发明名称 Distributed performance evaluation framework
摘要 A distributed performance evaluation framework provides functionality for evaluating the availability, reliability, and scalability of a network-based service. The framework includes a control and reporting tier and a load-generating tier. The control and reporting tier provides a control interface through which a request to evaluate the performance of a service may be received. In response to receiving such a request, the control and reporting tier creates a load-generating job for use by the load-generating tier. Load-generating instances in the load-generating tier are configured to perform the load-generating job by executing instances of a load-generating plug-in configured to generate requests to the service. The load-generating instances also periodically provide data regarding the status of each load-generating job to the control and reporting tier.
申请公布号 US9455887(B1) 申请公布日期 2016.09.27
申请号 US201313966867 申请日期 2013.08.14
申请人 Amazon Technologies, Inc. 发明人 Upadhyay Divyang Mahendrabhai;Abrahams David William
分类号 G06F15/173;H04L12/26 主分类号 G06F15/173
代理机构 Lee & Hayes, PLLC 代理人 Lee & Hayes, PLLC
主权项 1. A system for evaluating performance of a service, the system comprising: one or more computing systems configured to provide a control and reporting tier configured to receive a request by way of a load generating Web service application programming interface (API) to evaluate the performance of the service and, in response to receiving the request, to generate a load-generating job based upon the request and to provide the load-generating job to a load-generating tier; and one or more other computing systems configured to provide the load-generating tier, the load-generating tier comprising one or more load-generating instances configured to receive the load-generating job and to generate requests to the service based upon the load-generating job, wherein each of the load-generating instances is further configured to determine if at least a desired number of requests per time period are being submitted to the service and, if at least the desired number of requests per time period are not being submitted to the service, to enable other of the load-generating instances to also submit requests to the service.
地址 Seattle WA US