Skip to content

Latest commit

 

History

History
62 lines (46 loc) · 2.48 KB

roadmap.md

File metadata and controls

62 lines (46 loc) · 2.48 KB

Kubespray's roadmap

Self deployment (pull-mode) #320

  • the playbook would install and configure docker and the etcd cluster
  • the following data would be inserted into etcd: certs,tokens,users,inventory,group_vars.
  • a "kubespray" container would be deployed (kubespray-cli, ansible-playbook)
  • to be discussed, a way to provide the inventory
  • self deployment of the node from inside a container #321

Provisioning and cloud providers

  • Terraform to provision instances on:
    • GCE
    • AWS (contrib/terraform/aws)
    • OpenStack (contrib/terraform/openstack)
    • Equinix Metal
    • Digital Ocean
    • Azure
  • On AWS autoscaling, multi AZ
  • On Azure autoscaling, create loadbalancer #297
  • On GCE be able to create a loadbalancer automatically (IAM ?) #280
  • TLS bootstrap support for kubelet (covered by kubeadm, but not in standard deployment) #234 (related issues: kubernetes/kubernetes#20439 kubernetes/kubernetes#18112

Tests

  • Run kubernetes e2e tests
  • Test idempotency on single OS but for all network plugins/container engines
  • single test on AWS per day
  • test scale up cluster: +1 etcd, +1 control plane, +1 node
  • Reorganize CI test vars into group var files

Lifecycle

  • Upgrade granularity: select components to upgrade and skip others

Networking

  • Opencontrail
  • Consolidate roles/network_plugin and roles/kubernetes-apps/network_plugin

Kubespray API

  • Perform all actions through an API
  • Store inventories / configurations of multiple clusters
  • Make sure that state of cluster is completely saved in no more than one config file beyond hosts inventory

Addons (helm or native ansible)

  • Helm
  • Ingress-nginx
  • kubernetes-dashboard

Others

  • Organize and update documentation (split in categories)
  • Refactor downloads so it all runs in the beginning of deployment
  • Make bootstrapping OS more consistent
  • consul -> if officially supported by k8s
  • Flex volumes options (e.g. torrus support) #312
  • Clusters federation option (aka ubernetes) #329