Reliability issue

Fragile external call in n8n

Why API calls without retry logic make your workflows unreliable

What is this issue?

A fragile external call is an HTTP request or API call that has neither 'Retry on Fail' nor 'Continue on Fail' enabled. When the external service is temporarily unavailable, your entire workflow fails instead of gracefully handling the error.

Common scenarios:

  • HTTP requests to third-party APIs without retry
  • Database connections without error handling
  • Webhook calls to external services
  • API integrations during rate limiting or maintenance

Why is this dangerous?

Cascading failures

A single failed API call can abort an entire complex workflow, wasting all previous processing.

Transient error sensitivity

Network hiccups, temporary 503 errors, or rate limits cause permanent failures when they're often recoverable.

Data loss

Partially processed data may be lost when the workflow aborts mid-execution.

Manual intervention required

Every failure requires manual restart, increasing operational burden.

How to fix it

  1. 1

    Enable Retry on Fail

    In node settings, enable 'Retry on Fail' with appropriate max retries (2-3) and wait time between retries.

  2. 2

    Use Continue on Fail

    For non-critical calls, enable 'Continue on Fail' to allow the workflow to proceed even if this node fails.

  3. 3

    Add error branching

    Use the $node.error property in downstream nodes to handle failed calls gracefully with alternative logic.

  4. 4

    Implement circuit breaker

    For frequently failing services, consider adding logic to skip calls when the service is known to be down.

Scan your workflow now

Upload your n8n workflow JSON and instantly detect fragile API calls and other reliability issues.

Scan for reliability issues

Related resources

Related reliability issues