Custom Metrics Stackdriver Adapter vs Kube Metrics Adapter

Developers should use this adapter when running Kubernetes workloads on Google Kubernetes Engine (GKE) or other platforms that need to send custom metrics to Stackdriver for centralized monitoring meets developers should use kube metrics adapter when they need to autoscale kubernetes workloads based on custom or external metrics, such as scaling a microservice based on http request latency or a batch job based on queue depth in a message broker. Here's our take.

Custom Metrics Stackdriver Adapter

Nice Pick

Developers should use this adapter when running Kubernetes workloads on Google Kubernetes Engine (GKE) or other platforms that need to send custom metrics to Stackdriver for centralized monitoring

Pros

• It is particularly useful for applications that expose metrics in Prometheus format but require integration with Google Cloud's tools for dashboards, alerts, and analysis
• Related to: google-kubernetes-engine, prometheus

Cons

• Specific tradeoffs depend on your use case

Kube Metrics Adapter

Developers should use Kube Metrics Adapter when they need to autoscale Kubernetes workloads based on custom or external metrics, such as scaling a microservice based on HTTP request latency or a batch job based on queue depth in a message broker

Pros

• It is essential for applications where resource-based scaling (CPU/memory) is insufficient, enabling more responsive and efficient scaling in dynamic environments like e-commerce platforms or real-time data processing systems
• Related to: kubernetes, prometheus

Cons

• Specific tradeoffs depend on your use case

The Verdict

Use Custom Metrics Stackdriver Adapter if: You want it is particularly useful for applications that expose metrics in prometheus format but require integration with google cloud's tools for dashboards, alerts, and analysis and can live with specific tradeoffs depend on your use case.

Use Kube Metrics Adapter if: You prioritize it is essential for applications where resource-based scaling (cpu/memory) is insufficient, enabling more responsive and efficient scaling in dynamic environments like e-commerce platforms or real-time data processing systems over what Custom Metrics Stackdriver Adapter offers.

The Bottom Line

Custom Metrics Stackdriver Adapter wins

Developers should use this adapter when running Kubernetes workloads on Google Kubernetes Engine (GKE) or other platforms that need to send custom metrics to Stackdriver for centralized monitoring