All posts
Data Platform

Database Failover Downtime on AWS, Measured: RDS Multi-AZ 13–15 Seconds, Aurora Under 2

By

November 9, 2026

7 min read

We wrote to Postgres every 50 ms and forced failovers. RDS Multi-AZ lost writes for 13.0 and 14.8 seconds; in one trial the switch began three minutes after the reboot was requested. Aurora PostgreSQL lost writes for 2.0 and 0.5 seconds, one failed write each time. The full timelines, the errors your application will see, and what to set so it recovers on its own.

Benchmark
Postgres Performance
Failover
Aurora
High Availability
Trial 1132Trial 214.80.5

RDS for PostgreSQL, Multi-AZ instance

Aurora PostgreSQL, writer and one reader

Seconds during which writes failed after a forced failover, two trials each. Our measurements, November 2026. A write every 50 ms through a client that reconnects on error.

The data
RDS for PostgreSQL, Multi-AZ instanceAurora PostgreSQL, writer and one reader
Trial 1132
Trial 214.80.5

If your application can't tolerate 15 seconds of failed writes during a failover, that is a reason to choose Aurora that never shows up in a price comparison. The full report has both timelines, the exact errors, the three-minute delay we saw, and the client settings that decide whether your app recovers by itself.

Planning for database failover?

We'll test your application's behaviour during one with you, before a real one tests it for you. 30 minutes, free.

hoolco Monthly Review

Once a month: what changed in tech that month and what it means for you, from cloud and AI to security and the tools you build with. Every item links to its source. No sales pitch, unsubscribe in one click.

As of November 9, 2026. Provided as is, for general information, and accurate to the best of our knowledge on the date shown. Not legal, compliance, tax, financial or other professional advice. Prices, limits and vendor services change; check current vendor documentation before relying on anything here. Terms of use.