liveBook · Manning

Multitenancy

As explained in Grokking Software Architecture

Multi-tenancy is an architecture in which a shared software system serves multiple customers or tenant organizations while maintaining separation of their data and workloads.

In the public cloud, multitenancy means your application shares physical hardware—such as CPUs and network cables—with workloads belonging to other customers. Think of an apartment: you have a private space, but share the building’s foundation and utilities. section 9.1.1

When a neighbor gets noisy

If a startup sharing your server suddenly uses large amounts of network bandwidth or disk speed, your application may slow down even when its code is well written. This unpredictable performance impact is the noisy-neighbor problem. section 9.1.1

Treat shared infrastructure as an architectural constraint: cloud performance is not fully under your application’s control, so plan defensively for resource contention. section 9.1.1 section 9.12

How Grokking Software Architecture covers it

How deep it goes
Defines shared infrastructure and illustrates noisy-neighbor effects, but stops short of detailing mitigation techniques.
Examples
Uses an apartment-sharing analogy and a startup consuming bandwidth or disk speed to illustrate resource contention.
What you can do after
Evaluate shared-cloud deployment risk by recognizing that another tenant’s resource use can slow an otherwise well-written application.
How it connects
Connects public cloud infrastructure to noisy-neighbor latency and defensive architecture.

Also explained in

Compare how 5 books explain multitenancy →

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.