Your enterprise operates a distributed microservices architecture deployed across Compute Engine virtual machines, Google Kubernetes Engine (GKE) clusters, and Cloud Run services located in multiple Google Cloud projects. You need to implement a centralized observability strategy in Google Cloud Monitoring to collect and aggregate metrics, logs, and distributed traces from both infrastructure and applications to track Service-Level Objectives (SLOs) and application key performance indicators (KPIs) through unified dashboards. Which strategy aligns with Google Cloud recommended best practices?

Instrument services with OpenTelemetry to export OTLP metrics and traces; deploy the Ops Agent on Compute Engine with OTLP/Prometheus receivers, enable Google Cloud Managed Service for Prometheus on GKE, configure a central scoping project with a multi-project metrics scope, and build custom Cloud Monitoring dashboards combining metric widgets, log-based metrics, and trace data.: What the Recommended Observability Strategy Is This strategy establishes an end-to-end telemetry pipeline across hybrid and polyglot workloads using Google Cloud's modern observability architecture. It leverages open standards and fully managed services to capture, aggregate, and visualize metrics, logs, and traces without introducing management overhead or vendor lock-in. How It Meets the Requirements Application and Infrastructure Telemetry Collection: Instrumenting code with OpenTelemetry allows services to emit standardized OpenTelemetry Protocol (OTLP) metrics and distributed traces. The Google Cloud Ops Agent on Compute Engine VMs natively ingests system telemetry, application logs, and OTLP/Prometheus endpoints via built-in receivers. GKE Workload Integration: Enabling Google Cloud Managed Service for Prometheus provides fully managed, scalable scraping of Prometheus metrics and pod telemetry without operating standalone Prometheus servers. Cross-Project Aggregation: Setting up a multi-project metrics scope allows a central scoping project to query, chart, and monitor time-series data originating across multiple development, staging, and production projects. SLO and KPI Dashboards: Custom Cloud Monitoring dashboards correlate golden signals (latency, traffic, errors, saturation), custom application KPIs, log-based metrics (such as specialized error counts), and distributed trace latencies into unified operational views. Key Advantages Vendor-Neutral Telemetry: OpenTelemetry decouples application code from vendor APIs, simplifying migrations and future expansions. Unified Multi-Project Visibility: A centralized metrics scope simplifies global KPI tracking and service-level management across complex organizations. Optimized Data Pipeline: In-agent buffering, managed scrapers, and native metric ingestion minimize resource overhead and streamline maintenance. Comprehensive Dashboards: Integrates logs, error groups, alerts, and time-series metrics side-by-side to accelerate root-cause analysis. Why This Is Optimal This solution relies completely on Google Cloud's native, managed observability tooling and CNCF standards. It removes operational toil associated with managing custom ingestion proxies or manual batch upload scripts while delivering real-time metric evaluation for critical SLOs.