Impianti di Elaborazione
computer Infrastructures Nazzareno Bordi
KNOWLEDGE AND UNDERSTANDING:
The course introduces concepts, models and methodologies of performance engineering for sizing of computer systems, such as the concept of dependability and the key efficiency and reliability metrics, Markov chains, queuing networks models, operational analysis.CAPACITY TO APPLY KNOWLEDGE AND UNDERSTANDING:
At the end of the course, students will be able to estimate the work load of the resources of a computer infrastructure and properly sizing these resources to meet dependability requirements and provide users with an acceptable quality of service at a reasonable cost.TRANSVERSAL SKILLS:
The resolution of problems of wokload estimation and sizing of processing equipment, realized in groups and individually, will help improve the degree of independent judgment by the ability to properly analyze and esign, the ability to communicate, also deriving from the teamwork, by defining a common (written and spoken) language, the ability to learn independently and to draw conclusions.
Content of lectures (47 hours):
computer systems and their architectures, performance engineering and its correlation with the life cycle of a computer system; quality performance of a computer system: properties and quality of service metrics, models for performance engineering: Markov and queuing networks, performance engineering methodology, case studies: services, databases, web services, data centers, call centers, services for e-business.
Classroom exercises (25 hours):
exercises on all the lectures' topics, for example: model design of queuing networks (open and closed) of single class or multiclass computer systems, calculation of stationary probability distributions, calculation of the performance of the computer system resources (service time, service demand, residence time, utilization, bottleneck, ...) and of the same system (throughput, response time, throughput times of a transaction, response time and throughput asymptotic, ...), calculation of the foreseeable development of the system performance as a result of variation of the needs and / or of the Quality of Service, .....
Development of the examination
LEARNING EVALUATION METHODS
the assessment of the student knowledge level consists of two tests: a written test with the solution of three exercises on topics covered in the course, to be completed in an hour; an oral exam, consisting in the discussion of at least three topics covered in the course. The written test is preparatory to the oral exam and the student must have earned at least a pass in the written test and the vote in based on thirty. The oral exam must be done in the same session of the written test or in the subsequent session. In case of a negative outcome for the oral examination, the student has to repeat the written test.
LEARNING EVALUATION CRITERIA
to successfully pass the examination, the student must demonstrate, through the trials described above, that has understood the concepts presented during the course and must prove that has understood the criteria and the procedures for the design and dimensioning of the processing systems.
LEARNING MEASUREMENT CRITERIA
each of the two tests indicated above has a score between zero and thirty thirtieths; the overall grade, based on thirty, is given by the vote assigned in the oral test.
FINAL MARK ALLOCATION CRITERIA
the evaluation is positive if the student achieves at least the sufficiency, equal to eighteen points, in each of the two tests described above; each correctly done written exercise worth 10 points; the highest rating is achieved by demonstrating a thorough understanding of the course content by the tests; full marks with honours is reserved for students who, having done all the tests so correctly, have demonstrated full knowledge of the course topics.
- D. A. Menascè, V. A. F. Almeida, L. W. Dowdy: Performance by Design: Computer Capacity Planning by Example, Prentice Hall PTR, 2004.
- G. Bracchi, C. Francalanci, G. Motta, Sistemi Informativi dImpresa, McGraw-Hill, 2010.
- Ingegneria Informatica e dell'Automazione (Corso di Laurea Magistrale (DM 270/04))