CAFE: Adaptive VDI Workload Prediction with Multi-Grained Features
Authors: Yao Zhang, Wen-Ping Fan, Xuan Wu, Hua Chen, Bin-Yang Li, Min-Ling Zhang5821-5828
AAAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on real VDI customers data sets clearly validate the effectiveness of multi-grained features for VDI workload prediction. Furthermore, practical insights identified in our VDI data analytics are also discussed. |
| Researcher Affiliation | Collaboration | 1School of Computer Science and Engineering, Southeast University, Nanjing 210096, China 2VMware Information Technology (China) Ltd. 3Key Laboratory of Computer Network and Information Integration (Southeast University), Ministry of Education, China |
| Pseudocode | No | The paper describes the model mathematically and textually but does not include any pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide any statement or link indicating that the source code for the CAFE model is publicly available. |
| Open Datasets | No | The experimental data sets are collected from four pools of four different VDI customers over a period of 11 weeks from June 1st to August 17th, 2018. These are described as "real VDI customers data sets" and no specific link, DOI, or repository for public access is provided. |
| Dataset Splits | No | For each pool, the data is divided into two divisions. Division D1 uses data from June 1st to July 10th, 2018 as the training set and data from July 11th to July 17th, 2018 as testing set. Division D2 uses data from July 1st to August 10th, 2018 as the training set and data from August 11th to August 17th, 2018 as testing set. A separate validation set is not explicitly mentioned. |
| Hardware Specification | No | The paper mentions VDI infrastructure but does not provide any specific details about the hardware (e.g., CPU, GPU models, memory) used to run the experiments or train the models. |
| Software Dependencies | No | The paper refers to methods like GBDT, Prophet, and Holt-Winters, but does not list specific software libraries or their version numbers (e.g., scikit-learn version, Python version, specific deep learning frameworks). |
| Experiment Setup | Yes | For CAFE, the maximum action scope n is 1440 (24 hours). Furthermore, the granularity vector k is (1440, 720, 240, 180, 10, 2, 1) when t = 30 minutes and is (1440, 720, 240, 180, 20, 4, 1) when t = 60 minutes. Correspondingly, the action scope vector γ is (1440, 720, 240, 180, 60, 30, 1) when t = 30 minutes and is (1440, 720, 240, 180, 120, 60, 1) when t = 30 minutes. For Holt-Winters, we specify the seasonal parameter as 168 (24*7 hours) due to the weekly seasonal pattern in the data. |