Site Reliability Engineering

Reliability Targets

Set and defend realistic reliability goals.

30 min read intermediate 3 objectives

Status

Not started

What you will learn

  • Set SLOs from user needs
  • Use error budgets
  • Balance reliability and features

New to this? Start here

The basics, in plain English

Reliability targets are clear numbers that say how dependable a service should be, so the team is not just guessing. They are built from a measured indicator, a target goal, and a small budget for failure. These numbers help decide when to focus on stability versus new features.

Reliability target
A specific number describing how dependable a service must be.
SLI
The measured number, like the percent of requests that succeed.
SLO
The goal you aim for, like 99.9 percent of requests succeeding.
Error budget
The small amount of failure allowed before you break the goal.
Nines
Shorthand for uptime levels, like “three nines” meaning 99.9 percent.
Trade-off
Using the budget to balance shipping features against staying stable.
01

100% is wrong

Perfect reliability is unaffordable and unnecessary. Pick an SLO users accept, then spend the error budget on shipping faster.

Finished this topic?

Mark it done to earn 100 XP and keep your streak alive.