发明授权
US08108718B2 Checkpointing in massively parallel processing 有权
大规模并行处理中的检查点

Checkpointing in massively parallel processing
摘要:
One embodiment is a method that performs a local checkpoint at a processing node in a massively parallel processing (MPP) system that executes a workload with a plurality of processing nodes. The local checkpoint is stored in local memory of the processing node. While the workload continues to execute, a global checkpoint is performed from the local checkpoint stored in the local memory.
公开/授权文献
信息查询
0/0