Database Hash Join Bucket Division for Reduced Computation Time
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional Hash Join methods require excessive computation time due to the need to compare all records in directory entries during the join operation, leading to inefficient performance.
Innovation Solution
The proposed join method divides buckets when no free space is available, storing records based on differing upper bits, and performs join operations by reading buckets in order, utilizing sorting or hashing only on tables without primary keys, and merging records using binary search or Merge Sort techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all records in directory entries are compared for the Join operation, then the Join operation can be performed, but the computation time becomes excessively long
Solution Approach 1:
The patent divides the hash table into multiple buckets, each containing only records with the same hash value. This segmentation allows the join operation to compare records only within the same bucket rather than across all directory entries, significantly reducing computation time while maintaining join correctness.
Solution Approach 2:
The patent performs preliminary hashing and bucket assignment before the join operation. Records are pre-grouped into buckets based on their hash values, so that during the join operation, only records within the same bucket need to be compared. This preliminary organization eliminates the need to compare all records against all other records.
2Productivity
If sorting or hashing is performed on tables with primary keys, then the join operation can be optimized, but the overhead of hashing increases computation time
Solution Approach 1:
The patent applies hashing and sorting only to the necessary portions of data - specifically, only to tables without primary keys or to the foreign key columns. Tables with primary keys are processed differently, avoiding unnecessary hashing overhead. This local application of hashing optimizes the join operation by focusing computational resources only where needed.
Data Source
AI summary
A Join method in a database management system according to an embodiment of the invention comprises: in a memory of a storage medium, which processes data stored within a computer memory, when a new record is added to a bucket of a first table including only a basic key of another table, dividing the bucket in a case where there is no free space in the bucket; and on the basis of a fact that at least one bit of each of the new record and a first record having existed in the bucket before the dividing thereof is different from an upper bit thereof, storing one of the first record and the new record in the divided bucket.