| 摘要 |
PURPOSE: To provide a method and a system for deciding the set of tables capable of redistributing data among nodes inside a parallel data base. CONSTITUTION: The list of candidate tables is established from the tables 62, 64, 66 and 68 stored in a memory relating to the plural nodes 1-4 inside the parallel data base first and the list is sequenced corresponding to the respective operation loads of the candidate tables. Then, the respective candidate tables are tested and whether or not the candidate tables can be scanned within prescribed time in the respective plural nodes storing the candidate tables is judged. All the candidate tables scannable within the prescribed time are added to the set of the tables capable of redistributing the data.
|
