DevOps for Beginners
by Sergey Okinchuk
Safe Rollouts and RollbacksYou Shipped It — Now What?

Chapter Nine · Part 3: Keeping It Running

Monitoring & Observability

Chapter 8 got a change safely out the door. This chapter is about what happens next — because the moment Pageturn is live in front of real readers is the moment the running half of the work begins, not the moment it ends.

3 topics

It is easy to picture a deploy as a finish line: the change goes out, and you exhale. But software that is live is software that can break, slow down, or quietly start failing for some of the people using it. Nobody knows whether a release is actually fine until someone looks — and looking, on purpose and all the time, is the work this chapter is about.

Three short topics build the picture. First, why "it's live" is the start of the operations job rather than the end of it. Then the three plain things teams watch — logs, metrics, and alerts — and what each one tells you. And finally what happens when something does go wrong: on-call, incidents, and the blameless way good teams learn from them. Together they close the feedback loop from Chapter 1: watching the live app is what tells you what to change next.

Watching the live app is what feeds the next change — the loop closes here
Deployship the change
→
Watchis it healthy?
→
Learnread the signals
→
Next changeloop turns again

Topics in This Chapter

Topic 27
You Shipped It — Now What?
Going live is where the running half of the work begins. You can't fix what you can't see, so the first job of operations is watching the live app — and "you build it, you run it" means the team that shipped it is the team that watches.
Concept
Topic 28
Logs, Metrics, and Alerts
The three plain things teams watch: logs are the recorded diary of events, metrics are numbers measured over time, and alerts are automatic warnings when a number crosses a line. Together they're the vocabulary of every monitoring tool.
Concept
Topic 29
On-Call and Incidents
When an alert fires at 3am, someone has to respond — that's on-call. A serious production problem is an incident, and the blameless postmortem afterward studies why it happened, not who to blame. Where the "you run it" half meets DevOps culture.
Concept