> ## Documentation Index
> Fetch the complete documentation index at: https://docs.incident.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Alert ingestion rate limits

> How much alert traffic each source accepts, and how to give a busy source more headroom.

Alert ingestion has rate limits applied, using a leaky-bucket algorithm. By default, each alert source accepts 120 events per minute. If you need higher rate limits, consider [splitting its allowance](#split-a-sources-allowance) based on a property of the incoming events, or discuss with your account manager.

These limits are subject to change unless otherwise contracted.

## Handling a 429

Going over returns `429 Too Many Requests`, with the limit, what you have left, and when to retry in the [error body](/api-reference/introduction#errors). Retry from the sending tool: we drop the rejected event rather than queueing it, so no alert is created or updated.

Every event draws on the allowance, including resolve events, deduplicated events, and events your [source filters](/alerts/alert-sources#filter-incoming-alerts) drop later.

The alert source's request log lists rate-limited requests.

## Split a source's allowance

Shard a source so that one noisy team or priority can't consume its whole rate limit. Each distinct value at the path you choose gets its own allowance.

**To split a source's allowance:**

1. Open the alert source and expand **Advanced settings**
2. Under **Rate limit sharding**, choose **Split the limit by a value in the payload**
3. Enter a path into the payload, read like a deduplication key path (e.g., `$.metadata.team` or `$.query_params.region`)
4. Click **Save**

Point the path at a single value. A path landing on an object or an array does not split the rate limit.

**Rate limit sharding** is supported for most alert sources. It is currently not supported for ticketing sources (Jira, GitHub, ServiceNow, Zendesk), PagerDuty, Opsgenie, or Sentry Issues.

## FAQs

<AccordionGroup>
  <Accordion title="Should I split a high-volume tool across several alert sources?">
    No. Rate limit sharding does the same job without duplicating a source's template, attributes, and routing. Shard
    the one source on a value that separates your traffic, such as team or region.
  </Accordion>

  <Accordion title="Are these the same as the API rate limits?">
    No. These limits cover alert events arriving at an alert source. Requests you make with an API key are limited
    separately, and those limits are in the [API reference](/api-reference/introduction#rate-limits).
  </Accordion>
</AccordionGroup>