Advances, Systems and Applications
Table 2 Summary of the related work
From: Analysis and prediction of virtual machine boot time on virtualized computing environments
Attributes | VM Boot Time Prediction Models | |
|---|---|---|
Rule-based | ML-based | |
Prediction Methods | Nguyen et al. [24] | Regression Tree (Govindaraju et al. [1]) |
Overview | Use the expertise of researchers to create rules based on several features, including CPU time and I/O time to build a prediction model | Use several features, including VM image size, memory utilization, and network utilization to build prediction models |
Advantages | 1) easy to interpret | 1) provide higher accuracy |
2) fast processing time | 2) provide a better understanding of data and features | |
Limitations | 1) does not consider competition between hosts | 1) have not been applied for VM boot time prediction |
2) does not consider the number of CPU cores | 2) only applied for a small-scale cluster of four hosts | |
| Â | 3) does not provide feature analysis | |