Early Stopping Agreement Problem in the Distributed Cloud Computing Environment

View Statistics Email Alert RSS Feed

Information

Details

Project title

Code/計畫編號

NSC102-2221-E032-026

Translated Name/計畫中文名

分散式雲端網路環境下即時容錯協議問題之研究

Project Coordinator/計畫主持人

Chien-Fu Cheng

Funding Organization/主管機關

National Science and Technology Council

Department/Unit

Tamkang University Department of Computer Science and Information Engineering

Website

https://www.grb.gov.tw/search/planDetail?id=3108083

Year

2013

Start date/計畫起

01-08-2013

Expected Completion/計畫迄

31-07-2014

Bugetid/研究經費

531千元

ResearchField/研究領域

資訊工程--硬體工程

在前一年度的計畫中，我們探討了分散式雲端網路環境下的各種協議問題，藉此讓雲端網路中參與運作的各式處理器即使在發生損毀或被攻擊時仍能夠協同工作。在本計畫書中，我們將進一步地探討以下 3項議題於分散式雲端網路環境之中。議題 1：雲端運算環境下之 "即時" 容錯協議問題。議題 2：雲端運算環境下具 "雙重損毀模式" 之 " 即時" 容錯協議問題。議題 3：雲端運算環境下損毀處理器之 "隔離機制"。在議題 1中，我們將設計出適用於雲端網路環境之 Early Stopping機制，當收集到足夠的訊息時即停止訊息交換，以減少訊息交換所需之時間及空間複雜度，進一步地提升演算法之執行效率。在議題 2 中，我們將進一步地把損毀情形區分為硬性損毀(hardware-fault)以及軟性損毀(software-fault)，藉由分析不同程度之損毀情況，來增加可容忍之損毀處理器數量，以提升雲端系統之容錯能力，並且設計出能夠處理雲端網路環境中具雙重損毀模式下之即時容錯協議演算法。在議題 3中，我們將探討如何偵測/定位出雲端網路環境中的損毀處理器，藉此設計出適用於雲端網路環境之損毀處理器隔離機制。在時程方面，本計畫預計以兩年的時間來解決上述之問題。包含即時協議演算法之設計(降低協議達成所需之訊息交換次數，即降低達成協議所需之時間及空間複雜度)、雙重損毀模式下之即時容錯協議演算法之設計(增加可容忍的損毀處理器數量，即容錯能力之提升)以及損毀處理器之隔離機制(偵測/定位出損毀處理器，藉由隔離損毀處理器進一步地提升系統之可靠度)。 In previous project, we address the agreement-related problems in the distributed cloud environment to ensure robustness of the collaboration of all fault-free processors against the faulty influence from faulty processors. In this project, we attempt to visit the following three issues in the cloud environment: (1) the "early stopping" agreement problem in the distributed cloud environment; (2) the "early stopping" agreement problem with "dual failure mode" in the distributed cloud environment; (3) the "isolation mechanism" of faulty processor in the distributed cloud environment. This project is expected to take two years to complete the following three issues: In the first issue, an early stopping rule suited for the distributed cloud environment will be developed. With this rule, the system can terminate the message exchange operation if enough messages have been collected and enter the decision-making stage earlier. Through reduction of message exchange rounds, this rule can increase the efficiency of the protocol. In the second issue, we will divide faults into hardware-fault and software-fault and enhance the fault tolerance of the system through analysis of fault level. Moreover, we will also design an early stopping rule for dual failure mode in the distributed cloud environment. In the third issue, we will discuss the isolation mechanism of faulty processors in the distributed cloud environment and develop a mechanism that can help the system isolate faulty components.

Keyword(s)

雲端運算
即時容錯協議問題
硬性損毀
軟性損毀
隔離機制
cloud computing
early stopping agreement problem
hardware-fault
software-fault
isolation mechanism

DSpace CRIS

Early Stopping Agreement Problem in the Distributed Cloud Computing Environment

Details

Description