Skip to content
Start Assessment

OPS04 - How do you implement observability in your workload?

Best Practices

OPS04-BP01 BP01 - Identify key performance indicators OPS04-BP02 BP02 - Implement application telemetry OPS04-BP03 BP03 - Implement user experience telemetry OPS04-BP04 BP04 - Implement dependency telemetry OPS04-BP05 BP05 - Implement distributed tracing

Best Practices

This question includes the following best practices:

Key Concepts

Strategy and Governance

Telemetry strategy: Use this concept to guide architecture and operating decisions for this question area. Define measurable targets, assign clear ownership, and review results regularly against expected business outcomes.

Service-level indicators: Use this concept to guide architecture and operating decisions for this question area. Define measurable targets, assign clear ownership, and review results regularly against expected business outcomes.

Distributed tracing: Use this concept to guide architecture and operating decisions for this question area. Define measurable targets, assign clear ownership, and review results regularly against expected business outcomes.

Operational Execution

Actionable alerting: Use this concept to guide architecture and operating decisions for this question area. Define measurable targets, assign clear ownership, and review results regularly against expected business outcomes.

Operational dashboards: Use this concept to guide architecture and operating decisions for this question area. Define measurable targets, assign clear ownership, and review results regularly against expected business outcomes.

Incident diagnostics: Use this concept to guide architecture and operating decisions for this question area. Define measurable targets, assign clear ownership, and review results regularly against expected business outcomes.

Implementation Approach

1. Define observability objectives

  • Map business outcomes to SLIs and SLOs
  • Identify golden signals per critical component
  • Define logging, metrics, and tracing standards
  • Set alert priorities based on customer impact

2. Instrument workload components

  • Instrument applications for structured logging
  • Capture custom metrics at service boundaries
  • Enable end-to-end distributed tracing
  • Collect dependency health and latency data

3. Operationalize insights

  • Build role-specific dashboards for operators and product teams
  • Tune alarms to reduce noise and alert fatigue
  • Link alarms to runbooks and remediation workflows
  • Integrate telemetry with incident management channels

4. Continuously improve coverage

  • Review observability gaps after incidents
  • Add instrumentation for new services during delivery
  • Retire low-value metrics and alerts
  • Validate observability during resilience testing

AWS Services to Consider

Amazon CloudWatch

Collects metrics, logs, alarms, and dashboards so teams can detect issues early and track operational outcomes.

AWS X-Ray

Traces distributed requests to identify latency bottlenecks and dependency failures across microservices.

Amazon OpenSearch Service

Provides managed search and analytics engines for near real-time insights.

Amazon EventBridge

Routes events between services and triggers automated responses for operational events.

AWS Systems Manager

Provides operational automation, inventory, and runbooks to reduce manual effort and improve day-2 operations.

Common Challenges and Solutions

Challenge: High alert noise

Solution: Use SLO-based thresholds, composite alarms, and ownership-based routing to reduce non-actionable alerts.

Challenge: Missing context in logs

Solution: Adopt structured logging standards including correlation IDs and business transaction identifiers.

Challenge: Blind spots across dependencies

Solution: Trace requests across service boundaries and include third-party integration telemetry in dashboards.

Related Resources