A performance engineer needs to find all HTTP calls to the /api/checkout endpoint that took longer than 2 seconds AND returned a 5xx HTTP status code across all services in the last 6 hours. Which Instana capability is best suited for this investigation?

Unbounded Analytics in Instana's Calls view is designed for exactly this kind of ad-hoc, multi-condition, trace-level exploration—combining arbitrary filters on endpoint path, duration threshold, and HTTP status code without pre-defined aggregations. The Service Map shows high-level service health and does not support granular call-level filtering with combined duration and status conditions. Custom dashboard widgets show aggregated metrics rather than individual call records matching multi-condition filters. The correct answer is 'Unbounded Analytics in the Calls view with filters for endpoint, duration greater than 2s, and HTTP 5xx status'. The incorrect options — 'The Service Map filtered by error rate for the application perspective', 'The Infrastructure view filtered by service name and sorted by latency', 'A custom dashboard widget showing P99 latency percentiles for the checkout endpoint' — are wrong because they do not align with IBM Instana's architecture or recommended practices for this scenario. Understanding this concept is essential for the Domain 1: Operations domain of the IBM Instana Observability certification.