Metrics are quantitative measures that provide valuable insights into system performance and health. They can originate from various sources:

• Resource metrics: Provide information about the utilization and performance of system resources such as CPU, memory, and network.
• Infrastructure/cloud metrics: These metrics give insights into cloud and hardware infrastructures like AWS, GCP, and Azure. They monitor aspects like disk usage, instance performance, or cloud service consumption.
• Service/transaction metrics: Focus on the performance of specific services or transactions within an application, capturing data like response time, throughput, and error rates.
• Custom metrics: User-defined metrics tailored to capture unique aspects of their application or infrastructure not covered by standard metrics.

Metric Sources

There are many different supported metric sources for Cloud Observability, and we recommend going to our docs site for up-to-date integration information. Some popular sources include:

• OpenTelemetry Collector: With the Collector and OpenTelemetry Protocol, metrics from diverse providers can be ingested. This includes metrics from sources like Prometheus, StatsD, Kafka, and Java/JMX.
• AWS: Metrics are pushed to Lightstep through a Cloudwatch Metric Stream. Supported AWS metrics can be found here.
• Google Cloud Monitoring: Metrics, resource tags, and metadata are sent to Cloud Observability using a service account that grants read-only metric access.
• Azure AKS: Using our OpenTelemetry Collector, metrics from Azure's managed Kuberentes service, AKS, can be easily sent to Cloud Observability.

For those looking to integrate specific platforms, a comprehensive list of metric integrations offers detailed guidelines, examples, and more.

Using Metric data

After metric data is ingested into Cloud Observability, users can: