Login

eMail
Password
 
 
Forgot Password!

COMMENTS

With the Performance Engineering Team being the last team to sign-off on a release prior to its move to production it is imperative that the Performance team be very organized with repeatable processes to produce timely results for the Go/No Go decision.

Maturity Model

In order to reduce risks and cycle times, the Performance Team needs to determine the maturity level of the project, department, or organization. The goal of the performance team is to quickly migrate up the maturity model for repeatability of improvement.

Performance Accountabilities

The Performance team will be judged by the eCommerce user experience related to throughput, response times and the Distribution Center capacity related to orders and units shipped
The Performance team is judged on the efficiency of process for issue resolution and application performance compliance during the release cycle
The Performance team is judged on the quality and accuracy of the trending report that is associated with the reliability of each release baseline
The Performance team is judged by the hardware / software tuning efforts and the timeliness of the What if scenarios needed to maintain peak event application performance and availability to protect brand reputation
The Performance team is judged by their ability to identify issues and facilitate issue resolution

Business and Technology Goals

Develop baseline timings throughout the supply chain using peak season volumes to determine throughputs and compare with business defined SLAs.The business used transactions per second to compare throughput volumes
Install monitoring throughout the supply chain to determine transactional throughputs, cpu and memory utilization, garbage collection results and network traffic volumes to leveraged for decision-making
Generate large volumes of test data at each interface point; message queues, soap, ejb, servlets, api's and databases on demand. Ensure referential integrity with the backend
Once we generate the test data, provide the ability to dump the data into performance, QA, development and other environments on demand
Have the necessary tools in place to answer the What If scenarios from the business and infrastructure teams
Provide a framework where both on-shore and off-shore can leverage the same testing data
Utilize Flashback technologies to quickly restore testing data
Develop standardized reports and store the data to use in summarized graphs

PERF STRATEGY

This document provides a detailed description of the plan that is utilized to execute Component, Load and Duration Performance Engineering for eCommerce Applications.

  Runbooks
  Data Generation Guide
  Sterling Commerce
  Training

webmaster@performanceCOE.com

PORTAL OVERVIEW - Metrics, Monitoring and Measuring

This portal provides a framework that will be utilized to execute Performance Engineering for eCommerce applications. The eCommerce applications are an integration of a number of systems, some previously associated, but others developed and fielded as stand-alone. The eCommerce applications integrate a number of multi-tier components executing many real-time and batch transactions, and as such are particularly susceptible to performance degradation. Moreover, because of the significant development effort ongoing, not only in the integration of components but within the components themselves, individual components cannot be assumed as optimized at the time of integration with consumer applications. The Performance Engineering strategy thus intends to reduce risk by employing an incremental, 'maturity model' approach to validating eCommerce application performance.

Performance Engineering



eCOMMERCE Supply Chain