Distributed computing system and method

Disclosed are a distributed computing system and method. The system comprises: a plurality of clusters for conducting distributed computing; a global metadata server for saving and managing a multi-cluster data copy distribution view and a multi-cluster topology structure view; a cross-cluster data replication server for replicating designated data into a cluster designated by a replication instruction according to the replication instruction; and a data access allocation server for allocating location information about target data according to a cluster data usage request, the multi-cluster data copy distribution view and the multi-cluster topology structure view, and generating a replication instruction according to the location information about the target data, so that the cross-cluster data replication server replicates a target data copy corresponding to the target data into a cluster designated by the data usage request. The system can guarantee that a computing task acquires service data within a limited waiting time range in a multi-cluster environment with a limited network condition when service data copies are dynamically distributed in many locations.