Skip to main content
Bulkgrid run-based workflows are asynchronous. Your client should poll run status until the run reaches a terminal state.

​
Endpoint

curl "$BULKGRID_BASE_URL/api/v1/runs/$RUN_ID" \
  -H "x-api-key: $BULKGRID_API_KEY"

​
Terminal states

Stop polling when status is one of:
  • completed
  • failed
  • cancelled
A reasonable default:
  • first minute: every 2 to 5 seconds
  • larger crawl jobs: back off to 5 to 15 seconds
  • always apply a client-side maximum wait time

​
What to inspect while polling

  • status
  • queued
  • in_progress
  • done
  • failed
  • last_error
  • updated_at

​
What to do next

  • if completed, fetch results
  • if failed, inspect error fields and decide whether to retry
  • if cancelled, stop work and clean up any dependent jobs