Service A (Java) calls Service B (Node.js) via HTTP. In Instana, separate independent traces appear for each service but are never linked into a single distributed trace. Host metrics for both services appear normally. What is the most likely cause of the broken trace correlation?

Distributed trace correlation requires an active Instana agent on every participating service so that trace context headers (X-INSTANA-T and X-INSTANA-S) are both injected by the caller and consumed by the callee. If Service B lacks an instrumented Instana agent, its spans are recorded independently with no parent reference, resulting in separate unlinked traces. Confirming agent presence and successful instrumentation on Service B is the correct first diagnostic step. The correct answer is 'The Instana agent on Service B is not installed or not instrumenting the Node.js process, so downstream spans cannot be correlated back to the originating trace'. The incorrect options — 'The two services are deployed in different Instana monitoring zones, preventing cross-zone trace stitching', 'Service A is running an outdated Instana Java sensor that does not emit downstream trace context headers', 'The Instana backend trace correlation processor has a queue backlog causing multi-minute stitching delays' — 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 6: Troubleshooting domain of the IBM Instana Observability certification.