What is the difference between provisioned and on-demand capacity in OCI NoSQL?

Provisioned capacity lets you specify the exact read and write units per second you need, and you pay for that capacity whether used or not. On-demand capacity automatically scales to your actual traffic and charges per request. Provisioned is cheaper for predictable, steady workloads. On-demand is better for variable traffic patterns, development environments, or new applications where you cannot predict throughput requirements. You can switch between modes.

How do read and write capacity units translate to actual operations?

A read unit represents one eventually consistent read of an item up to 1 KB. Strongly consistent reads cost 2 read units. A write unit represents one write of an item up to 1 KB. Larger items consume proportionally more units — a 5 KB item write costs 5 write units. Query operations consume read units based on the data scanned, not just the results returned, so efficient queries with targeted key conditions minimize costs.

OCI NoSQL Cost Estimator

StorageOCI

Estimate NoSQL Database Cloud costs with provisioned and on-demand capacity modes.

Last verified: May 2026

OCI NoSQL Configuration

Capacity Mode

Read Units (per second)

Write Units (per second)

Storage (GB)

Include Free Tier (25 read/write units + 25 GB storage)

Raw Output

Output will appear here...

About This Tool

OCI NoSQL Database Cloud Service provides a fully managed, multi-model (key-value, document, columnar) database with single-digit millisecond latency and automatic scaling. Pricing is based on provisioned read and write capacity units, storage consumed, and optional on-demand capacity mode. This estimator calculates monthly costs based on your throughput requirements, storage volume, and capacity mode selection, helping you choose between provisioned and on-demand pricing models.

Real-World Scenario

Your team's session store handles 50K reads/sec and 10K writes/sec with 500 GB of data. The estimator compares: provisioned at exactly 50K read + 10K write capacity = ~$1,800/month; on-demand at the same workload = ~$5,400/month. Provisioned wins by 67% for this steady-state workload. The team also reviews item sizes — average item is 1.2 KB, suggesting they're paying 2 RU per read. Compressing the session JSON brings average item to 0.7 KB = 1 RU per read, halving the required provisioned capacity. Combined savings: provisioned + size optimization = $900/month vs the $5,400/month naive on-demand approach.

When to Use This Tool

•Estimating monthly costs for a session store that handles 50,000 read units/sec and 10,000 write units/sec with 500 GB of data
•Comparing provisioned capacity pricing versus on-demand pricing for a workload with unpredictable traffic patterns
•Budgeting for a NoSQL table used as a time-series database for IoT telemetry with high write throughput and periodic batch reads
•Calculating the cost of scaling a NoSQL table to handle seasonal traffic spikes (Black Friday, product launches)

Pro Tips

TIP

Provisioned capacity is dramatically cheaper than on-demand at steady state. The crossover is roughly when your throughput is consistently above 60% of provisioned for >12 hours/day. For unpredictable traffic, on-demand eliminates over-provisioning waste; for predictable workloads, provisioned wins on cost.

TIP

Strongly consistent reads cost 2x eventual reads. Most workloads tolerate eventual consistency (a session lookup returning slightly stale data is fine). Default to eventual consistency and only use strong consistency for genuinely consistency-critical operations (financial transactions, leaderboard updates).

TIP

Item size matters dramatically. A 1 KB read = 1 RU. A 5 KB read = 5 RU. Compress JSON values, use shorter field names, normalize related data into separate tables — these item-size optimizations directly reduce RU consumption AND make queries faster.

How It Works Under the Hood

The estimator computes OCI NoSQL monthly cost as: provisioned capacity (read units/sec × hourly rate + write units/sec × hourly rate × hours), OR on-demand capacity (per-request rates × monthly request count), plus storage GB × per-GB-month rate. Output presents both pricing models side-by-side at your stated throughput levels, plus the breakeven analysis showing when each becomes cheaper.

Frequently Asked Questions

What is the difference between provisioned and on-demand capacity in OCI NoSQL?: Provisioned capacity lets you specify the exact read and write units per second you need, and you pay for that capacity whether used or not. On-demand capacity automatically scales to your actual traffic and charges per request. Provisioned is cheaper for predictable, steady workloads. On-demand is better for variable traffic patterns, development environments, or new applications where you cannot predict throughput requirements. You can switch between modes.
How do read and write capacity units translate to actual operations?: A read unit represents one eventually consistent read of an item up to 1 KB. Strongly consistent reads cost 2 read units. A write unit represents one write of an item up to 1 KB. Larger items consume proportionally more units — a 5 KB item write costs 5 write units. Query operations consume read units based on the data scanned, not just the results returned, so efficient queries with targeted key conditions minimize costs.

Related Learning Guides

OCI Cost Optimization Strategies18 min read

Was this tool helpful?

Disclaimer: This tool runs entirely in your browser. No data is sent to our servers. Always verify outputs before using them in production. AWS, Azure, and GCP are trademarks of their respective owners.