Episode 7. HA DR and sev handling. Day 7 of your 10 day sprint.

Welcome. Technical depth gets you screened in. Sev handling gets you hired, because banks feel outages in customers, not in logs.

Say this first: Customer impact first, root cause second, stay-fixed third.

Concepts. RTO is how long recovery may take; RPO is how much data you may lose. Runbooks name owners and timeboxes. Escalation moves up on impact, not on elapsed time alone. Comms state impact, mitigation, next update time, never a cause guess.

Scenario. Payments writes stall during peak and the bridge is forming. Understand: affected flows and customer-visible errors. Prepare: RTO clock, mitigation options, last good state. Identify: fail over the service versus throttle the bad actor; pick the fastest customer relief. Support: execute, confirm errors stop, then root-cause with a stay-fixed action and owner.

Say this bridge once. Say: On sev calls I narrate impact before cause, and I close every sev with an automation or monitor so the class cannot recur.

Your Day 7 lab: run a five minute sev tabletop aloud, impact, mitigation, prevention, and time yourself under three minutes.
Teaser: Tomorrow, Episode 8, the AWS RDS gap. The honest gap line that earns trust.

Second scenario. A DR test reveals a four hour RTO against a one hour promise. Understand: which recovery step dominates the clock. Prepare: restore timings per step plus parallelism options. Identify: serial restore of archivelogs as the bottleneck; parallelize or pre-stage first. Support: re-run the test, publish the new RTO, schedule the test as recurring evidence.

Pause and say the sev sentence again: impact first, cause second, stay-fixed third. Twice.

Close with a STAR seed: a sev you owned with an RTO number and a customer outcome. Done.