Chapter 21. The Art and Science of the Service-Level Objective
You can’t meet or exceed expectations if no one agrees what the expectations are. In every job, you must understand your objectives to measure your success. In this chapter, we learn what it looks like when an SRE sets goals.
Why Set Goals?
The main objective of an SRE is to maintain the reliability of systems. Service-Level Objectives (SLOs) are the primary mechanism used by SREs to determine success in this objective. I’m sure you can see that it is difficult to “do your job well” without clearly defining “well.” SLOs provide the language we need to define “well.”
You might be more familiar with Service-Level Agreements (SLAs), so let’s begin there. SLAs are thought of by some as the heart of darkness, and by others as the light of redemption. Why the disparity? I believe it’s because of how they are defined. They can be defined in a way that enables producers and consumers to level-set their expectations, or they can be tool for despair, false assurances, or exposing one to dangerous financial liability. Let’s not spend too much time in the darkness.
Oftentimes today, the term SLO is used, and for the purposes of this text, an SLA is simply an SLO that two or more parties have “agreed” to. SLO and SLA are used fairly interchangeably herein, but we make an effort to consider SLAs as “external” multiparty agreements and SLOs as “internal” single-party goals.
Although the concept of an ...