AWS RDS Multi-AZ cost: you pay for the standby you hope never runs

Quick answer

RDS Multi-AZ runs a standby replica in another Availability Zone for automatic failover, which roughly doubles the instance cost and mirrors storage. The newer Multi-AZ cluster deployment runs two readable standbys (three instances total) for faster failover and read scaling, at higher cost. Use Multi-AZ for production databases where downtime is costly, single-AZ for dev and staging.

A single-AZ RDS instance is one database in one Availability Zone. Multi-AZ adds redundancy for automatic failover, and you pay for the redundancy whether or not it is ever exercised. The question is which environments genuinely need it.

Multi-AZ instance: pay for the standby

The classic Multi-AZ deployment runs a standby replica in a second AZ, kept in sync and ready to take over. You pay for both the primary and the standby, so compute cost roughly doubles, and storage is mirrored. The standby is not readable; it exists purely for failover, so you are buying availability, not capacity.

Multi-AZ cluster: two readable standbys

Deployment	Instances	Trade
Single-AZ	1	Cheapest, no automatic failover
Multi-AZ instance	2 (1 standby, not readable)	~2x cost, automatic failover
Multi-AZ cluster	3 (2 readable standbys)	Higher cost, faster failover + reads

The Multi-AZ cluster deployment runs three instances with two readable standbys, giving faster failover and some read-scaling, at correspondingly higher cost. It suits databases that need both high availability and read capacity.

When Multi-AZ is worth it

Multi-AZ is insurance: it is worth roughly doubling cost when database downtime is expensive, which is true for production systems of record. It is wasteful for dev, staging, and non-critical workloads, where single-AZ and good backups are enough. Matching the deployment to the environment, rather than defaulting everything to Multi-AZ, is the main saving, the same discipline as choosing the right RDS shape.

FAQ

How much more does RDS Multi-AZ cost?

Roughly double the single-AZ instance cost, because it runs a standby replica in another Availability Zone that you pay for alongside the primary, and storage is mirrored. The Multi-AZ cluster deployment (three instances) costs more still.

Is the RDS Multi-AZ standby readable?

In the classic Multi-AZ instance deployment, no, the standby exists only for failover and cannot serve reads. The newer Multi-AZ cluster deployment runs two readable standbys, giving read-scaling in addition to faster failover, at higher cost.

When should I use RDS Multi-AZ?

For production databases where downtime is costly, since it provides automatic failover. Use single-AZ for dev, staging, and non-critical workloads where the roughly doubled cost is not justified and good backups suffice.

What is the difference between Multi-AZ instance and cluster?

Multi-AZ instance runs one non-readable standby for failover, roughly doubling cost. Multi-AZ cluster runs two readable standbys (three instances total) for faster failover and read-scaling, at higher cost. Choose cluster when you need both HA and read capacity.

How do I reduce RDS Multi-AZ cost?

Use Multi-AZ only where downtime is genuinely costly, keep dev and staging single-AZ, right-size the instance so you are not doubling an oversized one, and use Reserved Instances or Savings Plans on the steady production database to offset the standby cost.

Does C3X estimate RDS Multi-AZ cost?

C3X reads the multi_az and deployment configuration of your aws_db_instance from Terraform and prices the standby accordingly, so the high-availability cost appears before deploy.

What to do next

Price high availability before you enable it. C3X reads your Terraform and prices RDS and the rest of your stack against a live catalog. Start with the quickstart.