kubespray/contrib/terraform/openstack

Kubernetes on Openstack with Terraform

Provision a Kubernetes cluster with Terraform on Openstack.

Status

This will install a Kubernetes cluster on an Openstack Cloud. It should work on most modern installs of OpenStack that support the basic services.

Known compatible public clouds

• Auro
• Betacloud
• CityCloud
• DreamHost
• ELASTX
• EnterCloudSuite
• FugaCloud
• Open Telekom Cloud : requires to set the variable wait_for_floatingip = "true" in your cluster.tfvars
• OVH
• Rackspace
• Ultimum
• VexxHost
• Zetta

Approach

The terraform configuration inspects variables found in variables.tf to create resources in your OpenStack cluster. There is a python script that reads the generated.tfstate file to generate a dynamic inventory that is consumed by the main ansible script to actually install kubernetes and stand up the cluster.

Networking

The configuration includes creating a private subnet with a router to the external net. It will allocate floating IPs from a pool and assign them to the hosts where that makes sense. You have the option of creating bastion hosts inside the private subnet to access the nodes there. Alternatively, a node with a floating IP can be used as a jump host to nodes without.

Kubernetes Nodes

You can create many different kubernetes topologies by setting the number of different classes of hosts. For each class there are options for allocating floating IP addresses or not.

• Master nodes with etcd
• Master nodes without etcd
• Standalone etcd hosts
• Kubernetes worker nodes

Note that the Ansible script will report an invalid configuration if you wind up with an even number of etcd instances since that is not a valid configuration. This restriction includes standalone etcd nodes that are deployed in a cluster along with master nodes with etcd replicas. As an example, if you have three master nodes with etcd replicas and three standalone etcd nodes, the script will fail since there are now six total etcd replicas.

GlusterFS

The Terraform configuration supports provisioning of an optional GlusterFS shared file system based on a separate set of VMs. To enable this, you need to specify:

• the number of Gluster hosts (minimum 2)
• Size of the non-ephemeral volumes to be attached to store the GlusterFS bricks
• Other properties related to provisioning the hosts

Even if you are using Container Linux by CoreOS for your cluster, you will still need the GlusterFS VMs to be based on either Debian or RedHat based images. Container Linux by CoreOS cannot serve GlusterFS, but can connect to it through binaries available on hyperkube v1.4.3_coreos.0 or higher.

Requirements

• Install Terraform 0.12 or later
• Install Ansible
• you already have a suitable OS image in Glance
• you already have a floating IP pool created
• you have security groups enabled
• you have a pair of keys generated that can be used to secure the new hosts

Module Architecture

The configuration is divided into three modules:

• Network
• IPs
• Compute

The main reason for splitting the configuration up in this way is to easily accommodate situations where floating IPs are limited by a quota or if you have any external references to the floating IP (e.g. DNS) that would otherwise have to be updated.

You can force your existing IPs by modifying the compute variables in kubespray.tf as follows:

k8s_master_fips = ["151.101.129.67"]
k8s_node_fips = ["151.101.129.68"]

Terraform

Terraform will be used to provision all of the OpenStack resources with base software as appropriate.

Configuration

Inventory files

Create an inventory directory for your cluster by copying the existing sample and linking the hosts script (used to build the inventory based on Terraform state):

$ cp -LRp contrib/terraform/openstack/sample-inventory inventory/$CLUSTER
$ cd inventory/$CLUSTER
$ ln -s ../../contrib/terraform/openstack/hosts
$ ln -s ../../contrib

This will be the base for subsequent Terraform commands.

OpenStack access and credentials

No provider variables are hardcoded inside variables.tf because Terraform supports various authentication methods for OpenStack: the older script and environment method (using openrc) as well as a newer declarative method, and different OpenStack environments may support Identity API version 2 or 3.

These are examples and may vary depending on your OpenStack cloud provider, for an exhaustive list on how to authenticate on OpenStack with Terraform please read the OpenStack provider documentation.

Declarative method (recommended)

The recommended authentication method is to describe credentials in a YAML file clouds.yaml that can be stored in:

• the current directory
• ~/.config/openstack
• /etc/openstack

clouds.yaml:

clouds:
  mycloud:
    auth:
      auth_url: https://openstack:5000/v3
      username: "username"
      project_name: "projectname"
      project_id: projectid
      user_domain_name: "Default"
      password: "password"
    region_name: "RegionOne"
    interface: "public"
    identity_api_version: 3

If you have multiple clouds defined in your clouds.yaml file you can choose the one you want to use with the environment variable OS_CLOUD:

export OS_CLOUD=mycloud

Openrc method

When using classic environment variables, Terraform uses default OS_* environment variables. A script suitable for your environment may be available from Horizon under Project -> Compute -> Access & Security -> API Access.

With identity v2:

source openrc

env | grep OS

OS_AUTH_URL=https://openstack:5000/v2.0
OS_PROJECT_ID=projectid
OS_PROJECT_NAME=projectname
OS_USERNAME=username
OS_PASSWORD=password
OS_REGION_NAME=RegionOne
OS_INTERFACE=public
OS_IDENTITY_API_VERSION=2

With identity v3:

source openrc

env | grep OS

OS_AUTH_URL=https://openstack:5000/v3
OS_PROJECT_ID=projectid
OS_PROJECT_NAME=username
OS_PROJECT_DOMAIN_ID=default
OS_USERNAME=username
OS_PASSWORD=password
OS_REGION_NAME=RegionOne
OS_INTERFACE=public
OS_IDENTITY_API_VERSION=3
OS_USER_DOMAIN_NAME=Default

Terraform does not support a mix of DomainName and DomainID, choose one or the other:

* provider.openstack: You must provide exactly one of DomainID or DomainName to authenticate by Username

unset OS_USER_DOMAIN_NAME
export OS_USER_DOMAIN_ID=default

or

unset OS_PROJECT_DOMAIN_ID
set OS_PROJECT_DOMAIN_NAME=Default

Cluster variables

The construction of the cluster is driven by values found in variables.tf.

For your cluster, edit inventory/$CLUSTER/cluster.tfvars.

Variable	Description
cluster_name	All OpenStack resources will use the Terraform variablecluster_name (defaultexample) in their name to make it easier to track. For example the first compute resource will be namedexample-kubernetes-1.
az_list	List of Availability Zones available in your OpenStack cluster.
network_name	The name to be given to the internal network that will be generated
network_dns_domain	(Optional) The dns_domain for the internal network that will be generated
dns_nameservers	An array of DNS name server names to be used by hosts in the internal subnet.
floatingip_pool	Name of the pool from which floating IPs will be allocated
external_net	UUID of the external network that will be routed to
flavor_k8s_master,flavor_k8s_node,flavor_etcd, flavor_bastion,flavor_gfs_node	Flavor depends on your openstack installation, you can get available flavor IDs through openstack flavor list
image,image_gfs	Name of the image to use in provisioning the compute resources. Should already be loaded into glance.
ssh_user,ssh_user_gfs	The username to ssh into the image with. This usually depends on the image you have selected
public_key_path	Path on your local workstation to the public key file you wish to use in creating the key pairs
number_of_k8s_masters, number_of_k8s_masters_no_floating_ip	Number of nodes that serve as both master and etcd. These can be provisioned with or without floating IP addresses
number_of_k8s_masters_no_etcd, number_of_k8s_masters_no_floating_ip_no_etcd	Number of nodes that serve as just master with no etcd. These can be provisioned with or without floating IP addresses
number_of_etcd	Number of pure etcd nodes
number_of_k8s_nodes, number_of_k8s_nodes_no_floating_ip	Kubernetes worker nodes. These can be provisioned with or without floating ip addresses.
number_of_bastions	Number of bastion hosts to create. Scripts assume this is really just zero or one
number_of_gfs_nodes_no_floating_ip	Number of gluster servers to provision.
gfs_volume_size_in_gb	Size of the non-ephemeral volumes to be attached to store the GlusterFS bricks
supplementary_master_groups	To add ansible groups to the masters, such as kube-node for tainting them as nodes, empty by default.
supplementary_node_groups	To add ansible groups to the nodes, such as kube-ingress for running ingress controller pods, empty by default.
bastion_allowed_remote_ips	List of CIDR allowed to initiate a SSH connection, ["0.0.0.0/0"] by default
master_allowed_remote_ips	List of CIDR blocks allowed to initiate an API connection, ["0.0.0.0/0"] by default
k8s_allowed_remote_ips	List of CIDR allowed to initiate a SSH connection, empty by default
worker_allowed_ports	List of ports to open on worker nodes, [{ "protocol" = "tcp", "port_range_min" = 30000, "port_range_max" = 32767, "remote_ip_prefix" = "0.0.0.0/0"}] by default
wait_for_floatingip	Let Terraform poll the instance until the floating IP has been associated, false by default.
node_root_volume_size_in_gb	Size of the root volume for nodes, 0 to use ephemeral storage
master_root_volume_size_in_gb	Size of the root volume for masters, 0 to use ephemeral storage
gfs_root_volume_size_in_gb	Size of the root volume for gluster, 0 to use ephemeral storage
etcd_root_volume_size_in_gb	Size of the root volume for etcd nodes, 0 to use ephemeral storage
bastion_root_volume_size_in_gb	Size of the root volume for bastions, 0 to use ephemeral storage
use_server_group	Create and use openstack nova servergroups, default: false

Terraform state files

In the cluster's inventory folder, the following files might be created (either by Terraform or manually), to prevent you from pushing them accidentally they are in a .gitignore file in the terraform/openstack directory :

• .terraform
• .tfvars
• .tfstate
• .tfstate.backup

You can still add them manually if you want to.

Initialization

Before Terraform can operate on your cluster you need to install the required plugins. This is accomplished as follows:

$ cd inventory/$CLUSTER
$ terraform init ../../contrib/terraform/openstack

This should finish fairly quickly telling you Terraform has successfully initialized and loaded necessary modules.

Provisioning cluster

You can apply the Terraform configuration to your cluster with the following command issued from your cluster's inventory directory (inventory/$CLUSTER):

$ terraform apply -var-file=cluster.tfvars ../../contrib/terraform/openstack

if you chose to create a bastion host, this script will create contrib/terraform/openstack/k8s-cluster.yml with an ssh command for Ansible to be able to access your machines tunneling through the bastion's IP address. If you want to manually handle the ssh tunneling to these machines, please delete or move that file. If you want to use this, just leave it there, as ansible will pick it up automatically.

Destroying cluster

You can destroy your new cluster with the following command issued from the cluster's inventory directory:

$ terraform destroy -var-file=cluster.tfvars ../../contrib/terraform/openstack

If you've started the Ansible run, it may also be a good idea to do some manual cleanup:

• remove SSH keys from the destroyed cluster from your ~/.ssh/known_hosts file
• clean up any temporary cache files: rm /tmp/$CLUSTER-*

Debugging

You can enable debugging output from Terraform by setting OS_DEBUG to 1 andTF_LOG toDEBUG before running the Terraform command.

Terraform output

Terraform can output values that are useful for configure Neutron/Octavia LBaaS or Cinder persistent volume provisioning as part of your Kubernetes deployment:

• private_subnet_id: the subnet where your instances are running is used for openstack_lbaas_subnet_id
• floating_network_id: the network_id where the floating IP are provisioned is used for openstack_lbaas_floating_network_id

Ansible

Node access

SSH

Ensure your local ssh-agent is running and your ssh key has been added. This step is required by the terraform provisioner:

$ eval $(ssh-agent -s)
$ ssh-add ~/.ssh/id_rsa

If you have deployed and destroyed a previous iteration of your cluster, you will need to clear out any stale keys from your SSH "known hosts" file ( ~/.ssh/known_hosts).

Metadata variables

The python script that reads the generated.tfstate file to generate a dynamic inventory recognizes some variables within a "metadata" block, defined in a "resource" block (example):

resource "openstack_compute_instance_v2" "example" {
    ...
    metadata {
        ssh_user = "ubuntu"
        prefer_ipv6 = true
	python_bin = "/usr/bin/python3"
    }
    ...
}

As the example shows, these let you define the SSH username for Ansible, a Python binary which is needed by Ansible if /usr/bin/python doesn't exist, and whether the IPv6 address of the instance should be preferred over IPv4.

Bastion host

Bastion access will be determined by:

• Your choice on the amount of bastion hosts (set by number_of_bastions terraform variable).
• The existence of nodes/masters with floating IPs (set by number_of_k8s_masters, number_of_k8s_nodes, number_of_k8s_masters_no_etcd terraform variables).

If you have a bastion host, your ssh traffic will be directly routed through it. This is regardless of whether you have masters/nodes with a floating IP assigned. If you don't have a bastion host, but at least one of your masters/nodes have a floating IP, then ssh traffic will be tunneled by one of these machines.

So, either a bastion host, or at least master/node with a floating IP are required.

Test access

Make sure you can connect to the hosts. Note that Container Linux by CoreOS will have a state FAILED due to Python not being present. This is okay, because Python will be installed during bootstrapping, so long as the hosts are not UNREACHABLE.

$ ansible -i inventory/$CLUSTER/hosts -m ping all
example-k8s_node-1 | SUCCESS => {
    "changed": false,
    "ping": "pong"
}
example-etcd-1 | SUCCESS => {
    "changed": false,
    "ping": "pong"
}
example-k8s-master-1 | SUCCESS => {
    "changed": false,
    "ping": "pong"
}

If it fails try to connect manually via SSH. It could be something as simple as a stale host key.

Configure cluster variables

Edit inventory/$CLUSTER/group_vars/all/all.yml:

• bin_dir:

# Directory where the binaries will be installed
# Default:
# bin_dir: /usr/local/bin
# For Container Linux by CoreOS:
bin_dir: /opt/bin

• and cloud_provider:

cloud_provider: openstack

Edit inventory/$CLUSTER/group_vars/k8s-cluster/k8s-cluster.yml:

• Set variable kube_network_plugin to your desired networking plugin.
  • flannel works out-of-the-box
  • calico requires configuring OpenStack Neutron ports to allow service and pod subnets

# Choose network plugin (calico, weave or flannel)
# Can also be set to 'cloud', which lets the cloud provider setup appropriate routing
kube_network_plugin: flannel

• Set variable resolvconf_mode

# Can be docker_dns, host_resolvconf or none
# Default:
# resolvconf_mode: docker_dns
# For Container Linux by CoreOS:
resolvconf_mode: host_resolvconf

• Set max amount of attached cinder volume per host (default 256)

node_volume_attach_limit: 26

• Disable access_ip, this will make all innternal cluster traffic to be sent over local network when a floating IP is attached (default this value is set to 1)

use_access_ip: 0

Deploy Kubernetes

$ ansible-playbook --become -i inventory/$CLUSTER/hosts cluster.yml

This will take some time as there are many tasks to run.

Kubernetes

Set up kubectl

1. Install kubectl on your workstation
2. Add a route to the internal IP of a master node (if needed):

sudo route add [master-internal-ip] gw [router-ip]

or

sudo route add -net [internal-subnet]/24 gw [router-ip]

3. List Kubernetes certificates & keys:

ssh [os-user]@[master-ip] sudo ls /etc/kubernetes/ssl/

4. Get admin's certificates and keys:

ssh [os-user]@[master-ip] sudo cat /etc/kubernetes/ssl/admin-kube-master-1-key.pem > admin-key.pem
ssh [os-user]@[master-ip] sudo cat /etc/kubernetes/ssl/admin-kube-master-1.pem > admin.pem
ssh [os-user]@[master-ip] sudo cat /etc/kubernetes/ssl/ca.pem > ca.pem

5. Configure kubectl:

$ kubectl config set-cluster default-cluster --server=https://[master-internal-ip]:6443 \
    --certificate-authority=ca.pem

$ kubectl config set-credentials default-admin \
    --certificate-authority=ca.pem \
    --client-key=admin-key.pem \
    --client-certificate=admin.pem

$ kubectl config set-context default-system --cluster=default-cluster --user=default-admin
$ kubectl config use-context default-system

7. Check it:

kubectl version

GlusterFS

GlusterFS is not deployed by the standardcluster.yml playbook, see the GlusterFS playbook documentation for instructions.

Basically you will install Gluster as

$ ansible-playbook --become -i inventory/$CLUSTER/hosts ./contrib/network-storage/glusterfs/glusterfs.yml

What's next

Try out your new Kubernetes cluster with the Hello Kubernetes service.