OpenAPI 3.0 Pagination Extension ↔ RESTful Service Implementation: Latency Under Concurrency

Context

The OpenAPI 3.0 specification includes a pagination extension point that allows service designers to expose link relations such as next and prev. However, the exact naming and parameter conventions are left to the implementation, which can lead to divergent client expectations.

Unresolved Decision

When multiple clients issue paginated requests simultaneously, the specification does not prescribe any normative latency behavior. Server frameworks may employ different strategies for offset‑based versus cursor‑based pagination, potentially causing unpredictable latency spikes that are not reflected in the OpenAPI contract.

Goal

Determine whether the OpenAPI 3.0 pagination extension can be relied upon to guarantee consistent latency characteristics across concurrent requests, and if not, identify what implementation‑level factors need to be specified.

Questions

• Does OpenAPI 3.0 define any expected latency or ordering guarantees for paginated endpoints when accessed concurrently?
• How do implementation choices between offset and cursor pagination affect server response times under high concurrency?
• What documentation or best‑practice recommendations exist for aligning OpenAPI pagination contracts with underlying server framework performance?