Releases: kubernetes-csi/external-health-monitor
Releases · kubernetes-csi/external-health-monitor
v0.4.0
Overall Status
The Volume Health Monitor is implemented in two components: External Health Monitor Controller and Kubelet.
- The External Health Monitor Controller has been Alpha since Kubernetes 1.19.
- When this feature was first introduced in Kubernetes 1.19, there was an External Health Monitor Agent that monitors volume health from the node side. In the Kubernetes 1.21 release, the node side volume health monitoring logic was moved to Kubelet as an Alpha feature to avoid duplicate CSI RPC calls.
Supported CSI Spec Versions
1.3-1.5
Container
docker pull registry.k8s.io/sig-storage/csi-external-health-monitor-controller:v0.4.0
Deprecations
- Volume Health Monitoring Agent is deprecated. The node side volume health monitoring logic is moved to Kubelet in Kubernetes 1.21.
- Remove the deployment file of external-health-monitor-agent (#68, @fengzixu)
Feature
- Add command line arguments
leader-election-lease-duration,leader-election-renew-deadline, andleader-election-retry-periodto configure leader election options for the external health monitor controller (#90, @deepakkinni)
Full Changelog
v0.3.0
Overall Status
The Volume Health Monitor is implemented in two components: External Health Monitor Controller and Kubelet.
- The External Health Monitor Controller has been Alpha since Kubernetes 1.19.
- When this feature was first introduced in Kubernetes 1.19, there was an External Health Monitor Agent that monitors volume health from the node side. In the Kubernetes 1.21 release, the node side volume health monitoring logic was moved to Kubelet as an Alpha feature to avoid duplicate CSI RPC calls.
Minimum CSI Spec Version
1.3.0
Container
docker pull registry.k8s.io/sig-storage/csi-external-health-monitor-controller:v0.3.0
Deprecations
- Volume Health Monitoring Agent is deprecated. The node side volume health monitoring logic is moved to Kubelet in Kubernetes 1.21.
- Remove the build option for agent (#64, @fengzixu)
Uncategorized
Full Changelog
v0.2.0
Overall Status
It is an alpha feature introduced in Kubernetes v1.19.
Minimum CSI Spec Version
1.3.0
Container
docker pull registry.k8s.io/sig-storage/csi-external-health-monitor-controller:v0.2.0
docker pull registry.k8s.io/sig-storage/csi-external-health-monitor-agent:v0.2.0
Deprecations
- For both the external-health-monitor-controller and external-health-monitor-agent, the
metrics-addressflag is deprecated and replaced byhttp-endpoint, which enables handlers from both metrics manager and leader election health check. (#49, @verult)
Feature
- Send the recovery event when volume returns to the healthy state (#36, @fengzixu)
- For both the external-health-monitor-controller and external-health-monitor-agent, the
metrics-addressflag is deprecated and replaced byhttp-endpoint, which enables handlers from both metrics manager and leader election health check. (#49, @verult)
Bug or Regression
- Added leader election health check at the metrics port + path "/healthz/leader-election". (#46, @pohly)
- process_start_time metric is now reported (#46, @pohly)
- When node plugin returns NotFound error for NodeGetVolumeStats
re-queuing should not be performed (#48, @kazimsarikaya)
Full Changelog
v0.1.0
Overall Status
This is the first release of external-health-monitor. It is an alpha feature introduced in Kubernetes v1.19.
Minimum CSI Spec Version
1.3.0
Container
docker pull registry.k8s.io/sig-storage/csi-external-health-monitor-controller:v0.1.0
docker pull registry.k8s.io/sig-storage/csi-external-health-monitor-agent:v0.1.0
New Features
- external-health-monitor agent sends the event of volume condition to the pod (#22, @fengzixu)
- Adds an external health monitor controller and an external health monitor agent. (#3, @NickrenREN)
Other Notable Changes
- build with Go 1.15 (#23, @pohly)
- publishing of images on k8s.gcr.io (#17, @xing-yang)