Advances, Systems and Applications
Table 8 Deadline-aware schedulers and their properties
From: MapReduce scheduling algorithms in Hadoop: a systematic study
Reference | Year | Key Ideas | Advantages | Disadvantages | Comparison Algorithms | Evaluation Techniques |
|---|---|---|---|---|---|---|
Ghazali et al. [84] | 2022 | CLQLMRS: improving cache locality in MapReduce job scheduling using Q‑learning | Reducing execution time. Improving the data locality. Improving Hadoop performance | Need to train the scheduling policy, which may be challenging if environmental changes occur rapidly and retraining becomes difficult | FIFO, adaptive cache local scheduling (ACL) | Implementation: Cluster with 8 nodes |
Naik et al. [85] | 2017 | A learning-based MapReduce scheduler in heterogeneous environments | Improving the data locality. Improving the performance of MapReduce | Not considering job deadline and data locality | Not considering | Not considering |
Naik et al. [86] | 2015 | Performance Improvement of MapReduce framework in heterogeneous context using reinforcement learning | Minimizing job completion time Does not require prior knowledge of environmental characteristics | Not considering job deadline and data locality | Not considering | Not considering |