Map RPO and RTO targets to a PostgreSQL HA topology
Turn 'no data loss, fast failover' into a concrete PostgreSQL topology: which target needs a synchronous standby, which an async region replica, which needs a consensus manager and fencing, and which existing lab measures each claim.
Problem
What you're actually looking at
The symptom as it shows up on a real server.
Availability targets are meaningless until each one maps to a mechanism and a measurement. RPO=0 within a failure domain needs synchronous commit to a standby. A regional budget needs a bounded-lag async replica. An RTO needs automatic failover with fencing, timed to first committed write. Skipping the mapping is how teams ship HA that quietly loses data or never actually fails over.
Meridian Freight has locked targets: RPO zero on single-AZ loss, <=60s across region, RTO <=60s for AZ failover. The database is still a single primary with async replication and manual promotion. This runbook is the assessment: read the current durability posture, decide the mechanism each target requires, and point at the measured lab that proves it.
Simple terms
Each availability promise is really a request for a specific Postgres mechanism. Lose nothing if one machine dies means a synchronous standby a commit waits for. Lose at most 60s if a region dies means an async replica whose lag you bound and watch. Be writable again within a minute means an automatic, fenced failover you have timed. On one Docker host you can prove the failover and lag mechanisms. True multi-AZ/region geography and a synchronous zero-loss promotion stay design-only until a real host measures them.
Full runbook for this incident
- The full identify checklist, the exact signals that tell you it's this incident
- Every diagnostic query; lab output is attached only to the steps we actually captured
- The resolution path and the pitfalls that make it worse
- Mitigation steps to stop it recurring, plus a verify-you're-done query
Card required. Cancel before day 7 and you are not charged.
Related runbooks
Continue the same incident path
Sibling procedures that cover the adjacent setup, recovery, or prevention step.
More in this category
Other Mission-critical platform runbooks
Neighbouring incidents that share the same diagnostic surface.
Connected
How this connects to the rest of the library
A live view of this page's real cross-references, what explains it, what fixes it, what to tune, and where to go next. Every link is an authored relationship, not a guess.
Fixes these errors
Need the full procedure?
Pro runbooks finish the incident path
Free runbooks teach the shape. Pro opens the full step transcript, edge cases, and prevention depth.