determined point-in-time (PIT). The method may also include activating the execution of the plurality of workloads on the secondary system based on the determined point-in-time (PIT), switching, and synchronizing.

Abstract: A data processing system and method of operation are disclosed. The data processing system includes a first site for processing data; a second site for processing data; and a transaction replay program configured to send a copy of data from the first site to the second site. A processor defines a maximum replication latency threshold and a reset threshold for the system, estimates a replication latency of a workload the system, switches the system to a first replication latency state when the estimated replication latency rises above the maximum replication latency and switches the system out of the first replication latency state to a second replication latency state when the estimated replication latency falls below the reset threshold.

Abstract: Embodiments of the disclosure are directed to methods, systems and computer program products for performing a planned workload switch. A method includes receiving a request to switch a site of an active workload and stopping one or more long running processes from submitting a new request to the active workload. The method also includes preventing a new network connection from accessing the active workload and processing one or more transactions in a queue of the active workload for a time period. Based on a determination that the queue of the active workload is not empty after the time period, the method includes aborting all remaining transactions in the queue of the active workload. The method further includes replicating all remaining committed units of work to a standby workload associated with the active workload.

Abstract: Embodiments of the disclosure are directed to methods, systems and computer program products for performing a planned workload switch. A method includes receiving a request to switch a site of an active workload and stopping one or more long running processes from submitting a new request to the active workload. The method also includes preventing a new network connection from accessing the active workload and processing one or more transactions in a queue of the active workload for a time period. Based on a determination that the queue of the active workload is not empty after the time period, the method includes aborting all remaining transactions in the queue of the active workload. The method further includes replicating all remaining committed units of work to a standby workload associated with the active workload.

Abstract: A data processing system and method of operation are disclosed. The data processing system includes a first site for processing data; a second site for processing data; and a transaction replay program configured to send a copy of data from the first site to the second site. A processor defines a maximum replication latency threshold and a reset threshold for the system, estimates a replication latency of a workload the system, switches the system to a first replication latency state when the estimated replication latency rises above the maximum replication latency and switches the system out of the first replication latency state to a second replication latency state when the estimated replication latency falls below the reset threshold.