Research Article
Optimizing Hadoop Performance for Big Data Analytics in Smart Grid
Table 6
GEP recommended configuration parameter settings.
| Configuration parameters | Optimized values |
| Number of data samples in million | 17.28 | 34.56 | 51.84 | 69.12 | 86.40 | io.sort.factor | 10 | 10 | 13 | 19 | 12 | io.sort.mb | 38 | 13 | 58 | 62 | 55 | io.sort.spill.percent | 0.90 | 0.90 | 0.90 | 0.89 | 0.90 | mapred.reduce.tasks | 14 | 14 | 13 | 2 | 16 | mapreduce.tasktracker.map.tasks.maximum | 8 | 6 | 8 | 8 | 8 | mapreduce.tasktracker.reduce.tasks.maximum | 1 | 1 | 1 | 2 | 1 | mapred.child.java.opts | 169 | 135 | 121 | 124 | 121 | mapreduce.reduce.shuffle.input.buffer.percent | 0.73 | 0.65 | 0.65 | 0.65 | 0.75 | mapred.inmem.merge.threshold | 200 | 200 | 201 | 202 | 201 |
|
|