Area(s)

area:k8s

What's missing?

I'd like to propose adding a k8s.service entity and associated metrics, in line with existing k8s entities. While pod-level health metrics exist, k8s services can aggregate endpoints across multiple deployments via label selectors. The proposed metrics provide a direct view of service-level availability without requiring users to discover and aggregate all matching pods.

Describe the solution you'd like

I would like to introduce the following attributes and metrics for a k8s.service entity.

New Metrics

New Attributes

Metric attributes:

• k8s.service.endpoint.condition (enum: ready, serving, terminating)
• k8s.service.endpoint.address_type (enum: IPv4, IPv6, FQDN)
• k8s.service.endpoint.zone (string)

Service entity:

• k8s.service.traffic_distribution (string, opt-in)
• k8s.service.publish_not_ready_addresses (bool, opt-in)

Listing out a few usecases for the proposed attributes and metrics -

• Service endpoint health during deployments - Track how many endpoints are ready to receive new traffic, still serving in-flight requests while terminating, or warming up. This is useful for assessing rolling update progress.
• LoadBalancer provisioning status - Confirm that external IPs or hostnames have been assigned by the load balancer controller, and detect when services lose their external endpoints.
• Zone-aware endpoint distribution - With topology-aware routing, operators need to verify each availability zone has sufficient endpoints and alert on zone imbalances.

Tip

_{React with 👍 to help prioritize this issue. Please use comments to provide useful context, avoiding +1 or me too, to help us triage it. Learn more here.}