PROBLEM TO BE SOLVED: To perform data rearrangement faster by excluding data movement between existent processors.SOLUTION: In a database server comprising a plurality of processors and a record storage part stored with different records for each processor on the basis of hash values, the processors each generate second processor information by replacing processor identification information, as part of first processor information in which each of all hash values and processor identification information on a processor referring to a record storage part stored with records of the hash values are made to correspond to each other, with processor identification information on the processor to be added, and rearrange the records in the record storage part that the added processor refers to from a record storage part that the local processor refers to, on the basis of the first processor information and second processor information.SELECTED DRAWING: Figure 1
展开▼