Active troubleshooting method for provisioning workflow for efficient cloud operation

A method for determining a category of intervention responses for provisioning a workflow. The method determines a provisioning characteristic of the provisioning step. The method performs outlier detection to identify and remove outliers from non-intervening data to produce a non-intervening normal data set. The method performs iterative grouping on the non-intervening normal data set to determine an effective variable in the provisioning feature. The method performs response mapping of the provisioning to classify the provisioning into response categories using results of iterative grouping including importance and presence of errors in non-intervening normal data and partially intervening data. 一种用于确定供应工作流的干预响应类别的方法。该方法确定供应步骤的供应特征。该方法执行离群值检测来从非干预数据中识别和移除离群值以产生非干预正常数据集。该方法对非干预正常数据集执行迭代分组以确定供应特征中的有效变量。该方法使用迭代分组的结果执行供应的响应映射，以将供应分类为响应类别，所述迭代分组的结果包括非干预正常数据和部分干预数据中的错误的重要性和错误的存在。