Rate limit

Rate limits control how many times a key can call the KMS API in a time period — protecting your workspace from runaway loops, cost spikes, and one application consuming all the quota.

Window	What it controls?
RPM — Requests per Minute	Short bursts. E.g., max 60 calls in any 60-second window
RPH — Requests per Hour	Sustained load. E.g., max 1,000 calls per hour
RPD — Requests per Day	Daily ceiling. E.g., max 10,000 calls per day

Rate Limit Inheritance

Global Default Policy <- set by Org Admin (all workspaces)

↓

Workspace Global Setting <- set by Org Admin (per workspace)

↓

Per-key Rate Limit <- set by you (Workspace Admin)

• No Workspace Global Setting: keys follow the Global Default Policy

• Workspace Global Setting exists: keys are bound to that Workspace Global Setting

• Keys with a manual limit are not affected when Org Admin changes the global policy

In short: your Org Admin sets the ceiling. You work within it.

When a key gets throttled

Your application receives a rate limit error and should retry after a short wait. In Oten KMS, the key status changes to Throttling and the event is recorded in the History Log. Check which window (RPM/RPH/RPD) is exceeded, wait for the reset, or contact your Org Admin for a higher ceiling.