Advances, Systems and Applications
Table 9 This table presents the related works of other types of failure prediction using other datasets. The algorithm in bold refers to the best model mentioned in the respective article. [Note: SML - Statistical Machine Learning Algorithms, TML - Traditional Machine Learning Algorithms, DL - Deep Learning Algorithms]
From: Cloud failure prediction based on traditional machine learning and deep learning
Article | Data source | Prediction scope | Feature studied | SML | TML | DL |
|---|---|---|---|---|---|---|
Guan et al. [37] | Private Data | System Failure | CPU usage, Memory Usage, Swap Space Utilization Page Faults, Interrupts, Network Activity, I/O and Data Transfer, Power Management | - | Bayesian Network, DT | - |
Adamu et al. [38] | NERSC | Component Failure | Failed Component, Failure Time | - | LR, SVM | - |
Pitakrat et al. [39] | Private Data | System Failure | Resource Usage, Failure Information, Failure Time | - | RF | - |
Zhang et al. [40] | Public Dataset | Switch Failure | Message Template Sequence, Frequency, Seasonality, Surge | HSMM | RF, SKSVM | - |
Lin et al. [41] | Private Cloud | Node Failure | Resource Usage, Group Policy, Domain Group, Rack Location | - | LR, SVM, RF | LSTM |
Han et al. [42] | Alibaba’s Cloud, Backblaze SMART | Disk Failure | SMART Log, SysLog, Trouble Ticket | - | LGBM | - |
Mohammed et al. [43] | NERSC | Component Failure and Service Failure | Multiple Sources of Failure | ARIMA | LDA, CART, RF, SVM, KNN | - |
Chen et al. [44] | Microsoft Cloud System | Outage Prediction | Storage Location, Physical Networking, Storage Streaming Component | - | SVM, PLR, Bayesian Network, XGBoost | - |
Li et al. [13] | System X | Node Failure | Resource Usage, Group Policy, Domain Group, Rack Location | - | RF | LSTM |
Rawat et al. [45] | Private Cloud | VM Failure | Number of VM Used | ARIMA | - | - |
Yu et al. [46] | Alibaba Cloud System | DRAM Failure | System Kernel Log, MCA Data Log | - | XGBoost | - |