1. join node add LB discoveryTokenAPIServers
2. kubeadm_config_api_fqdn support ipddress and domain_name

* calico upgrade to v3

* update calico_rr version

* add missing file

* change contents of main.yml as it was left old version

* enable network policy by default

* remove unneeded task

* Fix kubelet calico settings

* fix when statement

* switch back to node-kubeconfig.yaml

A typo in the adduser role prevents the createhome
variable to be taken into account.

Fix #3164

1.fix audit log not write
2.fix Parameter not recognized
3.delete kubedm futuregates auditing and use apiServerExtraArgs

Changed author

Signed-off-by: Jeff Bornemann <jeff.bornemann@oracle.com>

we use FQDN as inventory_hostname

Define apiserver flags directly instead of relying on auditPolicy section in order to have the ability to redirect audit log to stdout with kubeadm

Hostname must be a valid DNS name, which is checked as https://github.com/kubernetes/apimachinery/blob/master/pkg/util/validation/validation.go#L115

The situation I have encountered is that my hostname contained underscore which is disallowed and apiserver refused to start.

Nginx LB(default)
Other LB by kubeadm controlplane

Trigger reboot handler when version upgrade during update script

Upstream Changes:

-   cephfs-provisioner v2.0.0-k8s1.11 (https://github.com/kubernetes-incubator/external-storage/releases/tag/cephfs-provisioner-v2.0.0-k8s1.11)
-   Update ClusterRole

Our Changes:

-   Fix typo in defaults/main.yml (rs -> deploy)
-   Manifests cleanup

Fix DiscoveryTokenCACertHashes key to discoveryTokenCACertHashes in kubeadm-client.conf

Reset tasks specific to a network_plugin moved inside its role directory + Reset tasks specific to cilium