Enhancing Resource Utilization Efficiency in Serverless Education: A Stateful Approach with Rofuse
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsIn Section 4.3.2, the authors could provide more details on the memory management strategies for the FUSE server's in-memory cache, such as the eviction policy and cache size tuning. Cross-ref this - https://doi.org/10.1109/EECSI59885.2023.10295630
In Section 4.3.3, the authors could consider dynamic block size adaptation based on workload characteristics to optimize performance further. Cross-ref this - https://doi.org/10.1007/s10723-023-09727-1
In Section 4.3.4, the authors could provide more implementation details on the data block splitting process, such as handling sparse files or optimizing the metadata storage format.
In Section 5.1, the authors could explore the impact of different workloads (e.g., read-intensive vs. write-intensive) on the performance of Rofuse compared to other solutions.
In Section 5.2, the authors could investigate the scalability of Rofuse under different cluster sizes and container densities, as these factors may affect the cache hit ratio and overall performance.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe authors discuss very interesting topics related to improving the quality of education in higher education.
In the Introduction, the authors well present the reasons why they dealt with this topic. They point out well what they propose in the article, which makes it easier for the reader to form an opinion about the content of the article.
The next chapter shows the background and related work and it shows well the scientific background for the results. The only thing I could suggest is to enlarge this part by adding more current literature.
The description of the back framework shows well the essence and assumptions of the problem. A good addition is a graphical presentation of the proposed framework.
The subsequent descriptions supported by drawings well illustrate the results shown.
The proposed solution looks very promising, and the presented descriptions and results look very interesting.
The evaluation of the solution to verify the postulated advantages also deserves positive attention.
The article ends with an appropriate summary in an appropriate and non-trivial way.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsTraditional container orchestration platforms have the shortcoming of resource wastage in educational scenarios, while stateless serverless services struggle to meet the need for container state persistence during the teaching process.
To address these issues, AUTHORS propose a stateful serverless mechanism based on Containerd and Kubernetes, focusing on optimizing the startup process for container groups.
THEY first implement a checkpoint/restore framework for container states, providing fundamental support for managing stateful containers.
Building on this foundation, THEY propose the concept of "container groups" to address the problem in educational practice scenarios with characteristics of a large number of similar containers on the same node. Then, THEYpropose the Rofuse optimization mechanism, which employs delayed loading and block-level deduplication techniques, enabling containers within the same group to reuse locally cached file system data at the block level, thus reducing container restart latency.
THEIR Experimental results demonstrate that their stateful serverless mechanism can run smoothly in typical educational practice scenarios, and Rofuse reduces container restart time by approximately 50% compared to existing solutions.
THEY conclude that this research provides valuable exploration for serverless practices in the education domain, contributing new perspectives and methods to improve resource utilization efficiency and flexibility in teaching environments.
The study is interesting and written with enthusiasm.
It contributes to the literature.
The content is fine but needs improvements in the structure as it does not follow the standard architecture of a manuscript (intro/methods/results/discussion/conclusion). Furthermore the discussion is lacking.
These are my comments:
1. In the introduction “The main contributions are summarized as follows: 1. We propose …..2. The concept ….. 3. We conduct ….” Something does non flow well. Please uniform.
2. Modify the structure of the first two sections. The content is fine, however the logical sequence does no run. They are both introductive sections, in the second section there is the background. I suggest to insert the purpose at the end of the introductive discourse.
3. A presentation of the sections/organization of the study after the purpose could be useful
4. Revise the sections 3-4-5 into methods/results. It is not clear to me
5. Figure must be described in detail (Figure 2 for example).
6 Insert the discussion with comparison to other studies, the discussion of further work and the limitations encountered.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThe substantial revisions you have made have significantly improved the quality of the manuscript.
Thank you for your diligent efforts in improving the manuscript and incorporating the original feedback provided.
Reviewer 3 Report
Comments and Suggestions for AuthorsAuthors improved the ms following my cooments.
There are no futher comments.
