Research Article
Query Execution Optimization in Spark SQL
Table 2
Parameters in the cost model.
| Parameter | Meaning |
| X | Size of read/written data | C0 | Seeking time and rotation delay time | C1 | Time required to transmit 1 MB data | Α | Proportion of non-local data to total data | T | Number of I/O occurrences | |Din| | Size of stage input data | |Dout| | Size of stage output data tr time to read 1 MB data locally | | Time to write 1 MB data locally | tb | Time to transfer 1 MB data over network | B | Buffer size of spark task m task number in stage |
|
|