The "Code Hotspots" tab in the APM trace panel provides a direct link to data from the Datadog Continuous Profiler that corresponds to the trace's endpoint. The profiler captures granular, code-level performance data, including CPU usage, wall time, and contention metrics. "Lock wait time," which is the time a thread spends waiting to acquire a lock on a synchronized object, is a key contention metric captured by the profiler and displayed under this tab.

A. The "Infrastructure" tab correlates the trace with host-level metrics like CPU and memory usage, not code-level lock contention.

B. The "Network" tab displays metrics related to network communication between services, such as latency and traffic volume, not internal code execution.

C. The "Processes" tab shows information about the running processes on the host at the time of the trace, not specific code-level metrics like lock wait times.

1. Datadog Documentation, "Continuous Profiler > Code Hotspots". This section explicitly states, "The Code Hotspots tab in the trace view shows you the top 5 most time-consuming methods from the corresponding profile." It lists "Lock wait" as one of the available profile types.

2. Datadog Documentation, "APM Trace View". This page describes the various tabs available in the trace panel, clarifying that Code Hotspots connects the trace to profiling data for deep, code-level analysis.