研究生: |
蔡佾勳 Yi-Hsun Tsai |
---|---|
論文名稱: |
研製適用於容器化Android叢集伺服器之資源控管機制 Design and Implementation of Resource Management Framework on Container-based Android Cluster Server |
指導教授: |
呂政修
Jenq-Shiou Leu |
口試委員: |
陳維美
Wei-Mei Chen 吳晉賢 Chin-Hsien Wu 袁錦鋒 Kam-Fung Yuen 卓傳育 Chuan-Yu Cho 呂政修 Jenq-Shiou Leu |
學位類別: |
碩士 Master |
系所名稱: |
電資學院 - 電子工程系 Department of Electronic and Computer Engineering |
論文出版年: | 2018 |
畢業學年度: | 106 |
語文別: | 中文 |
論文頁數: | 59 |
中文關鍵詞: | 容器化Android 、虛擬手機架構 、叢集伺服器 、動態資源分配 、Linux容器 、cgroup |
外文關鍵詞: | Container-based Android, Virtual Mobile Infrastructure, cluster server, dynamic resource allocation, Linux container, cgroup |
相關次數: | 點閱:296 下載:3 |
分享至: |
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報 |
根據報告指出虛擬行動架構可有效解決自攜設備(Bring Your Own Device)帶來的企業資料安全與隱私權問題,企業可以透過建立一個大型的應用程式串流雲端架構來實現。然而在此架構中沒有一套可以公平分配資源給每個平台使用者的解決方案。當單一使用者使用過多系統資源時,在同一台主機上其餘的使用者會受到影響,導致使用者體驗下降。而在過往有關容器化Android的研究中,未針對此種情況進行設計與處理,因此在本論文中將提出一套基於容器化Android叢集伺服器之資源控管機制,本機制將使用物件導向程式技巧建立叢集伺服器管理系統進行叢集伺服器底層操作的實現與資訊的儲存,使用Linux內核cgroup功能建立節點資源動態分配機制,並實作一演算法可根據使用者要求進行主機分配。此外,將透過原生Android的CPU資源控管機制,使該機制可以相容於容器化Android,進而可以限制單一容器的CPU資源。為了達到更好的使用者體驗,必須確保使用者可以流暢地操作應用程式,意即應用程式可以在每秒60禎(60 frames per second)且沒有任何延遲與失禎(dropped frame)的情況下執行。因此本系統將使用FPS作為使用者體驗的量測指標。經由實驗量測結果驗證本系統可以透過改善容器化Android資源管理機制,有效地提升使用者體驗。
VMI (Virtual Mobile Infrastructure) has already been adopted as a better solution for the BYOD (Bring Your Own Device) problem. This service requires a scalable application streaming cloud infrastructure. However, there is no effective solution that can allocate system resources for each platform user fairly. When one user runs out of system resources, others in the same host computer may get affected, and their user experience may degrade. Nevertheless, all researches related to container-based Android in the past cannot resolve this issue effectively, so we design and implement a resource management framework to improve the user experience in the container-based Android. This framework includes three parts – (1) using a cluster manager implemented by object-oriented programming to communicate with cluster server and store the status about the server, (2) using cgroup from Linux kernel to implement resource control server that can dynamically allocate system resources, and (3) carrying out an algorithm that can allocate host computer by user requirement. In addition, we modify the native Android’s CPU resource management system to become container-aware in order to limit CPU resources for each container. For the better user experience, we should make sure that user interactions with your app are buttery smooth, running at a consistent 60 FPS (frames per second), without any dropped or delayed frames. Therefore, we evaluate the system availability of the proposed system by FPS as the user experience index. The experimental results show that the system can effectively improve the user experience by improving the containerized Android resource management mechanism.
[1] Lazar, M. (2017). BYOD Statistics Provide Snapshot of Future. Retrieved 7/18, 2018, from https://www.insight.com/en_US/learn/content/2017/01182017-byod-statistics-provide-snapshot-of-future.html
[2] Virtual Mobile Infrastructure: A Better Solution for the BYOD Problem. (2015). Retrieved 07/20, 2018, from https://www.trendmicro.com/vinfo/us/security/news/mobile-safety/virtual-mobile-infrastructure-a-better-solution-for-byod
[3] Heiser, G. (2008). The role of virtualization in embedded systems. Paper presented at the Proceedings of the 1st workshop on Isolation and integration in embedded systems.
[4] Morabito, R., Kjällman, J., and Komu, M. (2015). Hypervisors vs. lightweight virtualization: a performance comparison. Paper presented at the Cloud Engineering (IC2E), 2015 IEEE International Conference on.
[5] Soltesz, S., Pötzl, H., Fiuczynski, M. E., Bavier, A., and Peterson, L. (2007). Container-based operating system virtualization: a scalable, high-performance alternative to hypervisors. Paper presented at the ACM SIGOPS Operating Systems Review.
[6] Helsley, M. (2009). LXC: Linux container tools. IBM devloperWorks Technical Library, 11.
[7] Turner, P., Rao, B. B., and Rao, N. (2010). CPU bandwidth control for CFS. Paper presented at the Linux Symposium.
[8] Andrus, J., Dall, C., Hof, A. V. t., Laadan, O., and Nieh, J. (2011). Cells: a virtual mobile smartphone architecture. Paper presented at the Proceedings of the Twenty-Third ACM Symposium on Operating Systems Principles.
[9] Xu, L., Li, G., Li, C., Sun, W., Chen, W., and Wang, Z. (2015, March 30 2015-April 3 2015). Condroid: A Container-Based Virtualization Solution Adapted for Android Devices. Paper presented at the 2015 3rd IEEE International Conference on Mobile Cloud Computing, Services, and Engineering.
[10] 卓傳育, 林浩澄, and 洪茂榮. (2016). Smartphone Virtualization for Bring Your Own Device (BYOD) Security. [用於自攜裝置(byod)安全之智慧手機虛擬化技術]. 電腦與通訊(165), 57-67.
[11] Meera, A., and Swamynathan, S. (2013). Agent based resource monitoring system in IaaS cloud environment. Procedia Technology, 10, 200-207.
[12] Monsalve, J., Landwehr, A., and Taufer, M. (2015). Dynamic CPU resource allocation in containerized cloud environments. Paper presented at the Cluster Computing (CLUSTER), 2015 IEEE International Conference on.
[13] Keller, G., and Lutfiyya, H. (2010). Dynamic resource management in virtualized environments through virtual server relocation. International Journal on Advances in Software Volume 3, Number 3 & 4, 2010.
[14] Roh, H.-s., Lee, H.-w., and Lee, S.-h. (2014). A study on mobile virtualization. Paper presented at the Advanced Communication Technology (ICACT), 2014 16th International Conference on.
[15] Shah, M. A., Kamran, M., Khan, H., and Javaid, Q. (2016). Vdroid: A lightweight virtualization architecture for smartphones. Paper presented at the Future Technologies Conference (FTC).