Jason DeTiberus detiber

Pivot Testing

Initial testing, no Machine* objects, only a Cluster object in the default namespace.

Using cluster-api-provider-aws:

clusterctl create cluster -v 3   --bootstrap-type kind   --provider aws   -m ./cmd/clusterctl/examples/aws/out/machines.yaml   -c ./cmd/clusterctl/examples/aws/out/cluster.yaml   -p ./cmd/clusterctl/examples/aws/out/provider-components.yaml -a ./cmd/clusterctl/examples/aws/out/addons.yaml

sudo dnf config-manager --add-repo https://download.docker.com/linux/fedora/docker-ce.repo

# work around docker-ce-stable not being available yet for f29
sudo dnf install docker-ce --releasever=28

sudo dnf install golang
sudo dnf copr enable -y vbatts/bazel
sudo dnf install bazel

Cluster API Provider AWS Phases with KIND

Pre-demo Prep

Install KIND

go get -u sigs.k8s.io/kind

Cluster API Provider AWS Demo script

Pre-demo Prep

Build and install the artifacts

make clean
export DEV_DOCKER_REPO=gcr.io/$(gcloud config get-value project)
make clusterctl clusterawsadm docker-build-dev docker-push-dev

Testing kubeadm v1.12 - v1.13 stacked control plane upgrade

Install docker on each host:

amazon-linux-extras install docker=18.06.1
cat > /etc/docker/daemon.json <<EOF
{
  "exec-opts": ["native.cgroupdriver=systemd"]
}

Kubeadm stacked control plane install v1.12

Warning: hardcoded values below will need to be modified for testing against different hostnames and/or IPs

Common Steps

Install Docker

apt-get update && \

kubeadm Stacked HA Upgrade v1.11.x to v1.12.x

Warning: hardcoded values below will need to be modified for testing against different hostnames and/or IPs

Common Steps

Install Docker

apt-get update && \

kubeadm Stacked HA Upgrade v1.11 Minor upgrades

Warning: hardcoded values below will need to be modified for testing against different hostnames and/or IPs

Common Steps

Install Docker

apt-get update

Thoughts on future Kubeadm work required to streamline experience

Currently the kubeadm config and workflow is centered around a single control plane node and any number of worker nodes. This requires additional workarounds when trying to bootstrap and manage clusters with an HA control plane. I believe there is a path forward that would help streamline the effort required to stand up a fully managed HA control plane including etcd.

Etcd Config Changes
- Modify the local etcd config to have an additional parameter for exposing etcd ports or to expose the ports by default
- Add a attribute to override the exposed advertised ip, preferably by choosing an interface to use for detection
Make API Server Load Balancer config more explicit instead of just overriding the api config
- Add new attribute to specify a load balancer, which will implicitly set the api config if not explicitly set
Control Plane join/extend workflow

	package machine

	import (
	"fmt"

	"github.com/aws/aws-sdk-go/service/elbv2/elbv2iface"

	"github.com/aws/aws-sdk-go/service/iam/iamiface"

	"github.com/aws/aws-sdk-go/service/ec2/ec2iface"