FDE Work from Chapter 12
Production engineering instinct is the understanding of a system's operational state: knowing that a system is on and doing work every day.
Deployed Does Not Mean Working
Most software engineers have spent time building something that was supposed to work, then shipping it and hoping it kept working. This section looks at that moment from the other side. Once a system is running for real users, a new set of questions appears: is it on, is it doing its work, and who notices when it stops? The source frames these questions as a discipline called production engineering, and it argues that you do not have to start that discipline from zero.
What do you think happens?
Once a system has been deployed, is it automatically operational?
Reveal answer
No. A system can be deployed and still not be operational in any useful sense.
The source uses the word operational as the key idea. Production engineering instinct is the habit of asking whether the system is actually running and doing its job right now, not merely whether it was built correctly at some point in the past.
Reading a System's Operational State
Production engineering instinct is the understanding of a system's operational state: knowing that a system is on and doing work every day.
The source describes this as a feel for the system's daily state, which is why it is called an instinct rather than a checklist. It is a habit of mind. When you think about a running system, you keep asking whether it is on and doing its work, rather than only asking whether the code was written correctly.
One way to picture the operational state is as a system that moves over time between conditions. The diagram below is an illustrative model written for this article. The source itself does not name degraded or down as states, so treat those labels as a generated way to reason about the idea rather than as terms the source defines.
Three Practices, Three Questions
The source names three practices that come together under production engineering: on-call rotations, alerting, and incident response. Rather than memorizing them as three separate topics, it helps to see each one as an answer to a question that a running system raises. Who is watching? How do we hear about a problem? What do we do once we have heard?
The three practices are parts of one discipline and together they express the operational instinct. The source does not establish a fixed order among them, so do not assume one practice always triggers the next.
Tracing One Problem Through the Questions
A running system starts to slow
An illustrative system is running normally. Its response times begin to climb and some requests fail. Which of the three practice questions apply, and in what sense?
- 1Check the operational state
The system is still on, but it is no longer doing its work fully. In the illustrative model this is the degraded state, and the production engineering instinct asks you to notice it rather than assume the system is fine because it was deployed.
- 2Ask who is watching
The on-call question is whether someone is responsible for the system at this moment. The source places on-call rotations inside production engineering as the answer to who is watching.
- 3Ask how the problem is heard
The alerting question is how the degraded behavior reaches a person. The source names alerting as part of the discipline but does not describe its mechanism, so you should not assume how that signal is produced.
- 4Ask what happens next
The incident response question is what is done once the problem is known. The source names this practice but does not give its steps, so the response itself is outside what this section teaches.
The same running-system problem raises all three questions. The source supports naming them as three parts of one discipline, but it does not establish an order in which they must occur.
Where the Chapter 12 FDE Work Fits
The source makes a specific claim about where earlier work fits. The FDE work from Chapter 12 sits inside the production engineering discipline. That matters because the operational questions in this article are not a new subject bolted onto the course. They are the same discipline that the FDE work already belongs to, so you do not have to learn this from scratch.
The diagram shows the relationship the source describes: the instinct sits at the center and the three practices hang off it as parts of the same discipline. It does not show a sequence of steps. Read it as a map of membership, not a flow of events.
Mistakes When Reading the Operational Picture
Common mistake Treating a successful deployment as proof that the system is operational
What it looks likeA release goes out cleanly, so the team stops checking whether the system is doing its work.
Why it failsThe source separates being built correctly from being operational. Production engineering instinct asks about the system's state right now.
Do this insteadAfter deployment, keep asking whether the system is on and doing its work.
Common mistake Imposing a fixed order on the three practices
What it looks likeAssuming that every incident begins with an alert and always ends with a particular response step.
Why it failsThe source states that the three practices are parts of one discipline and does not establish a fixed order among them.
Do this insteadTreat any sequence you imagine as your own assumption, and check it against the actual system rather than the diagram.
Common mistake Memorizing the three practices as unrelated topics
Why it failsThe source presents them as three answers to the questions a running system raises, all expressing one operational instinct.
Do this insteadLink each practice back to the question it answers: who is watching, how we hear, and what we do next.
Where You Already Have a Head Start
The source says you do not have to start production engineering from zero. Most software engineers have already built something, shipped it, and cared whether it kept working. That experience is the starting point for the operational instinct.
Pick one system you have built or worked on in the FDE work from Chapter 12 or in earlier projects. For that system, write one sentence for each of the three questions: who would be watching it, how you would find out it had stopped doing its work, and what you would do once you knew. Then mark which of the three answers you could already give from your own experience and which you could not.
Hints
- Start with the operational state: is the system on and doing its work right now?
- If you cannot answer one question, note it as an area that is new to you rather than something you have already learned.
- The source does not describe how on-call, alerting, or incident response are implemented, so your answers can be honest about what you do not yet know.
Key Takeaways
- Production engineering instinct is the understanding of a system's operational state: knowing that a system is on and doing work every day.
- Deployment and operation are different questions; the instinct asks whether the system is doing its job right now.
- On-call rotations, alerting, and incident response are three parts of one discipline, answering who is watching, how we hear about a problem, and what we do once we have heard.
- The FDE work from Chapter 12 sits inside the production engineering discipline, so the operational questions extend work you have already begun.