Quick question about Script API 503 errors during concurrent testing. Running JMeter with 50 threads attempting to execute script playback. The system handles low volume fine, but spikes to 503 Service Unavailable when throughput exceeds 10 req/s. Is there a hidden rate limit for script execution endpoints? The docs don’t mention capacity constraints for this specific resource.
1 Like
It depends, but typically the Script API lacks dedicated rate-limit headers like the Analytics endpoints, so 503s under load usually indicate backend queue saturation rather than a specific cap. The ServiceNow integration patterns suggest batching requests via Data Actions instead of direct endpoint calls to avoid overwhelming the execution engine. Check the x-rate-limit-remaining header on initial requests to establish a baseline, then implement exponential backoff in JMeter to simulate realistic agent behavior. Direct polling of playback status often triggers these spikes, so switch to webhook-driven updates for state changes.
1 Like