首页> 外文期刊>CONCURRENCY PRACTICE & EXPERIENCE >Performance anomaly detection using isolation-trees in heterogeneous workloads of web applications in computing clouds
【24h】

Performance anomaly detection using isolation-trees in heterogeneous workloads of web applications in computing clouds

机译:在计算云中Web应用程序的异构工作负载中使用隔离树进行性能异常检测

获取原文
获取原文并翻译 | 示例

摘要

Cloud computing is amodel for on-demand access to shared resources based on the pay-per-usepolicy. In order to efficientlymanage the resources, a continuous analysis of the operational stateof the system is required to be able to detect the performance degradations and malfunctionedresources as soon as possible. Every change in the workload, hardware condition, or softwarecode can change the state of the system fromnormal to abnormal, which causes the performanceand quality of service degradations. These changes or anomalies vary from a simple gradualincrease in the load to flash crowds, hardware faults, software bugs, etc. In this paper, wepropose Isolation-Forest based anomaly detection (IFAD) framework based on the unsupervisedIsolation technique for anomaly detection in a multi-attribute space of performance indicatorsfor web-based applications. Unsupervised nature of the algorithm and its fast execution makethis algorithm most suitable for the environments with dynamic nature where the patternsof data change frequently. The experiment results demonstrate that IFAD can achieve gooddetection accuracy especially in terms of precision formultiple types of the anomaly. Moreover,we show the importance of validating the accuracy of anomaly detection algorithms with regardto both Area Under the Curve (AUC) and Precision-Recall AUC (PRAUC) in an extensive set ofcomparisons including multiple unsupervised algorithms. The demonstration of the effectivenessof each algorithm shown by PRAUC results indicates the importance of PRAUC in selectingsuitable anomaly detection algorithm, which is largely ignored in the literature.
机译:云计算是一种基于按使用量付费策略按需访问共享资源的模型。为了有效地管理资源,需要对系统的运行状态进行连续分析,以便能够尽快检测到性能下降和故障资源。工作负载,硬件状况或软件代码的每一次更改都会将系统状态从正常更改为异常,从而导致性能和服务质量下降。这些变化或异常从负载的简单逐渐增加到闪存人群,硬件故障,软件错误等不等。在本文中,我们提出了一种基于无隔离森林的异常检测(IFAD)框架,该框架基于无监督隔离技术,可在多系统中进行异常检测-基于Web的应用程序的性能指标的属性空间。该算法的无监督性质及其快速执行使该算法最适合于动态性质的环境,其中数据模式经常变化。实验结果表明,IFAD可以实现良好的检测精度,尤其是对于多种异常类型的精度而言。此外,我们在包括多个无监督算法的广泛比较中显示了验证异常检测算法相对于曲线下面积(AUC)和精确召回AUC(PRAUC)的准确性的重要性。 PRAUC结果显示的每种算法的有效性证明了PRAUC在选择合适的异常检测算法中的重要性,这在文献中被很大程度上忽略了。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号