Expand description
A trait for collecting and managing performance metrics within the pipeline.
The Metrics trait defines a set of asynchronous methods for initializing,
updating, flushing, and shutting down metrics in the carbon-core
framework. This trait is designed for tracking various types of metrics,
such as counters, gauges, and histograms, to monitor performance and
operational health in real time.
§Key Concepts
- Gauges: Track the value of a particular metric at a specific point in time. Useful for monitoring values that can fluctuate, like the number of queued updates.
- Counters: Track the number of times an event has occurred, such as successful or failed update processing.
- Histograms: Measure the distribution of values, such as processing times, allowing insights into latency or response times.
§Implementing the Trait
To implement Metrics, provide implementations for each method, typically
sending metrics data to a monitoring system or backend service for
visualization and alerting. The trait requires async functions, allowing
implementations to perform non-blocking I/O operations, such as network
requests or database writes.