You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently kubeletstats receiver only fetches metrics from /pods and /stats/summary
however I heard that there are other 2 endpoints /metrics and /metrics/caadvisor
I am not quite sure about this, I might have missed discussion this receiver does not add /metrics endpoints
If the receiver name is kubeletstats receiver, but we do not have metrics pertaining to the kubelet.
Example configuration for the component
I think current configuration is good, but if the cluster does NOT have caAdvisor, user can specifically mention that caAdivisor not exists.
Telemetry data types supported
The 2 more endpoints we can consider i.e, /metrics and /metrics/caadvisior
Prometheus data types would be returned for each of the new metric.
Code Owner(s)
No response
Sponsor (optional)
No response
Additional context
The reason is if kubelet itself looses due to some reasons, the whole node pods would be in not correct state. health of kubelet and it's operations metrics would definitely for the busy clusters.
No response
The text was updated successfully, but these errors were encountered:
developer1622
changed the title
Kubeletstats receiver: Adding feasibility in kubeletstats receiver to fetch kubelet metrics
[receiver/kubeletstatsreceiver] Kubeletstats receiver: Adding feasibility in kubeletstats receiver to fetch kubelet metrics
Nov 20, 2024
The purpose and use-cases of the new component
Currently kubeletstats receiver only fetches metrics from /pods and /stats/summary
however I heard that there are other 2 endpoints /metrics and /metrics/caadvisor
I am not quite sure about this, I might have missed discussion this receiver does not add /metrics endpoints
If the receiver name is kubeletstats receiver, but we do not have metrics pertaining to the kubelet.
Example configuration for the component
I think current configuration is good, but if the cluster does NOT have caAdvisor, user can specifically mention that caAdivisor not exists.
Telemetry data types supported
The 2 more endpoints we can consider i.e, /metrics and /metrics/caadvisior
Prometheus data types would be returned for each of the new metric.
Code Owner(s)
No response
Sponsor (optional)
No response
Additional context
The reason is if kubelet itself looses due to some reasons, the whole node pods would be in not correct state. health of kubelet and it's operations metrics would definitely for the busy clusters.
No response
The text was updated successfully, but these errors were encountered: