Research Article

Query Execution Optimization in Spark SQL

Table 3

Comparison of cache size with and without intermediate data correlation merging algorithm.

Id1234567891011

Input data (MB)256896384102417921664640384256640256
Spark buffer (MB)227792.2340908.815901476568340.8227.2568227.2
SSO buffer (MB)178625.8268715.212511162447268.2178.8447178.8