
What it does and who depends on it
- Who uses it, and what happens to them when it is down?
- Which other services call it, and which does it call?
- What is the one thing it must never get wrong?
Where it runs
- How does a change get to production? Show me one, start to finish.
- What settings and secrets does it need, and where are they kept?
- What runs on a schedule, and what happens if a run is missed?
How it breaks
- What were the last three incidents?
- Which alert do I take seriously, and which one is noise?
- What do you check first when something looks wrong?
- Is there anything that has to be done by hand, every week or every month?
What was never fixed
- Which part of the code are you afraid to touch?
- What would you rewrite if you had a month?
- What does the documentation say that is no longer true?
People answer these honestly in a handover in a way they never do in documentation.
Do one thing together
Before the hour ends, deploy one tiny change with them watching. The first deploy is where the missing permission, the undocumented step and the script on their laptop show up. Better now than at two in the morning.
Your first week alone
- Write up the handover while you remember it, and send it to them to correct.
- Read the entry points: routes, jobs and the schema.
- Find the dashboards and the logs, and look at a normal day so you can recognize a bad one.