Coverage Matrix

Chkk Curated Release Notesv0.11.0 to latest
Private RegistrySupported
Custom Built ImagesSupported
Safety, Health, and Readiness Checksv0.13.3 to latest
Supported PackagesHelm, Kustomize, Kube
EOL InformationAvailable
Version Incompatibility InformationAvailable
Upgrade TemplatesIn-Place, Blue-Green
PreverificationAvailable

Prometheus CloudWatch Exporter Overview

Prometheus CloudWatch Exporter integrates AWS CloudWatch metrics into Prometheus monitoring by converting AWS metrics into Prometheus-compatible format. It fetches metrics from AWS CloudWatch via AWS APIs, exposing them through a /metrics endpoint for Prometheus scrapes. This enables unified observability of Kubernetes workloads and AWS infrastructure metrics within the Prometheus ecosystem. The exporter supports querying diverse AWS services, allowing correlation of infrastructure and application metrics through PromQL queries and Grafana dashboards.

Chkk Coverage

Curated Release Notes

Chkk curates the CloudWatch Exporter’s official release notes, emphasizing impactful changes like AWS SDK upgrades, API optimizations, or new configuration fields. Rather than manually reviewing lengthy release notes, engineers receive tailored summaries relevant to their deployments. For instance, if a new exporter version introduces a configuration requiring additional IAM permissions, Chkk proactively flags this. Engineers stay informed about breaking changes or critical fixes, minimizing surprises during upgrades.

Preflight & Postflight Checks

Chkk’s preflight checks validate compatibility of existing configurations, IAM permissions, Kubernetes deployments, and container images before upgrading the CloudWatch Exporter. These checks catch deprecated configurations or new IAM requirements ahead of time. Postflight checks confirm successful deployment by verifying exporter health metrics, logs, and endpoint availability. Early detection of runtime errors or AWS throttling prevents metric collection disruptions.

Version Recommendations

Chkk continuously evaluates CloudWatch Exporter releases for stability and critical fixes, recommending upgrades when operational risk increases. Unlike projects with defined end-of-life schedules, Chkk highlights specific exporter versions known for issues like inefficient API usage or outdated AWS SDKs. Recommendations include details on community-reported stability, critical bug fixes, or support for newer AWS metrics. Engineers thus balance improvements against operational stability effectively.

Upgrade Templates

Chkk provides detailed Upgrade Templates covering both standard in-place upgrades and blue-green deployments. Templates guide engineers through version updates, configuration adjustments, and verification steps, clearly defining rollback strategies if needed. Canary upgrades allow side-by-side exporter deployments, minimizing disruption by validating metric consistency before full adoption. Templates integrate seamlessly into GitOps or CI/CD workflows, simplifying complex upgrades.

Preverification

Chkk’s preverification approach simulates CloudWatch Exporter upgrades in isolated environments using real-world configurations and credentials. This dry-run identifies configuration parsing issues, AWS authentication errors, or unexpected resource consumption before live deployment. Issues such as increased API calls, metric renaming, or higher memory usage can be addressed proactively. Preverification helps engineers anticipate and prevent upgrade-related incidents.

Supported Packages

Chkk supports multiple deployment methods for CloudWatch Exporter, including Helm charts, plain Kubernetes manifests, and Kustomize configurations. It recognizes custom builds, private images, and specialized configurations without requiring workflow changes. Chkk precisely identifies necessary adjustments like image tags or configuration updates for each upgrade method. This flexibility ensures seamless integration into existing GitOps or automated deployment practices.

Common Operational Considerations

  • Exact Metric Dimensions: Always specify required AWS metric dimensions exactly (e.g., AvailabilityZone for ELB metrics). Missing dimensions result in silently failed queries with no data returned.
  • AWS Credentials & Permissions: Exporter deployments require IAM permissions cloudwatch:ListMetrics, cloudwatch:GetMetricStatistics, and cloudwatch:GetMetricData (plus tag:GetResources when using aws_tag_select); missing or insufficient permissions will cause scraping errors despite the exporter process running normally.
  • CloudWatch Data Lag: CloudWatch metrics have inherent delays (default exporter delay is 10 minutes). Avoid aggressive scrape intervals to prevent missing or rejected metrics in Prometheus.
  • Over-collection & API Costs: Scraping excessive metrics can rapidly increase AWS API costs. Selectively query critical resources or dimensions to manage API usage and cost-effectiveness.
  • AWS API Rate Limits: Excessive API calls may trigger CloudWatch throttling, impacting metric collection reliability. Distribute metric scraping across multiple exporter instances or adjust scrape intervals to mitigate rate limiting.
  • Resource Usage & Scaling: The JVM-based exporter can consume significant memory and CPU with extensive metric collection. Appropriately size resource requests/limits and consider splitting large workloads across multiple instances.

Additional Resources