05c27f2d73
Rolling ugprade is an important feature for a managed k8s service, at this stage, two user cases will be covered: 1. Upgrade base operating system 2. Upgrade k8s version Known limitation: When doing operating system upgrade, there is no chance to call kubectl drain to evict pods on that node. Task: 30185 Story: 2002210 Change-Id: Ibbed59bc135969174a20e5243ff8464908801a23
18 lines
836 B
YAML
18 lines
836 B
YAML
---
|
|
features:
|
|
- |
|
|
Now the fedora atomic Kubernetes driver can support rolling upgrade for k8s
|
|
version change or the image change. User can call command
|
|
`openstack coe cluster upgrade <cluster ID> <new cluster template ID>` to
|
|
upgrade current cluster to the new version defined in the new cluster
|
|
template. At this moment, only the image change and the kube_tag change
|
|
are supported.
|
|
issues:
|
|
- |
|
|
There is a known issue when doing image(operating system) upgrade for k8s
|
|
cluster. Because when doing image change for a server resource, Heat will
|
|
trigger the Nova rebuild to rebuild the instnace and there is no chance to
|
|
call kubectl drain to drain the node, so there could be a very minior
|
|
downtime when doing(starting to do) the rebuild and meanwhile a request
|
|
is routed to that node.
|