03 Performance Metrics

03 Performance Metrics - CS-350: Fundamentals of Computing...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: CS-350: Fundamentals of Computing Systems Page 1 of 14Lecture Notes Performance Metrics of Computing SystemsIntroduction Performance metrics of computing systems are needed in order to evaluate the goodnessof a particular solution (architecture, protocol, resource allocation/management strategy, etc.) or compare that solution to other candidate solutions. Whose Performance? Depending on the objectives of the performance evaluation/comparison, performance metrics can be divided into two main classes: System centricand Process centric. System-centric (or resource-centric) performance metricsevaluate how well the overall system or any subset of its resources are being used, regardless of the quality of service (QoS) delivered to the individual entities (e.g. processes, transactions, database queries) making use of that system. Process-centric performance metricsevaluate how well a particular resource consumer (i.e. a specific process, a thread, a network connection, etc.) is being served, regardless of the overall efficiency of the system resources usage. One can understand the difference between the above two types of metrics by referring to our conception of processes as resource consumers. Basically, one can look at how well are the resources consumed and/or we can also look at how well served are the entities consuming the resources. System-centric and process-centric performance metrics do not necessarily go hand in hand. In particular, a solution that results in "poor" system-centric performance could well be providing "excellent" process-centric performance. System-centric and process-centric performance metrics do not necessarily go against each other either. Thus, it is possible for bothsystem-centric and process-centric performance metrics to be quite "good." How we quantify such goodness (i.e. "excellent," "poor," "acceptable," etc.) is the subject of this part of the course. A Motivating Analogy:To elucidate these concepts, we use an analogy (by now, you probably realize that I like to use analogies quite a bit!) Consider a restaurant with a maximum seating capacity of 50. From the perspective of the restaurant owner, the goal is to make the most money, and thus to make sure that there is no shortage of people waiting to be seated, and that every burner in the kitchen is being utilized all the time. This will imply that food is being sold and profit is being made at the maximal rate possible (given the kitchen resources, cook's capacity, waiters Azer Bestavros. All rights reserved. Reproduction or copying (electronic or otherwise) is expressly forbidden except for students enrolled in CS-350. CS-350: Fundamentals of Computing Systems Page 2 of 14Lecture Notes ability to serve food, etc.). From the perspective of a customer, the goal may be to eat as soon as possible (i.e. not having to wait in a long line, or not having to wait for a long time for food to be cooked, etc.) Obviously, getting the quickest possible service is unlikely when the restaurant cooked, etc....
View Full Document

This document was uploaded on 04/14/2011.

Page1 / 14

03 Performance Metrics - CS-350: Fundamentals of Computing...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online