The MongoDB Architecture

MongoDB enables you to meet the demands of modern apps with a developer data platform built on several core architectural foundations. It lets you access the best ways to innovate in building transactional, operational, and analytical applications. This chapter examines the MongoDB architecture with a special emphasis on two key elements: replication and sharding.

Replication is a crucial component in MongoDB's distributed architecture, ensuring data accessibility and resilience to faults. It enables you to spread identical datasets across various database servers, safeguarding against the failure of a single server.

Additionally, you will learn about sharding, a horizontal scaling strategy for spreading data across several machines. As applications grow in popularity, and the volume of data they produce increases, scaling across machines becomes essential to ensure sufficient read and write throughput.

This chapter will cover the following topics...

Replication vs sharding

People often confuse replication with sharding. While both are sets of systems utilized in database management, they serve distinct purposes and are employed for different reasons. Replication is a process where data is duplicated and stored in multiple locations to ensure redundancy and reliability, playing a vital role in data protection and accessibility.

On the other hand, sharding involves dividing a larger database into smaller, more manageable parts, called shards. Each shard stores a portion of the total dataset on a separate database server instance. However, it's important to note that each shard must also implement replication to maintain data integrity and availability.

The goal of combining sharding with replication is to ensure data durability and high availability. When a shard's server instance fails and there's only a single copy of data on that shard, it can result in unavailability of data until the server is restored...

Replication

A replica set in MongoDB refers to a collection of mongod processes that uphold the same dataset. They offer redundancy and high availability, serving as the foundation for all production implementations. By having numerous data copies across various database servers, replication ensures a degree of fault tolerance, protecting against the failure of a single database server.

Figure 2.1: A replica set

The primary node handles all write operations, and logs all dataset changes in its operations log (oplog). A MongoDB replica set can only have one primary node.

The secondary nodes replicate the primary's oplog, and implement the operations on their own datasets, ensuring that they mirror the primary's dataset. In the event of the primary becoming inaccessible, a qualified secondary will initiate an election to become the new primary.

With data stored across multiple servers, replication increases the reliability of the system....

Sharding

MongoDB supports horizontal scaling through sharding. Sharding involves the distribution of data across numerous processes and plays an essential role in managing and organizing large-scale data. This method divides a larger database into smaller, more manageable components known as shards. Each shard is stored on a separate database server instance, which distributes the load and offers an effective approach to data management. Moreover, this technique allows for the creation of distributed databases to support geographically distributed applications, enabling policies that enforce data residency within specific regions.

Why do you need sharding?

Consider a scenario where your data is expanding swiftly and your database is approaching its maximum capacity. This circumstance could manifest a multitude of challenges. Typically, the most pressing problem is performance deterioration. As your database grows, the time required to query and retrieve data can increase significantly...

New sharded cluster features in MongoDB 7.0

MongoDB 7.0 continues to simplify the management and understanding of sharded clusters for both operations and developer use cases. This version provides additional insights that assist in making the best decisions for both initial and future shard key selections. Furthermore, starting in MongoDB 7.0, developers can experience a consistent interface when using these commands on sharded or unsharded clusters, while still retaining the option to optimize for performance when necessary. Let's have a look at the new sharding features introduced in MongoDB 7.0.

Shard key advisor commands

Choosing a shard key is complex due to intricate data patterns and trade-offs. Yet, the new features in MongoDB 7.0 aim to ease this task:

• analyzeShardKey provides the ability to evaluate a candidate shard key against existing data. Shard key analysis in MongoDB 7.0 offers metrics to evaluate a shard key's suitability, including its...

Summary

Replication in MongoDB is a process that synchronizes data across multiple servers, providing redundancy and increased data availability. This is achieved through replica sets, a group of MongoDB servers that maintain the same dataset. Within a replica set, one node acts as the primary node, receiving all write operations, while the secondary nodes replicate the primary's operations to their datasets. This structure provides a robust system for failover and recovery. If a primary node fails, an election among the secondaries determines a new primary, allowing for continuous client operations.

Sharding in MongoDB is a method for splitting and distributing data across multiple servers or shards. Each shard is an independent replica set, and collectively, the shards make up a single logical database—the sharded cluster. This approach is used to support deployments with very large datasets and high-throughput operations, effectively addressing scalability issues...