| 摘要 |
PROBLEM TO BE SOLVED: To make it possible to access a target data base by performing full- text retrieval only in a specified range with all processor, by dividing data of a large-scale data base by repeating small-scale division, and performing full-text retrieval as to only necessary parts of many divisions. SOLUTION: An inputted inquiry described in SQL is found by taking a syntax analysis and suitably optimizing an access path to be adopted. After the access path is determined by the optimization, retrieval is carried out. when full-text retrieval is indicated to each node 1 of data access, the node 1 checks whether or not a key range condition is given and performs retrieval from only a storage area 30 stored with data meeting the condition in such a case. Therefore, only a part of key range divisions can be retrieved in spite of the full-text retrieval using all processors, and the retrieval is completed in a time which is a submulitple of the number of key divisions as compared with the full-text retrieval. |
