Skip to main content

​
Overview

While building integration logic, you may encounter rate limit errors from upstream APIs if you consume requests faster than the integration provider supports. Rate limits can be enforced by integration providers for either the requesting user or the requesting application (or both). They can be applied with different schedules (requests per minute, per hour, or per day). Paragon provides automatic protection from rate limit errors in the form of Smart Rate Limits, adapting to known rate limits from supported providers or rate limiting errors encountered during Workflow executions. Example: HubSpot rate limit
429 Too Many Requests

​
Usage

All Workflows for supported integrations will automatically use Smart Rate Limits. No action is required to enable Smart Rate Limits on your account or Workflows.

​
How it works

Smart Rate Limits automatically limit the pressure of all requests sent to an API by your Workflows, backing off and retrying when a rate limit is exceeded. When a Smart Rate Limit is enforced, your workflow will be paused until it is safe to proceed with additional requests.
You can see the status of your paused workflow on the Monitoring > Workflows page, along with the scheduled resumption time. When the workflow resumes, it will retry the last request that was suspended or failed.

​
Supported integrations

The following integrations have Smart Rate Limits built-in and will respect known rate limits for their APIs. We are actively expanding support for Smart Rate Limits across our complete integrations catalog.
  • Confluence
  • Gmail
  • Google Drive
  • Google Sheets
  • HubSpot (currently not available for Search Records or Get All Records actions)
  • Jira
  • NetSuite
  • Notion
  • OneDrive
  • Outlook
  • Salesforce
  • Zendesk

​
Custom Integrations and API Resources

Smart Rate Limits also apply to requests made through Custom Integrations and API Resources. By default, these requests are throttled reactively: when the upstream API returns a 429 Too Many Requests response, Paragon pauses subsequent requests using the response’s Retry-After or x-retry-after header (in seconds), then resumes automatically. For API Resources, you can also define a proactive request budget in the Resource’s Rate Limits step. When set, the Workflow Engine paces requests to stay under the published limit before the API returns a 429.
  • Request limit — the number of requests allowed per second, minute, hour, or day.
  • Limit scope — whether the budget is counted per Connected User or shared across every Workflow request to the Resource.