Rate Limiting

To guarantee the reliability, security, and consistent low latencies of our ingestion, the Antarctica.io AI Module employs automated rate-limiting policies protecting our telemetry pipelines.

Limits

External clients currently have base ingestion boundaries:

• Default Limit: 5,000 requests per IP address, per minute.
• Enterprise Limits: Can be elevated depending on plan configuration.

Handling Throttling

When limits are met, the API correctly responds with an HTTP status code 429 Too Many Requests. The response object contains retry headers.

Best Practices:

1. Exponential Backoff: Catch 429 status errors and use an exponential backoff with jitter prior to resubmission.
2. Batch Ingestion: If supported, batch your logs to keep HTTP overhead minimal.
3. Idempotency Keys: Use Idempotency-Key headers on retry loops so that events processed under heavy concurrency remain perfectly deduplicated.