Object storage

Object storage cost planning: model capacity, transfer and payment overhead

Object-storage comparisons work best when they model the workload at realistic usage levels rather than relying on a headline storage rate. Compare billable capacity, transfer, retrieval, request charges and retention rules using the same assumptions for every provider.

Convert provider units carefully

Providers may quote GiB per hour while product plans use decimal TB, and monthly examples often assume 730 hours. Normalize the units before comparing plans so the numbers describe the same amount of storage and time.

A one-terabyte decimal quota is about 931.3 GiB. A one-tebibyte quota is 1,024 GiB; they are not interchangeable.

Model realistic high usage

A fixed-capacity plan should be compared at both typical and near-full usage so the effect of storage, transfer, retrieval and request charges is visible before purchase.

Use the same workload assumptions for every service being compared and record any fees that change once usage increases.

Check transfer, requests and retrieval

Standard classes may include public transfer while archive classes can add retrieval charges or minimum retention periods. Request pricing can also matter for applications with tiny objects and very high operation counts.

Document which fees are currently included and put a review date beside the assumption. Provider pricing is not a permanent constant.

Recheck the estimate against actual usage

After a trial or first billing period, compare the estimate with the service usage report and invoice. Unexpected transfer, retrieval or request charges are a reason to revise the workload assumptions.

Repeat the comparison when retention, access patterns or provider pricing changes materially.

Related DotMoose serviceExplore DotMoose Object Storage

Keep reading

Related guides.

More object storage →