Performance Modeling and Design of Computer Systems: Queueing Theory in Action by Mor Harchol-Balter (PDF)

Description

Computer systems design is full of conundrums: •Given a choice between a single machine with speed s, or n machines each with speed s/n, which should we choose? •If both the arrival rate and service rate double, will the mean response time stay the same? •Should systems really aim to balance load, or is this a convenient myth? •If a scheduling policy favors one set of jobs, does it necessarily hurt some other jobs, or are these “conservation laws” being misinterpreted? •Do greedy, shortest-delay, routing strategies make sense in a server farm, or is what’s good for the individual disastrous for the system as a whole? •How do high job size variability and heavy-tailed workloads affect the choice of a scheduling policy? •How should one trade off energy and delay in designing a computer system? •If 12 servers are needed to meet delay guarantees when the arrival rate is 9 jobs/sec, will we need 12,000 servers when the arrival rate is 9,000 jobs/sec? Tackling the questions that systems designers care about, this book brings queueing theory decisively back to computer science. The book is written with computer scientists and engineers in mind and is full of examples from computer systems, as well as manufacturing and operations research. Fun and readable, the book is highly approachable, even for undergraduates, while still being thoroughly rigorous and also covering a much wider span of topics than many queueing books. Readers benefit from a lively mix of motivation and intuition, with illustrations, examples, and more than 300 exercises – all while acquiring the skills needed to model, analyze, and design large-scale systems with good performance and low cost. The exercises are an important feature, teaching research-level counterintuitive lessons in the design of computer systems. The goal is to train readers not only to customize existing analyses but also to invent their own.

User’s Reviews

Editorial Reviews: Book Description Written with computer scientists and engineers in mind, this book brings queueing theory decisively back to computer science. About the Author Mor Harchol-Balter is an Associate Professor in the Computer Science Department at Carnegie Mellon University. She is a recipient of the McCandless Chair, the NSF CAREER award, the NSF Postdoctoral Fellowship in the Mathematical Sciences, multiple best paper awards and several teaching awards, including the Herbert A. Simon Award for Teaching Excellence and the campus-wide Teaching Effectiveness Award. She is a leader in the ACM SIGMETRICS/Performance community, for which she recently served as Technical Program Chair, and has served on the Technical Program Committee twelve times. Harchol-Balter’s work integrates queueing theoretic analysis with low-level computer systems implementation. Her research is on designing new resource allocation policies (load balancing policies, power management policies and scheduling policies) for server farms and distributed systems in general, where she emphasizes integrating measured workload distributions into the problem solution.

Reviews from Amazon users which were colected at the time this book was published on the website:

⭐Unlike many of the other reviewers of this book, I am not a seasoned queueing professional, professor, or a student of Mor’s. I used this book for a course last semester (Spring 2014). It was my first year in the PhD program. I came from a liberal arts school where I double majored in CS and Math (unfortunately a bit lacking). Just so you get a sense for my background.Make no mistake: this is an excellent book. On many topics, the book is an abundantly clear source. The problems are challenging, and most of them fair. We made it through the first 23 chapters in our opening semester — and could have gotten through a few more, had we started out the semester properly. The other reviews cover the great aspects of the book, so I won’t say more on this. It really is a great book though.So why four stars then? I have some reservations about the book. First is the initial organization of the book. The overall organization is just fine. But the Introduction to Queueing section (Chapters 1 and 2) feels sorely misplaced. The book should start with the Necessary Probability Background section, then the Introduction to Queueing, and then the actual queueing chapters. Why? Because Chapter 2 provides good intuition for the rest of the book, and by the time you’ve made it through the probability review, it is not on the forefront of your mind. Similarly, in some of the problems, you’ll need “Slowdown”, which is only briefly defined in Chapter 2 and not again until Chapter 28. To any future users of the book, I would recommend re-reading Chapter 2 as you continue past the probability review. Maybe a couple of times. It’s an important one for mastering the concepts.Second, the book makes many assumptions about the mathematical background of the audience. Admittedly, the probability review is skimpy (William J. Stewart provides ~8 chapters of probability in his book), but at least it’s there. There are a few instances where it’s assumed you’ll recognize you need to use the limit definition of e^x and the taylor expansion of e^x. There are some derivations where a few steps are omitted that are necessary for clarity — for example, the steps between the last two lines of the derivation at the bottom of page 220. If you get confused at points, I would recommend trying to fill in the steps on your own in the textbook. Just be forewarned, if you’re mortal, you will feel lost at places.Third, I felt that there were some interesting pieces of queueing theory that were not emphasized in Mor’s book. Fortunately, these blurbs were well substituted by Adan & Resing, which is for free online (legally). On pages 38 and 39, Adan and Resign have great coverage of “Busy Periods,” a topic that seems to only appear in exercises in Mor’s book. Particularly, the equations in 4.6.1 are incredibly easy yet crucial for understanding queueing systems. Additionally, Mor was also quite skimpy on what she calls the “Resource Requirement” (defined on page 273). Adan & Resing don’t even acknowledge that they provide insight on the resource requirement. However, I implore you to do Exercise 25 in Adan & Resing and then look at the solution. Part (ii) of the question provides an interesting insight on how you can use the resource requirement. Or at least I, a non-expert, think its interesting. To be completely clear though, Mor’s book is better in nearly every way than Adan & Resings.Fourth, some of the problems are quite difficult. If you have a helpful TA or Prof then that should be okay. But be warned, if your TA and Prof are not helpful, AND you’re assigned one of the deathly hard problems, you’re absolutely screwed. This is not a fault of the book per say, but it should be known. On the other hand, I also found some of the problems did not pertain as well to the chapter as other ones, which is a bit annoying. I don’t want to struggle through a hard problem, fully knowing I will get little out of it.Last, I felt that there were some places the intuition was not highlighted very well. William J. Stewart spends quite a bit of time (and text) cultivating his reader’s intuition for certain topics, whereas again, Mor throws it down quickly and expects you to get the whole picture. To be fair, this only happened a few times and I can’t remember any specific examples right now, so take this blurb with a grain of salt.Outside of these issues, I also wish the book had a few more things. The first is more programming assignments. Perhaps we just weren’t assigned any of the good ones, but the few programming exercises we had were pretty trivial. There is much more potential for some simulation coverage. The second is case-studies at the end of every section. It would be nice to have a problem to pore over that integrates all of the topics in the past 5-8 chapters together. Essentially some mini hands-on experience that would be a decent representation of what an actual professional would do.Now that we’ve finished my desiderata, I want to reiterate, this book is very good. I just wanted to explain my four star ranking and cover some details other reviews might not have captured.

⭐I haven’t read the entire text yet but so far I like it a lot, with the exception noted below.Specifically regarding the formatting of the Kindle edition:Some but not all formulae and diagrams are formatted in a manner which is nearly impossible to read. It appears that some formulae are formatted in a way that uses a reasonable font which is comparable in size to the rest of the text, whereas other formulae are formatted as images which display at a size which is so small as to be very difficult to read.The problem does not seem to be significant enough for me to feel that this was a wasted purchase. (As has been the case with some other e-books having a much more severe manifestation of the same problem.)I have examined this using 2 devices, namely my Kindle Fire HD 8.9″ (2012-ish) and my Google Pixel C (2015 Android device with slightly larger screen than the Kindle, slightly lower resolution) running the Kindle App. The display is not identical on the two devices but is similar.

⭐My level of probability theory understanding was entry to intermediate. The book prepared me with enough probability theory material to enjoy the in-depth explanations of queue’s behaviors throughout the books. I would rate this book 5 star, if there are solutions to the exercises that I can reference.

⭐Applied mathematics at its best! If only all queueing theory books were like this one. Looking forward to the continuation with more analysis from the field.

⭐The book is awesome. The contents are great for self-study and the quality of the book is excellent.

⭐Very easy to understand. Small chapters and well written

⭐Nice book. I needed something well written and straightforward without being trivial while covering sufficient depth for practicing engineer to model performance of cloud based system

⭐Hands-on, practical advice for real world problems.

⭐Excellent book, thought for computer scientists.

⭐Very good book.

⭐Great book. Clear and useful for the subject.

Keywords

Free Download Performance Modeling and Design of Computer Systems: Queueing Theory in Action in PDF format
Performance Modeling and Design of Computer Systems: Queueing Theory in Action PDF Free Download
Download Performance Modeling and Design of Computer Systems: Queueing Theory in Action 2013 PDF Free
Performance Modeling and Design of Computer Systems: Queueing Theory in Action 2013 PDF Free Download
Download Performance Modeling and Design of Computer Systems: Queueing Theory in Action PDF
Free Download Ebook Performance Modeling and Design of Computer Systems: Queueing Theory in Action