Complementary in Time and Space: Optimization on Cost and Performance with Multiple Resources Usage by Server Consolidation in Cloud Data Center
Round 1
Reviewer 1 Report
The authors have designed a method of reducing the operating cost of the cloud by increasing the performance of the multiple resources usage by server consolidation in the Cloud Data Centers. The authors have also adopted convolution autoencoder-based filter approach to preprocess the data and utilize the attention based Recurrent Neural Network model to resolve the Minimizing Computing Resources Cost problem and resource utilization problems. Though, the manuscript has adopted a novel approach to prove its place, still some basic observations need proper justifications.
I suggest the authors to work on the following issues:
1. The content of the abstract is too colloquial and it should be precise.
2. In the Introduction part, the main contributions are chaotic and it is recommended to include the highlights of the proposed work to emphasize the authors’ novel work.
3. The authors presented the related work focusing on “Energy Consumption and Cost Models” & “Server Consolidation Solutions”. But it is recommended to provide a proper explanation for opting these specific concepts.
4. In Section 3, Line no. 203 the notations are not rendered properly. Also, verify that all the equations (in all sections) have proper notation and explanation. (i.e. for eg: - In Eqn. 26, the authors failed to explain the notations)
5. It is recommended to enhance the quality of the figures: Figure 1, Figure 3 & 4.
6. Why the performance of the proposed model is compared with only one baseline method?
7. Does reducing the cost of SLAV can reduce the overall cost of the model? Justify your answer
8. The entire manuscript has to be proofread by a native English speaker. While reading the manuscript, scientific/technical touch is missing, it is colloquial
Comments for author File: 
Comments.pdf
Author Response
Response to Reviewer 1 Comments
Point 1: The content of the abstract is too colloquial and it should be precise.
Response 1: Thank you very much for your suggestion.
We have rewritten the abstract. Following is the new abstract:
The recent COVID-19 pandemic has accelerated the use of cloud computing. The surge in the number of users presents cloud service providers with severe challenges in managing computing resources. Guaranteeing the QoS of multiple users while reducing the operating cost of the cloud data center (CDC) is a major problem that needs to be solved urgently. To solve this problem, this paper establishes a cost model based on multiple computing resources in CDC, which comprehensively considers the hosts' energy cost, virtual machine (VM) migration cost, and SLAV penalty cost. To minimize this cost, we design the following solution. We employ a convolutional autoencoder-based filter to preprocess the VM historical workload and use an attention-based RNN method to predict the computing resource usage of the VMs in future periods. Based on the predicted results, we trigger VM migration before the host enters an overloaded state to reduce the occurrence of SLAV. A heuristic algorithm based on the complementary use of multiple resources in space and time is proposed to solve the placement problem. Simulations driven by the VM real workload dataset validate the effectiveness of our proposed method. Compared with the existing methods, our proposed method reduces the energy consumption of the hosts and SLAV and reduces the total cost by 26.1% ~ 39.3%.
Point 2: In the Introduction part, the main contributions are chaotic and it is recommended to include the highlights of the proposed work to emphasize the authors’ novel work.
Response 2: Thank you very much for your suggestion.
We have described our contributions on Page 2~3 in the revised manuscript.
Our contributions are as follows:
- A) We formally identify the multi-resource-based cost model for server consolidation, which involves host energy consumption, VM migration, and SLAV. Based on the cost model, the optimization problem is given.
- B) A convolutional auto-encoder-based filter is leveraged to denoise the VM workload trace. Then we propose an attention-based RNN method to predict the future workloads of the VMs. Based on the prediction results, a host workload detection policy is proposed.
- C) To minimize the total cost of server consolidation, we propose a VM selection policy and a VM placement algorithm which consider the multi-resource demands of VMs in the present and future.
- D) We conduct the simulations to evaluate the performance of our proposed solution ARP-TSCP. The simulations’ results indicate that ARP-TSCP can reduce host energy consumption by 18.5% ~30.3%, SLAV cost by 38%~52%, and total cost by 26.1%~39.3% as compared to the baseline methods.
Point 3: The authors presented the related work focusing on “Energy Consumption and Cost Models” & “Server Consolidation Solutions”. But it is recommended to provide a proper explanation for opting these specific concepts.
Response 3: Thank you very much for your suggestion.
In section 2 (Page 2), we inserted the following explanation before subsection 2.1 and 2.2 of the revised manuscript:
“In this section, we first survey the energy consumption models and cost models in cloud server consolidation, then we review the server consolidation solutions.”
Then we revised the title of subsection 2.1 to ‘Sever Consolidation Cost Models’, and we explain that ‘The cost of server consolidation in the cloud is mainly related to the host energy consumption, VM migration, and SLAV.’ And the next three paragraphs are related to host energy consumption, VM migration, and SLAV, respectively.
To view more details, please check the highlights (blue color) in section 2.
Point 4: In Section 3, Line no. 203 the notations are not rendered properly. Also, verify that all the equations (in all sections) have proper notation and explanation. (i.e. for eg: - In Eqn. 26, the authors failed to explain the notations)
Response 4: Thank you very much for your suggestion. We have modified these mistakes.
Please check Line No. 203-204, Page 5，Line No. 210, Page 5, and Line No.307, Page 9 in the revised manuscript.
Point 5: It is recommended to enhance the quality of the figures: Figure 1, Figure 3 & 4.
Response 5: Thank you very much for your suggestion. We have updated these figures’ positions and sizes in the revised manuscript. Please check Page 11, 13, and 14. Due to the sizes of the figures, the PDF reader application needs time to load. And these figures can be zoomed in to see more details.
Point 6: Why the performance of the proposed model is compared with only one baseline method?
Response 6: Thank you very much for your suggestion. In the evaluation of the revised manuscript, four overloading detection algorithms (MAD [6], IQR [6], and LR [6]), three VM selection algorithms (MMT [6,30,39], MC [6,55], and RS [6]) and one VM placement algorithm (PABFD [6]) are combined as 9 methods to compare with our proposed solution. Hence, we rewrote subsection 5.2. Please check the revised manuscript for more details.
Point 7: Does reducing the cost of SLAV can reduce the overall cost of the model? Justify your answer
Response 7: Thank you very much for your suggestion.
In server consolidation, a tradeoff between energy consumption and QoS needs to be achieved. To reduce energy overhead means that VMs must be concentrated on as few hosts as possible. The problem with this is that the VMs would over-compete the hosts' resources, which would degrade performance and have a negative impact on QoS. Therefore, the essential purpose of reducing SLAV is to ensure performance and QoS.
The proposed ARP-TSCP method can not only make the VMs highly concentrated on the hosts, but also be aware of the occurrences of excessive resource contention (SLAV) in advance, and take countermeasures to avoid the actual occurrences of SLAV as much as possible. This means that our model and method not only reduce the energy cost but also reduce the occurrences of SLAV. Therefore, for our method, reducing the cost of SLAV does reduce the overall cost.
Point 8: The entire manuscript has to be proofread by a native English speaker. While reading the manuscript, scientific/technical touch is missing, it is colloquial.
Response 8: Thank you very much for your suggestion. We have improved the writing. Please check the highlights in the revised manuscript.
Author Response File: Author Response.pdf
Reviewer 2 Report
This paper proposes a cost model based on multiple computing resources in CDC, which 5 comprehensively considers the hosts energy cost, virtual machine (VM) migration cost, and SLAV 6 penalty cost.
I would suggest that in the related work part, or in the discussion part, a comparison between the proposed solution and the solutions from the literature should be included.
Otherwise, I have no observations to make regarding the way of presentation.
Author Response
Response to Reviewer 2 Comments
Point 1: I would suggest that in the related work part, or in the discussion part, a comparison between the proposed solution and the solutions from the literature should be included.
Otherwise, I have no observations to make regarding the way of presentation.
Response 1: Thank you very much for your suggestion. We have rewritten the abstract and the performance evaluation section, and we also reorganized the related work section. The comparisons between our proposed methods and multiple solutions (as the baseline methods) from the literature. Please check the highlights in the manuscript.
Author Response File: Author Response.pdf
