| 摘要 |
<p>A computer system for which multiple computers execute analysis processes in parallel with respect to data sets containing a plurality of data formed with keys and data values, with each computer holding, for each data set, partition information that manages partition position keys, which are keys indicating the partition positions for partition regions resulting when the aforementioned data sets are partitioned for each of prescribed key ranges, and with all of the partition position keys included in the partition information for each data set being the same. When a new data set is stored in the file system, the computer system determines, on the basis of the data size of each partition region subsequent to the storing of the new data set, whether a target region, which is a partition region having a data size greater than a prescribed threshold value, exists. When it is determined that a target region exists, the target region is partitioned into multiple new partition regions.</p> |
