When would you run a database yourself on VMs instead of using the managed service?
What they are really testing: Whether you can reason about operational burden honestly. Both extremes are wrong answers, and the interesting part is what the managed service actually buys and what it takes away.
A real interview question
When would you run a database yourself on VMs instead of using the managed service?
What most people say
drag me
“Self-managing is cheaper, so at scale you should always run your own database.”
It compares instance prices and ignores the expensive part. The cost of self-managing is a person who can restore a corrupted database under pressure at 3am, plus upgrade projects and backup verification, and that routinely exceeds the price difference.
The follow-ups they ask next
How would you verify backups are actually usable?
Restore them on a schedule to a scratch environment and run a data integrity check, automatically. A backup nobody has restored is an assumption, and the middle of an incident is a bad time to discover it.
What is the hidden cost of a managed service during an incident?
You lose some visibility and control. You cannot always see the host, and you may wait on vendor support during a critical failure. That trade-off is usually still worth it, but it should be conscious.
The team wants self-managed for a specific extension. What do you propose?
Check whether the managed service added it or offers an alternative, whether the extension is truly required or a convenience, and whether that one need justifies owning the whole operational burden. Sometimes a small self-managed component alongside a managed primary is the answer.
What the interviewer is listening for
- Frames managed as buying an operational function
- Counts on-call and expertise in the cost
- Names specific legitimate exceptions
- Raises backup verification unprompted
What sinks the answer
- Compares only instance prices
- Self-manages by default for control
- No restore testing story
- Cannot name a legitimate exception either way
If you genuinely do not know
Say this instead of freezing. Reasoning out loud from what you do know beats silence every single time, and a good interviewer is listening for exactly that.
“Default managed, because it buys [backups with point-in-time recovery, patching, failover and someone awake at 3am], which is [a role, not a server]. The real cost of self-managing is [an engineer permanently, plus being able to restore under pressure]. Genuine exceptions are [an unsupported extension or version, kernel-level tuning, regulatory constraints, or scale where the premium is enormous].”
Keep going with cloud fundamentals
Senior
Design a CI/CD system for 15 microservices owned by 4 teams deploying several times a day.
Senior
Design an observability stack for a system where nobody can currently answer why a request was slow.
Senior
Leadership asks you to prove the platform investment is working. What do you measure?
Senior
You inherit 200 Jenkins jobs with no documentation and are asked to migrate to a modern CI system. How do you approach it?
Senior
Four teams want to share one Kubernetes cluster. How do you isolate them, and when would you give them separate clusters instead?
Senior
You are incident commander for a total outage. It is 2am, 6 engineers are online, and nobody knows the cause. What do you do?
Knowing the answer is not the same as recalling it under pressure
Sign in to send the questions you fumble to spaced recall, so they come back right before you would forget them, and learn the concepts behind them with hands-on labs.
Start free