A Framework for the Study of Grid Inter-Operation Mechanisms
The study of the history of computing infrastructures reveal(I rate highly my own work...)
A Framework for the Study of Grid Inter-Operation Mechanisms
The study of the history of computing infrastructures reveals an integration trend. For example, the explosive growth of the Internet in the 1990s was the result of an integration process started in the 1960s with the emerging networks of computers. By using the Internet, millions of users were capable of accessing information anytime and anywhere, much like other daily utilities such as water, electricity, and telephone. However, an important category of users remained under-served: the users with large computational and storage requirements, e.g., the scientists, the companies that focus on data analysis, and the governmental departments that manage the interaction between the state and the population (such as census, tax, and public health). Thus, in the mid-1990s, the vision of the Grid as a universal computing utility was formulated. The main benefits promised by the Grid are similar to those of other integration efforts: extended and optimized service of the integrated network, and significant reductions of maintenance and operation costs through sharing and better scheduling.
While the universal Grid has yet to be developed, large-scale distributed computing infrastructures that provide their users with seamless and secure access to computing resources, individually called Grid parts or simply grids, have been built throughout the world -- in different countries, for different sciences, and both for production work and for computer-science research. At the same time, the main technological alternatives to grids, that is, supercomputers and large clusters, have evolved into much larger, scalable, and reliable systems. Thus, the integration of existing grids into larger infrastructures and finally into The Grid is key in keeping the grid vision attractive for its potential users.
The integration of grids raises a double challenge, one related with the efficient scaling of a distributed computing system, the second associated with the operation of a system across different ownership and administrative domains. Thus, many of the traditional approaches for inter-operating computer systems, such as those based on completely centralized or purely decentralized system ar- chitectures, are eliminated from the start. To mark the distinction between the typical problem of integrating smaller components into a larger system and the double challenge of grid integration, we call the latter the problem of grid inter-operation. In this thesis we approach the problem of grid inter-operation with two main objectives: to design a comprehensive framework for the study of grid inter-operation mechanisms, and to provide an initial but good solution for this problem.
Our framework provides both the theoretical support and the tools for finding new and improved solutions for this problem. The tools are assembled into a research toolbox for the study of grid inter-operation mechanisms. This research toolbox addresses two problems that have hampered the grid community in the past decade: the lack of knowledge about the workloads and resources of real grids, and the lack of tools for grid simulation and performance evaluation in real environments. Research using unrealistic characteristics or characteristics that are specific to other types of environments is being limited in scope and applicability, and may even miss the problems that are specific to grids. Thus, real data and realistic models of grid workloads and resources are critical for designing efficient and scalable architectures. Using for simulation and for performance evaluation in real environments tools that have not been adapted to the requirements of grids leads to slower progress and to results that are difficult to compare. Thus, tools adapted to grids and aimed at producing results that can be shared with other researchers are needed.
The contents of this thesis is split into four logical parts: the introduction, a toolbox for grid inter-operation research, a method for grid inter-operation, and the conclusion.
We begin the thesis with an introduction to the problem of grid inter-operation that focuses on the challenges of grid inter-operation addressed by this thesis. In Chapter 1 we also present an overview of the framework for the study of grid inter-operation mechanisms introduced in this thesis. In Chapter 2 we introduce a basic model for grid inter-operation. This model, required to understand the remainder of the thesis, defines the components of a grid system, the types of applications that can be found in a grid, the system users, and the grid job execution model.
The toolbox for grid inter-operation research is described in Chapters 3, 4, 5, and 6, which we describe in turn. In Chapter 3 we present the Grid Workloads Archive (GWA). We design the GWA with a focus on building a grid workload data repository, and on establishing a community center around the archived data. One of the important design achievements is the formulation of a grid workload format for storing job-level information that can be extended for higher-level information such as co-allocated jobs or resource reservations. We develop a comprehensive set of tools for collecting, processing, and using grid workloads. To make the GWA accessible by non-expert users, we devise a mechanism for automated trace ranking and selection. So far, the GWA contains traces from nine well-known grid environments, with a total content of more than 2,000 users submitting more than 7 million jobs over a period of over 13 operational years, and with working environments spanning over 130 sites comprising 10,000 resources.
In Chapter 4 we describe the extension of the basic model for grid environments into a comprehensive model for (multi-)grids. By analyzing real data such as long-term system traces of real grids, we find that grid resources exhibit a highly dynamic availability both over the course of single days and over whole years. We also find that grid workloads are very different from the workloads of other related systems such as parallel production environments and distributed web servers. Based on the results of this analysis, we design and validate a comprehensive model for grid resource dynamics and evolution, and for grid workloads that include parallel jobs and/or bags-of-tasks.
In Chapter 5 we introduce the GrenchMark testing framework. The main focus of this framework is on testing large-scale distributed computing systems with synthetically generated yet realistic workloads. We test and validate our reference implementation of the GrenchMark framework, and show that GrenchMark has been successful in testing real multi-cluster grids and pools of resources. The experimental results show that a grid testing tool focusing on realistic workloads can indeed be used to assess important characteristics of real systems that are otherwise not available, such as scalability limits, overheads, and reliability.
To conclude the presentation of our grid research toolbox, in Chapter 6 we introduce the DGSim grid simulation framework. The main focus of this framework is on facilitating repeated simulations of multi-cluster and multi-grid environments under realistic workload. We test and validate our reference implementation of the DGSim framework, and show that DGSim has been successful as the simulation tool for several design space exploration studies of grid settings that are larger than the previous state-of-the-art.
The method for grid inter-operation and a solution for the grid inter-operation problem are described in Chapters 7 and 8, which we describe in turn. In Chapter 7 we study the existing alternatives for grid inter-operation, and introduce a novel architecture for grid inter-operation. We classify real grid systems according to their architectural and operational components. The practical limitations of the centralized grid inter-operation approaches are evaluated in a real environment. These two preliminary steps allow us to assess the grid inter-operation ability of existing grid resource management systems; we find that this ability is limited. Thus, we introduce a novel architecture for grid inter-operation with a better potential of fulfilling the requirements of grid inter-operation. The architecture is a hybrid between hierarchical and purely decentralized architectures. The set of architectures investigated here provides a comprehensive architectural space for the problem of grid inter-operation.
In Chapter 8 we introduce a novel approach for grid inter-operation, Delegated MatchMaking. Our approach, which couples the hybrid architecture introduced in the previous chapter with a novel inter-operation mechanism, is compared with five alternatives through trace-based simulations, and is found to deliver the best performance especially when the system is heavily loaded. While many other mechanisms can be designed in the future, our experiments prove that the Delegated MatchMaking approach already is a good solution for the problem of grid inter-operation. Our experiments also demonstrate that the inter-operation of existing grids can lead to significant performance gains in comparison with leaving them operate independently.
At the end of this thesis, Chapter 9 summarizes our main achievements and presents future direc- tions for this work. The direct use of the framework for the study of grid inter-operation mechanisms holds good promise for future research. In particular, "How many clusters are best?" and other related questions about the system structure can find answers under this framework, leading to important contributions to automating system provisioning and administration. With extensions, our framework can be used to investigate important classes of resource management problems, such as mechanisms and incentives for more system decentralization, scheduling for specific classes of applications or scheduling under less strict information availability assumptions, and guarantees for Quality-of-Service for commercial workloads. We have already taken initial steps in several of these directions....more
Katherine Losse's The Boy Kings claims to be a memoir written by former Facebook employee #51. Graduated from top-level liberal arts colleges with a PKatherine Losse's The Boy Kings claims to be a memoir written by former Facebook employee #51. Graduated from top-level liberal arts colleges with a Ph.D. in English literature, Kate navigates jobs until reaching Facebook. Initially attracted by a customer service job, she finds herself trapped in a world of tecchies and dreams of 'world domination (her words), which conflicts strongly with her liberal arts education and personal prejudice. Unable to cope, she descends into a miserable state, until progressive promotions and eventually financial independence liberate her. She stops working for Facebook and writes this book. Voila!
(I got to read The Boy Kings through a chain of recommendations: I was reading Laszlo Bock's Work Rules!, which mentioned Dave Egger's The Circle; because I have read before books that in my view Eggers plagiarized, I decided to check if others did not believe this was the case also for The Circle; this is how I reached The Boy Kings.)
Overall, a horrible read. There is little fact in this book, and lots of tormented soul. Even the title is misleading, because there seems to be little heart in the money-grabbing employee who quits as soon as the shares vest and then writes a revenge book. Very bad book, if only I could have avoided reading this!
1. Some of the critique of the tech geek seems reasonable: the early decision of how life will work, the quest for structure and algorithm, the belief in tech as the ultimate defense of reason, etc. (Negative here: the author calls this critique analysis, but it is always negative and devoid of constructive advice).
2. "The only thing more powerful than celebrity is to own the tool that makes it" (Loc. 1399).
3. Surprisingly reasonable insight into her own state of mind appears from place to place, such as "I hate Judge-book, I hate rankings, I hate algorithms, I thought, in a moment of total rage at everything—the company, these boys—that was near, but also far beyond my control." followed by "I just wanted to be happy and loved for who I was" (Loc. 1870); and "As much as I had once made fun of the Facebook boys for staring at their phones more often than they looked up, I had become one of them." (Loc. 2122). Because it contrasts so much with the (willing?) hypocrisy, it makes the latter so much more difficult to palate.
4. The author does have a few humble moments, such as "His friend wanted to come to Coachella but couldn’t afford it, and I was reminded how lucky I was that Dustin gave me his ticket." (Loc. 1622) Better than nothing, but so very rare.
5. She identifies the true issues early Facebook employees had, relative to top talent brought in later: "Chamath was young, brash, and masculine in style but, unlike most Facebook engineers, he had experience managing a company." (Loc. 1965). Unfortunately, this happens two-thirds into the book. Another, about how Facebook had "compartmentalized just like in the American institutions we had wanted to leave behind.", appears three-quarters into the book.
6. Some pieces of very good text, such as "Sometimes, that year, I got a sick feeling in my stomach that I didn’t want this world in which we are all ranked virtually, by virtual strangers, on the basis of popularity and appearance. Even worse, I felt like I might not have a choice in the matter." (Loc. 1779) or, simply, "It was the Normandy of technology wars" (Loc. 2150).
7. In the end, she understands her role at this company: "Lol, I thought. That was a good description of my entire job. I was only important because he [n.b. Mark Zuckerberg] is." (Loc. 2869)
Unfortunately, the ugly things are overwhelming:
1. The first quarter of the book reads like an ugly attack of a desperate person on a former lover. Very, very poor manners, style, and credibility. She hates her perspectives in life. She hates Baltimore. She hates all men using AOL, whom she believes are all out to seduce and abandon her. She hates Johns Hopkins. She hates her first job. She hates male employees at Facebook, whom she decides are immature (she uses "juvenile") and entitled (wannabe kings). etc. Some of these aspects reappear througout the book, as if sprinkled by a careful copy-editor.
2. There is little in the aspects criticized in this book that is specific to Silicon Valley. The critique that Silicon Valley tries to present itself as different from the rest of the dog-eat-dog US economy is valid. In contrast, intense competition, unequal pay for unequal value created for the company, courting of superstars, much more relaxed workplaces, increased technology solutionism (for a much better account, read To Save Everything, Click Here: The Folly of Technological Solutionism), etc., are all general mores of the modern economy. We can criticize them, but not ascribe them exclusively to the Silicon Valley.
3. The feminism is here taken to levels where it acts as clear sexism. The negative traits are masculine. The negative examples involve exclusively men. Generalizations from single examples abound. As a minority student and then employee often at the receiving end of generalizations, I cringe at this approach.
4. There is much that points out that the author is often driven by self-interest, rather than concern for other women. Her condescending remarks towards the other customer support personnel of identical gender. Sheryl Sandberg and the author's neutral reaction to her, even when the conditions for women improve markedly at Facebook as a result of Sheryl's work. Also, after so much blaming towards the others (in particular of of being "boy kings"), she gets "a heady feeling" and talks about being "queen of the world", simply because she was able to claim the email address Kate@facebook.com. She is particularly happy that no other Kate will match her singularity, "like a new country in which I was the only Kate there, queen of a world in which every other Kate would be derived from my archetype". What a self-aggrandizing person!
5. The writing abounds in rhetorical tricks. For example, early on, she describes the five high-visibility hires from the Seattle areas as the "gang of five", then goes on to say that they looked like a fraternity, only to then go to great lengths to describe how malign fraternities are in general. This collocation leads the reader to associate the negative aspects to the five hires. Convicting by association and without proof is what we call bigotry. (Also, sororities are excluded from the conversation.) As a second example, she discusses how the engineers were sexual predators, then gives a personal example involving a guy from sales. Sexual harassment is very serious, but so are vile words that ruin collective reputations.
6. That the author hates tech is thinly veiled, but only stated clearly at the end. She also despises engineers, which is perhaps worse, and collides with her so much touted humanity. "I didn’t want to live in a world where I appeared only for a bunch of engineers to judge me and shoo me away." (Loc. 1859) is but one example.
7. The inconsistency in her positions begs questions of ethical writing. She may think of herself as an ethical person, but cheats on her first job and thinks it's ok because they don't pay her much; she first sells her shares and then discusses the issue with Facebook, because "she worked for it". The author first complains engineers spend too little time with her (because they love tech and the virtual, see?), then expresses her happiness to be alone. The author first complains about how she does not care about being pretty, then remarks "What would happen to me? I wondered. Was I pretty enough to make it past the bouncers?" (Loc. 1856). etc.
8. Her confused relationship with Thrax deserves special mention. He is introduced as a ghostly white hero, and then plays a sometimes on sometimes off role throughout the book. He acts conveniently for her and very out of character for the hero he is in the rest of the book "we stopped short of a kiss. “I can’t have a relationship story show up in News Feed,” he explained". When they finally have sexual contact (that does not seem like making love), it is because she has agreed to temporarily submit. The story goes on and on, tormented.
9. The author indicates to have never understood her role at Facebook, which was low in value relative to tech and top management positions. All the complaints derived from it sound hypocritical, because in the end she enjoys her relative privilege and position. For a lightweight example, she wants to change the world, but does not complain when she gets a position in which she can afford skipping days of work or doing much at all ("These were recent roles that Mark had invented, jobs that were not so much about doing things as being something" (Loc. 2595)). Hypocrisy (or just flip-flopping) also appears related to her desire to be valued for herself, when her actions were to shut down personal communication... unless it suited her better. "I liked my autonomy, my privacy, the fact that I was different from everyone else—a unique individual." (Loc. 2685) contradicts with "I was only important because he [n.b. Mark Zuckerberg] is." (Loc. 2869)
10. The authors has never understood how her complaints about money seem to anyone who is not a rich person. She mentions early "my hack: to live as richly as possible with next to nothing" (Loc. 841), when she was paid over 20$/hour at Facebook; at the same time, people in the same region could not get a decent minimal wage, and others were getting by or nearly starving in San Fran; not to mention an entire world of hunger, outside the US. She stops mentioning the salary once it has become sufficiently high.
etc. etc. etc.
In the end, I was left wondering if the author has ever considered that:
1. If she would have been good enough to stay in the academia, the engineers in the building would have been the marginalized and little appreciated employees.
2. If she was better at her skills, say in the excellent class, she would have been recruited for a better, more visible position. See Sheryl Sandberg.
3. The balanced payment approach proposed by her is already used in many places in the US academia, and in many places in the world, both industry and academia. Many countries in Europe make payment balanced through pressure from a trade union. Maybe she would have been happier in those places. In contrast, fast-growing startups, banks, and many other competitive industries pay in the US using a very imbalanced pay structure, see for example Work Rules! for imbalances in payment at Google (and the reasoning behind it). ...more
Overall, I really liked this book. I learned much and discovered even more. Multumesc mult, Laszlo!
1. In the end, and in bits throughout the book, the reader discovers that Google uses not so much a ground-breaking process, but rather a data-driven iteration of well-known HR (and to some extent also managerial) processes.
2. Very good analysis of many HR processes, including detailed and important references. I particularly liked the identification of references from a few decades ago, such as Andrew S. Grove's High Output Management (Intel processes, mid-1990s). I enjoyed the summary dismissal of tradition: "Command-oriented, low-freedom management is common because it’s profitable, it requires less effort, and most managers are terrified of the alternative." Also, good reference to Dave Eggers' The Circle (2013), a dystopian novel that seems to describe Googlife.
3. The analysis of the "two tails", the best and the worst performers, is nicely done. In traditional management, with narrow remuneration bands, best-performers should always quit after a great delivery, to seek to maximize their value through competitive market forces. Good observation that "most talented people on the planet are increasingly physically mobile, increasingly connected through technology, and—importantly—increasingly discoverable by employers.". At Google, they are rewarded much closer to their contribution. In traditional management, worst-performers are fired, and failures are never acceptable. At Google, risk is encouraged and failure from which much is learned is rewarded.
4. The simple but powerful idea of using checklists, including the 10-point checklist that summarizes the book.
5. Objectives and Key Results (OKRs) must be specific, measurable, verifiable (so, achievable, relevant, and timed). SMART criteria exist already for many decades (Laszlo cites I believe George Doran's 1980s paper on "the S.M.A.R.T. way"), but here OKRs are revised each quarter and employees are supposed to set goals that far exceed their performance (and results of "achieved 70%" are considered very good).
Other positive aspects:
6. The inspiring text about "trusting first", which works in an environment of positive and ethical people.
7. The notion that "public recognition is one of the most effective and most underutilized management tools". Very good point, albeit gamification could have been mentioned more here. The cafeteria setup and the difference between a job, a career, and calling also points to gamification, with its many tracks of advancement and support for Achievers/Explorers/Socializers/Winners.
8. The simple but powerful idea that full transparency is necessary in modern institutions. (This goes in contrast to the politicking seen in so many traditional companies.)
9. The focus on 'spreading the wealth', here, to make the best share their knowledge to others, and to analyze the best vs the worst to identify true best-practices.
10. Description of new interview practices at Google. All more standardized, enabling cross-comparisons. Focus also on candidate experience with the process. Loved the data.
11. Success at job depends on personal 'scope, impact, and leadership'. Title follows leadership, and, even then, no more pompous titles. Also helps with retention: bad for people trying to move to another company, because it is more difficult to explain what your work was about.
12. Googlegeist as tool to collect feedback about each person, also from peers.
13. Performance assessment focusing on personal development, instead of ratings and rewards. At least the two processes should be separated. (This is an old HR approach, with obvious pros and cons.)
14. Lesson learned: "Expanding the proportion of people receiving the top rating better reflected their actual performance".
15. The calibration processes used at Google, especially the peer-review of decisions by collectives of managers, match those used in so many other companies...
16. The discussion about primary and secondary education vs training is very interesting. In short: annually, companies spend on training about a quarter of what is spent on primary and secondary education, but get less than a tenth of the results of education. In the US, $156 billions spent in 2011 for training that resulted in disappointingly little.
17. Discussion about training practices that work, mention to Ericsson's "deliberate practice: intentional repetitions of similar, small tasks with immediate feedback, correction, and experimentation. Simple practice, without feedback and experimentation, is insufficient."
18. Interesting observations about what many of us do. Among others: how we ascribe aesthetic and personal value based on how much we paid.
19. Excellent tips for onboarding starters. This follows up on the long-running thread on the importance of having high retention of employees (is it useful? or just a way to justify HR's practices? the author does not address these questions)
20. Laszlo's hierarchy of needs for HR departments.
(Personally, I also liked that the author presents a chain of though that matches my own "Modern West vs Old Russian education system" analogy: "You either believe people are fundamentally good or you don’t. [...] If people are good, they should be free. [...vs...] Taylor, who told Congress in 1912 that management needs to tightly control workers, who were too feeble-minded to think for themselves". Very funny!)
1. Running against own claims (see main positive 1), the author tries occasionally to emphasize how new a part of the process is. (The title is an example in this sense.) We see claims of novelty regarding processes and mechanisms that have been identified and studied before, sometimes even decades before Google started using them. This claim for novelty could be correct, as much larger scale and a very different environment can change things, but not if the findings are the same and the process seems to have been trivially adapted. There is one more inconsistency here, in that for some of the processes (such as awards), even Google only has a few samples (real people put under the microscope) to base its decisions upon.
2. The defense of failed Google products. Wave, "an entirely new way of interacting online"?! Please, more geek speak and less corporate talk.
3. The unnecessarily manipulative text. The ode to HR departments. The constant jibes at Yahoo and other competitors of Google; Marissa Meyer is in particular a target. The thinly veiled attempt to discredit competition. For example, near the end, the author identifies several major companies that now use People Operations instead of Human Resources; he immediately claims having spoken to one of the top people running this at a nong-Google organization, who purportedly claims that it's just a word trick, not the real thing ("it’s just regular HR. We just like calling it that [n.b.: People Operations]."). As another example, Bill Gates is quoted out of context with a complaint that his foundation's actions do not get the same recognition as Google's (smaller) deeds; this makes Bill Gates sound petty, whereas his claim is correct: eradicating malaria in Africa vs a mug-shot.
4. (a critique on Google, rather than this book:) Nudges. Nudges are ways to influence behavior, stopping only short of enforcing it; for example, building a corridor with only one exit would enforce using it, but building one with two doors, of which one is highlighted, would nudge people to use the highlighted door instead of forcing them to do so. The argument for using "nudges" at Google looks very similar to what Big Brother would argue in 1984 (all is fair for the greater good). The extent to which Google seems to already apply nudges is already scary (hint: everything is measured, many things are engineered to manipulate people). "Would the results hold for thousands of Googlers?" seems a question commonly asked; "Would it be ethical to try?" not so much. (Recently, Facebook has been involved in a scandal) As a consequence, I am reconsidering my career options.
5. Many of the stories are personal, often funny, but do not advance the cause of this book. They do match the author's useful story 'Frank Flynn, a professor at the Stanford Graduate School of Business, once told me the secret to high student evaluations: “Tell lots of jokes and lots of stories. Grad students love stories.” He went on to explain that it’s a constant trade-off between being engaging and imparting knowledge.'
TODO: +++ Google chief tells their story. Enough said. +/- main message: Google same as previous tech companies, but for the focus on A-level personnelTODO: +++ Google chief tells their story. Enough said. +/- main message: Google same as previous tech companies, but for the focus on A-level personnel. So it's about human operations. - Less interesting than Laszlo Bock's book about Google (human) operations -- too much about Eric Schmidt 's opinions. --- too much trivia about minute details. Is Eric like this also as a manager?...more