Support healing Kubernetes master/worker nodes with Mgmt driver
In this spec, the healing operation is supported after deploying a Kubernetes
cluster with MgmtDriver as VNF Lifecycle Management Interface in ETSI
NFV-SOL003 v2.6.1 [1]. When healing a master or worker node, not only the
default heal operation but also "heal_end" operation to register the node to
the existing Kubernetes cluster.
When healing a master node, the configuration for etcd and HA proxy are
needed. Also, when healing a worker node, applying the required configuration
to join the cluster is needed as almost the same procedure with the case of
scaling-out/in.
[1] https:/
Blueprint information
- Status:
- Complete
- Approver:
- None
- Priority:
- Undefined
- Drafter:
- Yoshito Ito
- Direction:
- Needs approval
- Assignee:
- Yoshito Ito
- Definition:
- New
- Series goal:
- Proposed for wallaby
- Implementation:
-
Implemented
- Milestone target:
-
wallaby-3
- Started by
- Yoshito Ito
- Completed by
- Yoshito Ito
Related branches
Related bugs
Sprints
Whiteboard
Gerrit topic: https:/
Addressed by: https:/
Support Healing Kubernetes Master/Worker-nodes with Mgmtdriver