Skip to content

Repository files navigation

Trade-offs for Each Approach

App-native Metrics

Pros:

  • Simplicity: Directly integrates with the application, reducing the need for additional components.
  • Performance: Lower latency as metrics are collected and exposed directly by the application.
  • Consistency: Metrics are always in sync with the application's state.

Cons:

  • Coupling: Tightly couples the metrics collection with the application logic, making it harder to maintain and test.
  • Scalability: May require changes to the application code if the metrics collection needs to be updated or scaled.
  • Resource Usage: Increases the resource usage of the application container, potentially affecting its performance.

Aggregator Sidecar

Pros:

  • Decoupling: Separates metrics collection from the application logic, making it easier to maintain and test.
  • Flexibility: Can be updated or scaled independently of the application.
  • Isolation: Reduces the impact on the application's performance by offloading metrics collection to a separate container.

Cons:

  • Complexity: Adds an additional component to the deployment, increasing the overall complexity.
  • Latency: Potentially higher latency as metrics are collected from logs rather than directly from the application.
  • Synchronization: Requires careful handling to ensure that the logs and metrics are in sync.

About

Demonstration of native and external metrics publishing for a python application

Resources

Stars

0 stars

Watchers

1 watching

Forks

Releases

Packages

Used by

Contributors

Languages