Documentdb Elastic Cluster Pricing

Understanding the pricing model is critical because it shifts from "paying for the server" to "paying for shards and compute power." The Core Pricing Components

| Feature | Elastic Cluster Pricing | Instance-Based Pricing | | :--- | :--- | :--- | | | ACU (Auto-scaling) | vCPU/RAM (Fixed Instance Size) | | Idle Cost | Near Zero (Can scale to 0) | High (You pay for the instance 24/7) | | Peak Performance | Instant scale-up to max capacity | Limited to the instance size provisioned | | Commitment | On-Demand only | Available (Reserved Instances provide discounts) | | Best For | Unpredictable workloads, dev/test, intermittent jobs | Steady, predictable, high-throughput workloads | documentdb elastic cluster pricing

You pay for the total number of vCPUs provisioned across your shards. Understanding the pricing model is critical because it

Region: US East (Ohio) Cluster type: Elastic Number of shards: 3 DCUs per shard (min): 2 DCUs per shard (max): 12 Average DCU usage: 5 Storage: 750 GiB Backup retention: 14 days (assume 2x storage) Data transfer out: 100 GB/month intermittent jobs | Steady