liveBook · Manning

Scalability

As explained in Grokking Software Architecture

Scalability is the ability of a system to maintain acceptable behavior as its workload or size increases.

A link goes viral, and traffic jumps from 10 users a minute to 10,000 users a second. Can the system handle the rush without crushing itself under its own success? Scalability is the ability to handle increasing workloads, including by adding resources. It is a quality requirement: it describes how well a system handles growth, not what the system does. section 10.5 section 2.8

Scalability is related to performance, but they ask different questions. Performance concerns speed, including response time and throughput; scalability asks how well the system handles more load. section 2.4

Where should favorites live?

Putting favorites directly in user profiles is simple and quick to ship. But counting everyone who favorited one article requires reading every profile, so the work grows with the user population. A dedicated favorites storage area can grow independently of the main profiles, but needs a new table and more complex queries. section 1.5.5

When extra capacity is justified, two broad options are scaling up and scaling out. Scaling up makes one worker more powerful; scaling out adds more identical workers. Independent scaling can also let a busy service add capacity without scaling every other service, though services must coordinate over a network. section 10.5.1 section 12.1

Figure 10.7 Scaling Up (Vertical) makes a single worker more powerful, but Scaling Out (Horizontal) offers a practically infinite, resilient scale by adding more identical workers.

Compare a more powerful single worker with multiple identical workers. section 10.5.1

Good enough for now

An internal CSV export generated directly in a web request is quick to build, but it can time out for users with vast amounts of data. Accepting that limit may be reasonable for an immediate need; security, reliability, and data integrity may still be non-negotiable. section 2.5.1

Scale is a business decision, not a contest to build the biggest system. Describe the expected workload, the consequences of falling short, and the cost of extra capacity. An internal tool for 500 employees may not need infrastructure built for a viral traffic spike. section 10.5.1 section 2.5.1

How Grokking Software Architecture covers it

How deep it goes
Defines scalability and contrasts scaling up with scaling out, alongside practical costs and storage tradeoffs.
Examples
Illustrates growth with favorites storage, a synchronous CSV export, and a viral traffic spike.
What you can do after
Evaluate scalability choices against expected demand, implementation complexity, and business cost.
How it connects
Connects scalability with performance, maintainability, resilience, and non-functional requirements.

Also explained in

Compare how 37 books explain scalability →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Grokking Software Architecture. The section links point to the parts of the book it draws on.