Data-Intensive Text Processing with MapReduce. Jimmy Lin
Читать онлайн книгу.
2 http://www.dbms2.com/2009/05/11/facebook-hadoop-and-hive/
3 http://public.web.cern.ch/public/en/LHC/Computing-en.html
4As an aside, it is interesting to observe the evolving definition of large over the years. Banko and Brill’s paper in 2001 was titled Scaling to Very Very Large Corpora for Natural Language Disambiguation, and dealt with a corpus containing a billion words.
5As in, so stupid it couldn’t possibly work.
6This title was inspired by a classic article titled The Unreasonable Effectiveness of Mathematics in the Natural Sciences [155]. This is somewhat ironic in that the original article lauded the beauty and elegance of mathematical models in capturing natural phenomena, which is the exact opposite of the data-driven approach.
7On Exactitude in Science [23]. A similar exchange appears in Chapter XI of Sylvie and Bruno Concluded by Lewis Carroll (1893).
8What is the difference between cloud computing and grid computing? Although both tackle the fundamental problem of how best to bring computational resources to bear on large and difficult problems, they start with different assumptions. Whereas clouds are assumed to be relatively homogeneous servers that reside in a datacenter or are distributed across a relatively small number of datacenters controlled by a single organization, grids are assumed to be a less tightly-coupled federation of heterogeneous resources under the control of distinct but cooperative organizations. As a result, grid computing tends to deal with tasks that are coarser-grained, and must deal with the practicalities of a federated environment, e.g., verifying credentials across multiple administrative domains. Grid computing has adopted a middleware-based approach for tackling many of these challenges.
9The first example is Facebook, a well-known user of Hadoop, in exactly the manner as described [68]. The second is, of course, Google, which uses MapReduce to continuously improve existing algorithms and to devise new algorithms for ad selection and placement.
10Adapted from a post by Ted Dunning on the Hadoop mailing list.
11For more detail, Jacobs [76] provides real-world benchmarks in his discussion of large-data problems.
12See also DeWitt and Gray [50] for slightly different definitions in terms of speedup and scaleup.
13Note that this idea meshes well with utility computing, where a 100-machine cluster running for one hour would cost the same as a 10-machine cluster running for ten hours.
14Guess when this was written? You may be surprised.
15More than a theoretical model, the PRAM has been recently prototyped in hardware [153].
16Nevertheless, it is important to understand the relationship between MapReduce and existing models so that we can bring to bear accumulated knowledge about parallel algorithms; for example, Karloff et al. [82] demonstrated that a large class of PRAM algorithms can be efficiently simulated via MapReduce.
CHAPTER 2
MapReduce Basics
The only feasible approach to tackling large-data problems today is to divide and conquer, a fundamental concept in computer science that is introduced very early in typical undergraduate curricula. The basic idea is to partition a large problem into smaller sub-problems. To the extent that the sub-problems are independent [5], they can be tackled in parallel by different workers—threads in a processor core, cores in a multi-core processor, multiple processors in a machine, or many machines in a cluster. Intermediate results from each individual worker are then combined to yield the final output.1
The general principles behind divide-and-conquer algorithms are broadly applicable to a wide range of problems in many different application domains. However, the details of their implementations are varied and complex. For example, the following are just some of the issues that need to be addressed:
• How do we break up a large problem into smaller tasks? More specifically, how do we decompose the problem so that the smaller tasks can be executed in parallel?
• How do we assign tasks to workers distributed across a potentially large number of machines (while keeping in mind that some workers are better suited to running some tasks than others, e.g., due to available resources, locality constraints, etc.)?
• How do we ensure that the workers get the data they need?
• How do we coordinate synchronization among the different workers?
• How do we share partial results from one worker that is needed by another?
• How do we accomplish all of the above in the face of software errors and hardware faults?
In traditional parallel or distributed programming environments, the developer needs to explicitly address many (and sometimes, all) of the above issues. In shared memory programming, the developer needs to explicitly coordinate access to shared data structures through synchronization primitives such as mutexes, to explicitly handle process synchronization through devices such as barriers, and to remain ever vigilant for common problems such as deadlocks and race conditions. Language extensions, like OpenMP for shared memory parallelism,2 or libraries implementing the Message Passing Interface (MPI) for cluster-level parallelism,3 provide logical abstractions that hide details of operating system synchronization and communications primitives. However, even with these extensions, developers are still burdened to keep track of how resources are made available to workers. Additionally, these frameworks are mostly designed to tackle processor-intensive problems and have only rudimentary support for dealing with very large amounts of input data. When using existing parallel computing approaches for large-data computation, the programmer must devote a significant amount of attention to low-level system details, which detracts from higher-level problem solving.
One of the most significant advantages of MapReduce is that it provides an abstraction that hides many system-level details from the programmer. Therefore, a developer can focus on what computations need to be performed, as opposed to how those computations are actually carried out or how to get the data to the processes that depend on them. Like OpenMP and MPI, MapReduce provides a means to distribute computation without burdening the programmer with the details of distributed computing (but at a different level of granularity). However, organizing and coordinating large amounts of computation is only part of the challenge. Large-data processing by definition requires bringing data and code together for computation to occur—no small feat for datasets that are terabytes and perhaps petabytes in size! MapReduce addresses this challenge by providing a simple abstraction for the developer, transparently handling most of the details behind the scenes in a scalable, robust, and efficient manner. As we mentioned in Chapter 1, instead of moving large amounts of data around, it is far more efficient, if possible, to move the code to the data. This is operationally realized by spreading data across the local disks of nodes in a cluster and running processes on nodes that hold the data. The complex task of managing storage in such a processing environment is typically handled by a distributed file system that sits underneath MapReduce.
This chapter introduces the MapReduce programming model and the underlying distributed file system. We start in Section 2.1 with an overview of functional programming, from which MapReduce draws its inspiration. Section 2.2 introduces the basic programming model, focusing on mappers and reducers. Section 2.3 discusses the role of the execution framework in actually