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
When the liveness check fails, the container should be restarted. Currently, the failure of the check can only be determined through the /metrics path with csi_liveness as shown below:
$ curl 10.46.1.6:8680/metrics|more
# HELP csi_liveness Liveness Probe
# TYPE csi_liveness gauge
csi_liveness 0
This is not very friendly for pod probe checks in Kubernetes. A common way to check pods is through the following method:
Describe the feature you'd like to have
When the liveness check fails, the container should be restarted. Currently, the failure of the check can only be determined through the /metrics path with csi_liveness as shown below:
This is not very friendly for pod probe checks in Kubernetes. A common way to check pods is through the following method:
Could you provide an endpoint for liveness to facilitate probe checks?
What is the value to the end user? (why is it a priority?)
When a health check fails, the container can be restarted instead of showing as running but not being able to do anything.
How will we know we have a good solution? (acceptance criteria)
The text was updated successfully, but these errors were encountered: