Layer 3 · Go deeper · 7 min
Go deeper
Size it, build it, then make the calls an interviewer will push you on.
Back of the envelope
Size it up
Tap through each step of the arithmetic.
With backups only, how long until the order service is back?
How busy can each of three zones run and survive losing one?
Build-up · 6 steps
Watch the clock after a region dies
The order service on a warm standby, writing 50 orders a second, with an async replica in another region and a 60-minute RTO.
Step 1 of 6
A replica trails the primary
Every order lands on the primary first and reaches the replica in another region about 1 s later. The newest copy that survives the disaster sets your real RPO.
4 interview moments
You’re the candidate
The interviewer pushes on the hard parts. Pick an approach and see how it holds up.
Deep dive 1 of 4
Interviewer
“The business wants an RPO of one minute and an RTO of one hour for orders. Which DR strategy do you build?”
Your move. Pick an approach:
Deep dive 2 of 4
Interviewer
“At 10:42 a bad DELETE runs on the orders database. Your cross-region replica is 1 s behind. How do you recover?”
Your move. Pick an approach:
Deep dive 3 of 4
Interviewer
“Health checks say your primary region is impaired. Should failover to the DR region start automatically?”
Your move. Pick an approach:
Deep dive 4 of 4
Interviewer
“Attackers with stolen credentials hunt for reachable backups to delete or encrypt. Where do you keep yours?”
Your move. Pick an approach: