Alert on Anthos/GKE request latency exceeding a threshold for a defined duration per SRE practice.

The correct approach is to install Anthos Service Mesh (ASM) because it provides the necessary telemetry for request latency. ASM integrates with Cloud Monitoring, allowing you to define Service Level Objectives (SLOs) directly in the Cloud Console based on service mesh metrics. Once an SLO is defined, you can easily create an alerting policy that triggers when the SLO is violated, aligning with SRE best practices for reliability. Enabling Cloud Trace provides detailed trace data but isn't ideal for real-time alerting on aggregate latency metrics across a service. Cloud Profiler is for optimizing code performance, not for continuous monitoring and alerting on service-level latency. Anthos Config Management is for declarative configuration of clusters, not for defining and monitoring SLOs directly.