Site reliability engineering is the practice of treating operations as a software problem and reliability as a measurable target with a cost. Both halves matter: the automation without the target produces busywork, and the target without automation produces heroics.
One hundred percent is the wrong target
Perfect availability is unattainable and pursuing it is ruinously expensive. Beyond a point users cannot distinguish improvements, because their own network and devices fail more often than your service does. The correct target is the point where further reliability stops being noticed.
Reliability competes with feature velocity
Every engineering hour spent on reliability is an hour not spent on features. Making that trade explicit and numeric, rather than an argument between teams, is the central contribution of the discipline.
Operations work should shrink
If a service’s operational burden grows in proportion to its traffic, it does not scale. The measure of the function is whether load can grow substantially without a matching increase in human effort.
Action step
Estimate the hours your team spent on operational work last month and what proportion of the total that was.
Tech Skills Library is independent and is not affiliated with, accredited by, or endorsed by any cloud provider, tooling vendor or certification body named in this course. Product features change frequently; always confirm current details in the official documentation. This course teaches engineering practice and does not issue a certification or credential.