摘要 |
A parallel computer (1A) includes a plurality of calculation nodes (50) executing a calculation job distributively in parallel, and a management node (60) managing the plurality of calculation nodes. The calculation node (50) includes a calculation side retention unit (14) retaining job information corresponding to a plurality of generations acquired by the calculation node (50), and the management node (60) includes a management side retention unit (34) retaining job information corresponding to a plurality of generations collected from each calculation node (50). Therefore, by using the job information corresponding to the plurality of generations retained in the calculation side retention unit (14) of each calculation node (50), the management node (60) secures an accurate snapshot of the calculation job that is being executed in the parallel computer (1A). |