book

MongoDB Applied Design Patterns

by Rick Copeland

March 2013

Intermediate to advanced

174 pages

3h 58m

English

O'Reilly Media, Inc.

Read now

Unlock full access

AudienceAssumptions This Book MakesContents of This BookPart I: Design PatternsPart II: Use CasesConventions Used in This BookUsing Code ExamplesSafari® Books OnlineHow to Contact UsAcknowledgments
Relational Data Modeling and NormalizationWhat Is a Normal Form, Anyway?So What’s the Problem?Denormalizing for PerformanceMongoDB: Who Needs Normalization, Anyway?MongoDB Document FormatEmbedding for LocalityEmbedding for Atomicity and IsolationReferencing for FlexibilityReferencing for Potentially High-Arity RelationshipsMany-to-Many RelationshipsConclusion
Polymorphic Schemas to Support Object-Oriented ProgrammingPolymorphic Schemas Enable Schema EvolutionStorage (In-)Efficiency of BSONObject-Document MappersPolymorphic Schemas Support Semi-Structured Domain DataConclusion
The Relational Approach to ConsistencyCompound DocumentsUsing Complex UpdatesOptimistic Update with CompensationConclusion
Storing Log DataSolution OverviewSchema DesignOperationsInserting a log recordBulk insertsFinding all events for a particular pageAside: managing index sizeFinding all the events for a particular dateFinding all events for a particular host/dateRules of index designCounting requests by day and pageSharding ConcernsLimitationsOption 1: Shard by timeOption 2: Shard by a semi-random keyOption 3: Shard by an evenly distributed key in the data setOption 4: Shard by combining a natural and synthetic keyTest with your own dataManaging Event Data GrowthCapped collectionsTTL collectionsMultiple collections, single databaseMultiple databasesPre-Aggregated ReportsSolution OverviewSchema DesignOne document per page per day, flat documentsOne document per page per day, hierarchical documentsSeparate documents by granularity levelOperationsLog an eventPre-allocateRetrieving data for a real-time chartGet data for a historical chartSharding ConcernsHierarchical AggregationSolution OverviewSchema DesignMapReduceOperationsCreating hourly views from event collectionsDeriving day-level dataWeekly and monthly aggregationRefactor map functionsSharding Concerns
Product CatalogSolution OverviewConcrete-table inheritanceSingle-table inheritanceMultiple-table inheritanceEntity attribute valuesAvoid modeling product data altogetherThe MongoDB answerOperationsFind products sorted by percentage discount descendingFind albums by genre and sort by year producedFind movies based on starring actorFind movies with a particular word in the titleConclusion: Index all the things!Sharding ConcernsScaling read performance without shardingCategory HierarchySolution OverviewSchema DesignOperationsRead and display a categoryAdd a category to the hierarchyChange the ancestry of a categoryRename a categorySharding ConcernsInventory ManagementSolution OverviewSchemaOperationsAdd an item to a shopping cartModifying the quantity in the cartChecking outReturning inventory from timed-out cartsError handlingSharding Concerns
Metadata and Asset ManagementSolution OverviewSchema DesignGridFSOperationsCreate and edit content nodesUpload a photoLocate and render a nodeSearch for nodes by tagGenerate a feed of recently published blog articlesSharding ConcernsStoring CommentsSolution OverviewApproach: One Document per CommentOperation: Post a new commentOperation: View paginated commentsOperation: Retrieve comments via direct linksApproach: Embedding All CommentsOperation: Post a new commentOperation: View paginated commentsOperation: Retrieve a comment via direct linksApproach: Hybrid Schema DesignOperation: Post a new commentOperation: View paginated commentsOperation: Retrieve a comment via direct linksSharding Concerns
Solution OverviewDesign 1: Basic Ad ServingSchema DesignOperation: Choose an Ad to ServeOperation: Make an Ad Campaign InactiveSharding ConcernsDesign 2: Adding Frequency CappingSchema DesignOperation: Choose an Ad to ServeShardingDesign 3: Keyword TargetingSchema DesignOperation: Choose a Group of Ads to Serve

Solution OverviewSchema DesignIndependent CollectionsDependent CollectionsOperationsViewing a News Feed or Wall PostsCommenting on a PostCreating a New PostMaintaining the Social GraphSharding
Solution OverviewSchema DesignCharacter SchemaItem SchemaLocation SchemaOperationsLoad Character Data from MongoDBExtract Armor and Weapon Data for DisplayExtract Character Attributes, Inventory, and Room Information for DisplayPick Up an Item from a RoomRemove an Item from a ContainerMove the Character to a Different RoomBuy an ItemSharding
Where Do I Go from Here?

Overview

Whether you’re building a social media site or an internal-use enterprise application, this hands-on guide shows you the connection between MongoDB and the business problems it’s designed to solve. You’ll learn how to apply MongoDB design patterns to several challenging domains, such as ecommerce, content management, and online gaming. Using Python and JavaScript code examples, you’ll discover how MongoDB lets you scale your data model while simplifying the development process.

Many businesses launch NoSQL databases without understanding the techniques for using their features most effectively. This book demonstrates the benefits of document embedding, polymorphic schemas, and other MongoDB patterns for tackling specific big data use cases, including:

Operational intelligence: Perform real-time analytics of business data
Ecommerce: Use MongoDB as a product catalog master or inventory management system
Content management: Learn methods for storing content nodes, binary assets, and discussions
Online advertising networks: Apply techniques for frequency capping ad impressions, and keyword targeting and bidding
Social networking: Learn how to store a complex social graph, modeled after Google+
Online gaming: Provide concurrent access to character and world data for a multiplayer role-playing game