The Role of the Performance Engineer - Case Study Example

Engineering performance Designing and development of the software of telecommunications products calls for continuous devotion to issues of capacity as well as performance. Both are essential necessities for the products produced by a particular company. The STC is the Software Technology Center in Bell Labs Advanced Technologies has come up with the definition of performance engineering, terming it as the best professional activity that help the project meet the capacity and performance prescription the product. The research, describe the practices, to be used by various companies in previous years and a good number of projects. The study will also contain a detailed prescription of the role of the performance engineer at any given time. Performance is a persistent quality of software systems; everything has an impact on it, ranging from software itself to the rest of underlying layers, such as hardware, system, communications network and middleware. (Woodward M) Software Performance Engineering incorporates efforts to explain and better the performance, with two distinct dimensions: a late cycle measurement approach and an early-cycle predictive model-based analysis approach. The present progress and future trends in the two approaches that explain the tendency for them to unite. The natural weakening of a radio signal increases as the distance increases from the transmitter. It allows a single carrier frequency to be used to convey to the two or more conversations at age, provided that the connected radio channels are geographically adequately remote from each other. Cellular Radio Systems have been improved to enjoy the merits of radio signal attenuation in a highly organized manner. In order to provide the services of mobile radiotelephone nationally, for all who are willing to pay for it, the issues in engineering a cellular radio system to attain the overall set performance goals are achieved. A primary system mechanism that examines, and various improvement changes in the system are put into consideration and directed to give potentially positive improvements in the traffic carrying ability of the system Reuse of assets presents a greater opportunity for improving performance of various manufacturing companies, where the matter approaches on an ad hoc basis. The challenge is to gauge and better performance, paying much attention to critical assets. Adequate engineering reuse draws on existing assets such as process, modules, designs, suppliers and teams at a minimum cost than starting again from scratch. Successfully implemented, the technique reduces the cost of innovation, technology and engineering management processes in the company. Excluding the reclamation and recycling of waste products and buying in of assets from a different firm. That is; it reuses product platforms to gain reduced time to market for the current products, and increase the coverage of the market. Reusing designs to cut down development time cost and risk standardizing parts to attain merits of high volumes. The study presents an outline for recycled drawn from industry and literature research. The framework will entail of two major parts. Firstly, a characterization mechanism for assets that are reusable based on technological definition. It addresses the mechanism to reuse, and, secondly, the process to be used to model. The data flow diagram will show the process, agents, and data flows that boost t and facilitate engineering reuse. The practical reuse outline of the methodology will also be given, based on the framework provided. To engineer for performance, one need to implant the culture of performance in development life cycle, and need a process follow without much interference. When one has a process to follow, and also knows the point to start and how to go on with the operation, and finally one needs to know when to finish. Performance modeling helps an individual to apply engineering knowledge to the production process. The basic approach is to set goals and to measure the progress as per the set objectives. Performance modeling assists one to set goals for the application scenarios. Measuring will go on in the entire life cycle and helps to determine whether one is moving towards the performance goals. Set Objectives and Measure Performance has to be given the necessary consideration from the beginning. If one view performance to relevant, then one has to consider it in the entire life cycle. The guide improves a structured and repeatable methodology to performance that one can implant into the application life cycle. It enables an individual to alleviate performance risk at the beginning of the project. One can work towards the set performance objectives, test and design for performance, measure and tune the performance of the entire life cycle. Actionable categories and areas of responsibility of engineering performance: Performance objectives allow one to know when the application attains the performance goals. Performance modeling gives a structured and repeatable means to achieving the performance goals. Design and Architecture framework enable one to restructure the production from an early stage. The performance, as well as scalability frame, allows one to arrange and prioritize the performance issues. Measuring gives one an opportunity to know whether the performance is towards the set goals and objectives. Providing clear segmentation assists architects, testers, developers and administrators comprehend their responsibilities in the application life cycle. Setting Performance Objectives The projects goals have to include measurable and attainable performance objective objectives have to be in consideration from the start in order to enable one attain the set aims and targets. Employ the planning phase to manage and control project risk to the right level of the project. Performance goals are unique in terms of the following: The response time; it is the time that it takes the server to act on a given request. Throughput; It is the number of requests that can observe the application in a unit time. Performance measures as applications of transactions in a second. Resource utilization; use of resources is the measure of server and network resources that are consumed by ones application. Resources entail the memory, disk I/O, CPU and network I/O. Workload; it requires the total number of active users, data volumes, and transaction volumes at a given time. One can identify the costs of resources on a per scenario basis. Scenarios might entail browsing a product catalog and placing an order. One can also measure resource costs for a given user load, by using a particular workload profile. A workload profile entails of a representative mix of the clients performing various tasks. Metrics Metrics is the mechanisms that one employs to ascertain the scenarios against the performance objectives. For instance, one can use the response time, throughput, as well as and resource utilization as per the metrics. The performance goals for each metric are the value that will be acceptable. Match the actual value of given parameters, to the objectives to verify that it is meeting or failing to meet the performance goals. Knowing the Budgets The funds represent a report of the maximum cost that a particular unit in the project can afford to pay in line with the key performance objectives. One needs to distinguish between budgets with performance targets. For instance, one might have a budget of 15-second response time. If it goes past the defined budget, the software has failed. Furthermore, one should set performance goals four to six seconds to leave room for amplified load from different sources. Also, one need to spread the budget between the various functionalities related to the processing the request. Funds are made to be specific depending on the terms of execution time, as well as the resource utilization, but also entail less tangible factors like the program resource costs. The budget will contain the following: Network; Network considerations will require bandwidth. Hardware; it will include payments items, such as servers, CPUs, and memory. Resource dependencies; it will entail items, like Web service connections and available database connections. Shared resources; will include payments items, such as the bandwidth that one has, the CPU amount that one get. If there is the share of the server with other different applications, and the amount of memory that one will get. The budgeting activity will enable one to determine if the set performance objectives are realistically attainable. Performance has got different meanings to different individuals, but performance and scalability are about meeting the set goals. Ones set goals, and targets illustrate how long a given r operations must take as well as the number resources acceptable for the actions to take at different load levels. The straight method to performance is to ignore it until positioning time. Nevertheless, many performances related shortfalls introduce a particular design, architecture, and technology choices that one makes at the beginning of the development cycle. Once the decisions are in line, and the application raised, the problems are quite difficult and very expensive to fix. The framework promotes a holistic, life-cycle based mechanism to performance where one engineer for performance from the star of the design for development, testing, and deployment. The engineering technique revolves around the principle of setting goals and measuring. When one always measure performance in the entire life cycle, one will know whether there is trending toward the set target objectives or still far away them. An essential tool that will assist an individual in the performance process is the performance modeling technique. Performance modeling gives a structured as well as repeatable mechanism for modeling the performance indicators of the software. In the entire engineering performance, a balanced technique is necessary. A risk management approach will help one to decide on how far to go into any given profession and helps one to determine the situation at which further analysis is premature. \ Work cited Woodward M. Computer and telecommunication System Performance Engineering. London: Pentech, 1994. Read More