Practical guide
Forecast AlloyDB read-pool cost
Price AlloyDB read scaling by adding read-pool compute, memory, storage, backup, and network assumptions to the primary cluster.
Updated · Sources checked
Add read capacity explicitly
AlloyDB read pools improve read capacity but introduce their own configured resources. Price each pool’s vCPU and memory hours, then add cluster storage, backup, and data-transfer assumptions. Avoid treating read scaling as a zero-cost setting on the primary.
If two read-pool instances each run 4 vCPUs and 32 GB for 400 peak hours, they contribute 3,200 vCPU-hours and 25,600 GB-hours before storage and networking. Compare that cost with measured read demand and the operational objective.
Use the AlloyDB Cost Calculator to total N2 vCPU, memory, and regional-storage quantities across primary and pool scenarios. It excludes backups and network charges; confirm precise prices in Google’s pricing page.
FAQ
Does a read pool remove egress charges?
No. Network egress remains a separate usage component.