Scholarlink

<p>Cloud Computing is an easy access computation model, in which data is stored remotely by the owner of the data to enjoy the huge aspect of applications on demand. Cloud system experiences the huge number of failures due to large-scale diverseness and distributed nature. Various fault-tolerance techniques (FTT) are used to handle failures. Occasionally for tolerating failures, more intensive FTT i.e., CheckPointing fault tolerance technique is terrific, this technique handles the failure efficiently but gracefully causes an increase in the overhead. Moreover, less intensive FTT i.e., retrying is used for tolerating failure, but this technique cannot handle failure gracefully. Furthermore, to achieve high throughput, availability and efficiency in task execution we launch replica on certain resources. Replication technique generates overhead for keeping multiple replicas. Currently, there is no prediction mechanism on the basis of which suitable fault tolerance techniques can be applied. In this paper, failure prediction-based fault tolerance techniques (FPBFTT) is presented that first predict the resource failure, and on the basis of failure percentage, suitable FTT were applied. CheckPointingFTT will be applied to the resources when the failure percentage became greater than the upper threshold. RetryingFTT will be used when the failure percentage became less than the lower threshold. Furthermore, the replication technique has been used when the failure rate was in between the upper and lower threshold. To evaluate the performance of the proposed technique following parameters (throughput, response time, makespan, cost, SLA violation, budget, and deadline) were used. Simulation results acquired through experiments, concluded that our proposed technique yield better results in all conditions. Results revealed that in the case of 25 jobs, the response time of FPBFTT was 9.24% better than RBFTT, 15.1% better than CPLCA and 22.5% better than job retry technique. In the case of 50 jobs, the throughput of FPBFTT was 17.4% better than RBFTT, 37.8% better than CPLCA and 64.2% better than job retry technique. In the case of 100 jobs, the makespan of FPBFTT was 16.08% better than RBFTT, 26.8% better than CPLCA and 38.7% better than job retry technique.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Failure prediction-based fault-tolerant resource management techniques in cloud computing

  • 期刊论文

  • 发表日期：

    2025年12月17日

查看全文
  • Asmara Jadoon,
  • Babar Nazir,
  • Zulfiqar Ahmad

摘要

Cloud Computing is an easy access computation model, in which data is stored remotely by the owner of the data to enjoy the huge aspect of applications on demand. Cloud system experiences the huge number of failures due to large-scale diverseness and distributed nature. Various fault-tolerance techniques (FTT) are used to handle failures. Occasionally for tolerating failures, more intensive FTT i.e., CheckPointing fault tolerance technique is terrific, this technique handles the failure efficiently but gracefully causes an increase in the overhead. Moreover, less intensive FTT i.e., retrying is used for tolerating failure, but this technique cannot handle failure gracefully. Furthermore, to achieve high throughput, availability and efficiency in task execution we launch replica on certain resources. Replication technique generates overhead for keeping multiple replicas. Currently, there is no prediction mechanism on the basis of which suitable fault tolerance techniques can be applied. In this paper, failure prediction-based fault tolerance techniques (FPBFTT) is presented that first predict the resource failure, and on the basis of failure percentage, suitable FTT were applied. CheckPointingFTT will be applied to the resources when the failure percentage became greater than the upper threshold. RetryingFTT will be used when the failure percentage became less than the lower threshold. Furthermore, the replication technique has been used when the failure rate was in between the upper and lower threshold. To evaluate the performance of the proposed technique following parameters (throughput, response time, makespan, cost, SLA violation, budget, and deadline) were used. Simulation results acquired through experiments, concluded that our proposed technique yield better results in all conditions. Results revealed that in the case of 25 jobs, the response time of FPBFTT was 9.24% better than RBFTT, 15.1% better than CPLCA and 22.5% better than job retry technique. In the case of 50 jobs, the throughput of FPBFTT was 17.4% better than RBFTT, 37.8% better than CPLCA and 64.2% better than job retry technique. In the case of 100 jobs, the makespan of FPBFTT was 16.08% better than RBFTT, 26.8% better than CPLCA and 38.7% better than job retry technique.