Add Google Cloud fedora-atomic module

* Network load balancer for ingress doesn't work yet
because Compute Engine packages are missing
* port-forward / socat is broken
This commit is contained in:
Dalton Hubble 2018-04-15 20:40:00 -07:00
parent 24d230505a
commit 2b74aba564
17 changed files with 1062 additions and 0 deletions

View File

@ -0,0 +1,23 @@
The MIT License (MIT)
Copyright (c) 2017 Typhoon Authors
Copyright (c) 2017 Dalton Hubble
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
THE SOFTWARE.

View File

@ -0,0 +1,22 @@
# Typhoon <img align="right" src="https://storage.googleapis.com/poseidon/typhoon-logo.png">
Typhoon is a minimal and free Kubernetes distribution.
* Minimal, stable base Kubernetes distribution
* Declarative infrastructure and configuration
* Free (freedom and cost) and privacy-respecting
* Practical for labs, datacenters, and clouds
Typhoon distributes upstream Kubernetes, architectural conventions, and cluster addons, much like a GNU/Linux distribution provides the Linux kernel and userspace components.
## Features <a href="https://www.cncf.io/certification/software-conformance/"><img align="right" src="https://storage.googleapis.com/poseidon/certified-kubernetes.png"></a>
* Kubernetes v1.10.0 (upstream, via [kubernetes-incubator/bootkube](https://github.com/kubernetes-incubator/bootkube))
* Single or multi-master, workloads isolated on workers, [Calico](https://www.projectcalico.org/) or [flannel](https://github.com/coreos/flannel) networking
* On-cluster etcd with TLS, [RBAC](https://kubernetes.io/docs/admin/authorization/rbac/)-enabled, [network policy](https://kubernetes.io/docs/concepts/services-networking/network-policies/)
* Ready for Ingress, Prometheus, Grafana, and other optional [addons](https://typhoon.psdn.io/addons/overview/)
## Docs
Please see the [official docs](https://typhoon.psdn.io) and the Google Cloud [tutorial](https://typhoon.psdn.io/google-cloud/).

View File

@ -0,0 +1,95 @@
# TCP Proxy load balancer DNS record
resource "google_dns_record_set" "apiserver" {
# DNS Zone name where record should be created
managed_zone = "${var.dns_zone_name}"
# DNS record
name = "${format("%s.%s.", var.cluster_name, var.dns_zone)}"
type = "A"
ttl = 300
# IPv4 address of apiserver TCP Proxy load balancer
rrdatas = ["${google_compute_global_address.apiserver-ipv4.address}"]
}
# Static IPv4 address for the TCP Proxy Load Balancer
resource "google_compute_global_address" "apiserver-ipv4" {
name = "${var.cluster_name}-apiserver-ip"
ip_version = "IPV4"
}
# Forward IPv4 TCP traffic to the TCP proxy load balancer
resource "google_compute_global_forwarding_rule" "apiserver" {
name = "${var.cluster_name}-apiserver"
ip_address = "${google_compute_global_address.apiserver-ipv4.address}"
ip_protocol = "TCP"
port_range = "443"
target = "${google_compute_target_tcp_proxy.apiserver.self_link}"
}
# TCP Proxy Load Balancer for apiservers
resource "google_compute_target_tcp_proxy" "apiserver" {
name = "${var.cluster_name}-apiserver"
description = "Distribute TCP load across ${var.cluster_name} controllers"
backend_service = "${google_compute_backend_service.apiserver.self_link}"
}
# Backend service backed by unmanaged instance groups
resource "google_compute_backend_service" "apiserver" {
name = "${var.cluster_name}-apiserver"
description = "${var.cluster_name} apiserver service"
protocol = "TCP"
port_name = "apiserver"
session_affinity = "NONE"
timeout_sec = "60"
# controller(s) spread across zonal instance groups
backend {
group = "${google_compute_instance_group.controllers.0.self_link}"
}
backend {
group = "${google_compute_instance_group.controllers.1.self_link}"
}
backend {
group = "${google_compute_instance_group.controllers.2.self_link}"
}
health_checks = ["${google_compute_health_check.apiserver.self_link}"]
}
# Instance group of heterogeneous (unmanged) controller instances
resource "google_compute_instance_group" "controllers" {
count = "${length(local.zones)}"
name = "${format("%s-controllers-%s", var.cluster_name, element(local.zones, count.index))}"
zone = "${element(local.zones, count.index)}"
named_port {
name = "apiserver"
port = "443"
}
# add instances in the zone into the instance group
instances = [
"${matchkeys(google_compute_instance.controllers.*.self_link,
google_compute_instance.controllers.*.zone,
list(element(local.zones, count.index)))}"
]
}
# TCP health check for apiserver
resource "google_compute_health_check" "apiserver" {
name = "${var.cluster_name}-apiserver-tcp-health"
description = "TCP health check for kube-apiserver"
timeout_sec = 5
check_interval_sec = 5
healthy_threshold = 1
unhealthy_threshold = 3
tcp_health_check {
port = "443"
}
}

View File

@ -0,0 +1,17 @@
# Self-hosted Kubernetes assets (kubeconfig, manifests)
module "bootkube" {
source = "git::https://github.com/poseidon/terraform-render-bootkube.git?ref=db36b92abced3c4b0af279adfd5ed4bf0cf8c39f"
cluster_name = "${var.cluster_name}"
api_servers = ["${format("%s.%s", var.cluster_name, var.dns_zone)}"]
etcd_servers = ["${null_resource.repeat.*.triggers.domain}"]
asset_dir = "${var.asset_dir}"
networking = "${var.networking}"
network_mtu = 1440
pod_cidr = "${var.pod_cidr}"
service_cidr = "${var.service_cidr}"
cluster_domain_suffix = "${var.cluster_domain_suffix}"
# Fedora
trusted_certs_dir = "/etc/pki/tls/certs"
}

View File

@ -0,0 +1,124 @@
#cloud-config
write_files:
- path: /etc/etcd/etcd.conf
content: |
ETCD_NAME=${etcd_name}
ETCD_DATA_DIR=/var/lib/etcd
ETCD_ADVERTISE_CLIENT_URLS=https://${etcd_domain}:2379
ETCD_INITIAL_ADVERTISE_PEER_URLS=https://${etcd_domain}:2380
ETCD_LISTEN_CLIENT_URLS=https://0.0.0.0:2379
ETCD_LISTEN_PEER_URLS=https://0.0.0.0:2380
ETCD_LISTEN_METRICS_URLS=http://0.0.0.0:2381
ETCD_INITIAL_CLUSTER=${etcd_initial_cluster}
ETCD_STRICT_RECONFIG_CHECK=true
ETCD_TRUSTED_CA_FILE=/etc/ssl/certs/etcd/server-ca.crt
ETCD_CERT_FILE=/etc/ssl/certs/etcd/server.crt
ETCD_KEY_FILE=/etc/ssl/certs/etcd/server.key
ETCD_CLIENT_CERT_AUTH=true
ETCD_PEER_TRUSTED_CA_FILE=/etc/ssl/certs/etcd/peer-ca.crt
ETCD_PEER_CERT_FILE=/etc/ssl/certs/etcd/peer.crt
ETCD_PEER_KEY_FILE=/etc/ssl/certs/etcd/peer.key
ETCD_PEER_CLIENT_CERT_AUTH=true
- path: /etc/systemd/system/cloud-metadata.service
content: |
[Unit]
Description=Cloud metadata agent
[Service]
Type=oneshot
Environment=OUTPUT=/run/metadata/cloud
ExecStart=/usr/bin/mkdir -p /run/metadata
ExecStart=/usr/bin/bash -c 'echo "HOSTNAME_OVERRIDE=$(curl\
-H "Metadata-Flavor: Google"\
--url http://metadata.google.internal/computeMetadata/v1/instance/hostname\
--retry 10)" > $${OUTPUT}'
[Install]
WantedBy=multi-user.target
- path: /etc/systemd/system/kubelet.service.d/10-typhoon.conf
content: |
[Unit]
Requires=cloud-metadata.service
After=cloud-metadata.service
Wants=rpc-statd.service
[Service]
ExecStartPre=/bin/mkdir -p /opt/cni/bin
ExecStartPre=/bin/mkdir -p /etc/kubernetes/manifests
ExecStartPre=/bin/mkdir -p /etc/kubernetes/cni/net.d
ExecStartPre=/bin/mkdir -p /etc/kubernetes/checkpoint-secrets
ExecStartPre=/bin/mkdir -p /etc/kubernetes/inactive-manifests
ExecStartPre=/bin/mkdir -p /var/lib/cni
ExecStartPre=/bin/mkdir -p /var/lib/kubelet/volumeplugins
ExecStartPre=/usr/bin/bash -c "grep 'certificate-authority-data' /etc/kubernetes/kubeconfig | awk '{print $2}' | base64 -d > /etc/kubernetes/ca.crt"
Restart=always
RestartSec=10
- path: /etc/kubernetes/kubelet.conf
content: |
ARGS="--allow-privileged \
--anonymous-auth=false \
--client-ca-file=/etc/kubernetes/ca.crt \
--cluster_dns=${k8s_dns_service_ip} \
--cluster_domain=${cluster_domain_suffix} \
--cni-conf-dir=/etc/kubernetes/cni/net.d \
--exit-on-lock-contention \
--kubeconfig=/etc/kubernetes/kubeconfig \
--lock-file=/var/run/lock/kubelet.lock \
--network-plugin=cni \
--node-labels=node-role.kubernetes.io/master \
--node-labels=node-role.kubernetes.io/controller="true" \
--pod-manifest-path=/etc/kubernetes/manifests \
--register-with-taints=node-role.kubernetes.io/master=:NoSchedule \
--volume-plugin-dir=/var/lib/kubelet/volumeplugins"
- path: /etc/kubernetes/kubeconfig
permissions: '0644'
content: |
${kubeconfig}
- path: /etc/systemd/system/bootkube.service
content: |
[Unit]
Description=Bootstrap a Kubernetes cluster
ConditionPathExists=!/var/bootkube/init_bootkube.done
[Service]
Type=oneshot
RemainAfterExit=true
WorkingDirectory=/var/bootkube
ExecStartPre=/bin/mkdir -p /var/bootkube
ExecStart=/usr/local/bin/bootkube-start
ExecStartPost=/bin/touch /var/bootkube/init_bootkube.done
[Install]
WantedBy=multi-user.target
- path: /var/bootkube/.keep
- path: /usr/local/bin/bootkube-start
permissions: '0755'
content: |
#!/bin/bash -e
# Wrapper for bootkube start
[ -n "$(ls /var/bootkube/assets/manifests-*/* 2>/dev/null)" ] && mv /var/bootkube/assets/manifests-*/* /var/bootkube/assets/manifests && rm -rf /var/bootkube/assets/manifests-*
/usr/bin/docker run --rm --name bootkube \
--net=host \
--volume /etc/kubernetes:/etc/kubernetes:Z \
--volume /var/bootkube/assets:/assets:Z \
--entrypoint=/bootkube \
quay.io/coreos/bootkube:v0.12.0 start --asset-dir=/assets
- path: /etc/selinux/config
owner: root:root
permissions: '0644'
content: |
SELINUX=permissive
SELINUXTYPE=targeted
bootcmd:
- [setenforce, Permissive]
- [systemctl, disable, firewalld, --now]
runcmd:
- [systemctl, daemon-reload]
- "atomic install --system --name=etcd quay.io/dghubble/etcd:0265e6680d2533f3fbf4512af868d29ff07451ca"
- [systemctl, start, --no-block, etcd.service]
- [systemctl, enable, cloud-metadata.service]
- "atomic install --system --name=kubelet quay.io/dghubble/kubelet:8767d4433f7c5a38c55edf6e682efb53fcd06f61"
- [systemctl, start, --no-block, kubelet.service]
users:
- default
- name: fedora
gecos: Fedora Admin
sudo: ALL=(ALL) NOPASSWD:ALL
groups: wheel,adm,systemd-journal,docker
ssh-authorized-keys:
- "${ssh_authorized_key}"

View File

@ -0,0 +1,91 @@
# Discrete DNS records for each controller's private IPv4 for etcd usage
resource "google_dns_record_set" "etcds" {
count = "${var.controller_count}"
# DNS Zone name where record should be created
managed_zone = "${var.dns_zone_name}"
# DNS record
name = "${format("%s-etcd%d.%s.", var.cluster_name, count.index, var.dns_zone)}"
type = "A"
ttl = 300
# private IPv4 address for etcd
rrdatas = ["${element(google_compute_instance.controllers.*.network_interface.0.address, count.index)}"]
}
# Zones in the region
data "google_compute_zones" "all" {
region = "${var.region}"
}
locals {
# TCP proxy load balancers require a fixed number of zonal backends. Spread
# controllers over up to 3 zones, since all GCP regions have at least 3.
zones = "${slice(data.google_compute_zones.all.names, 0, 3)}"
controllers_ipv4_public = ["${google_compute_instance.controllers.*.network_interface.0.access_config.0.assigned_nat_ip}"]
}
# Controller instances
resource "google_compute_instance" "controllers" {
count = "${var.controller_count}"
name = "${var.cluster_name}-controller-${count.index}"
zone = "${element(local.zones, count.index)}"
machine_type = "${var.controller_type}"
metadata {
user-data = "${element(data.template_file.controller-cloudinit.*.rendered, count.index)}"
}
boot_disk {
auto_delete = true
initialize_params {
image = "${var.os_image}"
size = "${var.disk_size}"
}
}
network_interface {
network = "${google_compute_network.network.name}"
# Ephemeral external IP
access_config = {}
}
can_ip_forward = true
tags = ["${var.cluster_name}-controller"]
}
# Controller Cloud-Init
data "template_file" "controller-cloudinit" {
count = "${var.controller_count}"
template = "${file("${path.module}/cloudinit/controller.yaml.tmpl")}"
vars = {
# Cannot use cyclic dependencies on controllers or their DNS records
etcd_name = "etcd${count.index}"
etcd_domain = "${var.cluster_name}-etcd${count.index}.${var.dns_zone}"
# etcd0=https://cluster-etcd0.example.com,etcd1=https://cluster-etcd1.example.com,...
etcd_initial_cluster = "${join(",", formatlist("%s=https://%s:2380", null_resource.repeat.*.triggers.name, null_resource.repeat.*.triggers.domain))}"
kubeconfig = "${indent(6, module.bootkube.kubeconfig)}"
ssh_authorized_key = "${var.ssh_authorized_key}"
k8s_dns_service_ip = "${cidrhost(var.service_cidr, 10)}"
cluster_domain_suffix = "${var.cluster_domain_suffix}"
}
}
# Horrible hack to generate a Terraform list of a desired length without dependencies.
# Ideal ${repeat("etcd", 3) -> ["etcd", "etcd", "etcd"]}
resource null_resource "repeat" {
count = "${var.controller_count}"
triggers {
name = "etcd${count.index}"
domain = "${var.cluster_name}-etcd${count.index}.${var.dns_zone}"
}
}

View File

@ -0,0 +1,149 @@
resource "google_compute_network" "network" {
name = "${var.cluster_name}"
description = "Network for the ${var.cluster_name} cluster"
auto_create_subnetworks = true
}
resource "google_compute_firewall" "allow-ssh" {
name = "${var.cluster_name}-allow-ssh"
network = "${google_compute_network.network.name}"
allow {
protocol = "tcp"
ports = [22]
}
source_ranges = ["0.0.0.0/0"]
target_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
}
resource "google_compute_firewall" "allow-apiserver" {
name = "${var.cluster_name}-allow-apiserver"
network = "${google_compute_network.network.name}"
allow {
protocol = "tcp"
ports = [443]
}
source_ranges = ["0.0.0.0/0"]
target_tags = ["${var.cluster_name}-controller"]
}
resource "google_compute_firewall" "allow-ingress" {
name = "${var.cluster_name}-allow-ingress"
network = "${google_compute_network.network.name}"
allow {
protocol = "tcp"
ports = [80, 443]
}
source_ranges = ["0.0.0.0/0"]
target_tags = ["${var.cluster_name}-worker"]
}
resource "google_compute_firewall" "internal-etcd" {
name = "${var.cluster_name}-internal-etcd"
network = "${google_compute_network.network.name}"
allow {
protocol = "tcp"
ports = [2380]
}
source_tags = ["${var.cluster_name}-controller"]
target_tags = ["${var.cluster_name}-controller"]
}
# Allow Prometheus to scrape etcd metrics
resource "google_compute_firewall" "internal-etcd-metrics" {
name = "${var.cluster_name}-internal-etcd-metrics"
network = "${google_compute_network.network.name}"
allow {
protocol = "tcp"
ports = [2381]
}
source_tags = ["${var.cluster_name}-worker"]
target_tags = ["${var.cluster_name}-controller"]
}
# Calico BGP and IPIP
# https://docs.projectcalico.org/v2.5/reference/public-cloud/gce
resource "google_compute_firewall" "internal-calico" {
count = "${var.networking == "calico" ? 1 : 0}"
name = "${var.cluster_name}-internal-calico"
network = "${google_compute_network.network.name}"
allow {
protocol = "tcp"
ports = ["179"]
}
allow {
protocol = "ipip"
}
source_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
target_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
}
# flannel
resource "google_compute_firewall" "internal-flannel" {
count = "${var.networking == "flannel" ? 1 : 0}"
name = "${var.cluster_name}-internal-flannel"
network = "${google_compute_network.network.name}"
allow {
protocol = "udp"
ports = [8472]
}
source_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
target_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
}
# Allow Prometheus to scrape node-exporter daemonset
resource "google_compute_firewall" "internal-node-exporter" {
name = "${var.cluster_name}-internal-node-exporter"
network = "${google_compute_network.network.name}"
allow {
protocol = "tcp"
ports = [9100]
}
source_tags = ["${var.cluster_name}-worker"]
target_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
}
# kubelet API to allow kubectl exec and log
resource "google_compute_firewall" "internal-kubelet" {
name = "${var.cluster_name}-internal-kubelet"
network = "${google_compute_network.network.name}"
allow {
protocol = "tcp"
ports = [10250]
}
source_tags = ["${var.cluster_name}-controller"]
target_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
}
resource "google_compute_firewall" "internal-kubelet-readonly" {
name = "${var.cluster_name}-internal-kubelet-readonly"
network = "${google_compute_network.network.name}"
allow {
protocol = "tcp"
ports = [10255]
}
source_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
target_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
}

View File

@ -0,0 +1,17 @@
output "ingress_static_ip" {
value = "${module.workers.ingress_static_ip}"
}
output "network_self_link" {
value = "${google_compute_network.network.self_link}"
}
# Outputs for worker pools
output "network_name" {
value = "${google_compute_network.network.name}"
}
output "kubeconfig" {
value = "${module.bootkube.kubeconfig}"
}

View File

@ -0,0 +1,25 @@
# Terraform version and plugin versions
terraform {
required_version = ">= 0.10.4"
}
provider "google" {
version = "~> 1.6"
}
provider "local" {
version = "~> 1.0"
}
provider "null" {
version = "~> 1.0"
}
provider "template" {
version = "~> 1.0"
}
provider "tls" {
version = "~> 1.0"
}

View File

@ -0,0 +1,89 @@
# Secure copy etcd TLS assets to controllers.
resource "null_resource" "copy-controller-secrets" {
count = "${var.controller_count}"
connection {
type = "ssh"
host = "${element(local.controllers_ipv4_public, count.index)}"
user = "fedora"
timeout = "15m"
}
provisioner "file" {
content = "${module.bootkube.etcd_ca_cert}"
destination = "$HOME/etcd-client-ca.crt"
}
provisioner "file" {
content = "${module.bootkube.etcd_client_cert}"
destination = "$HOME/etcd-client.crt"
}
provisioner "file" {
content = "${module.bootkube.etcd_client_key}"
destination = "$HOME/etcd-client.key"
}
provisioner "file" {
content = "${module.bootkube.etcd_server_cert}"
destination = "$HOME/etcd-server.crt"
}
provisioner "file" {
content = "${module.bootkube.etcd_server_key}"
destination = "$HOME/etcd-server.key"
}
provisioner "file" {
content = "${module.bootkube.etcd_peer_cert}"
destination = "$HOME/etcd-peer.crt"
}
provisioner "file" {
content = "${module.bootkube.etcd_peer_key}"
destination = "$HOME/etcd-peer.key"
}
provisioner "remote-exec" {
inline = [
"sudo mkdir -p /etc/ssl/etcd/etcd",
"sudo mv etcd-client* /etc/ssl/etcd/",
"sudo cp /etc/ssl/etcd/etcd-client-ca.crt /etc/ssl/etcd/etcd/server-ca.crt",
"sudo mv etcd-server.crt /etc/ssl/etcd/etcd/server.crt",
"sudo mv etcd-server.key /etc/ssl/etcd/etcd/server.key",
"sudo cp /etc/ssl/etcd/etcd-client-ca.crt /etc/ssl/etcd/etcd/peer-ca.crt",
"sudo mv etcd-peer.crt /etc/ssl/etcd/etcd/peer.crt",
"sudo mv etcd-peer.key /etc/ssl/etcd/etcd/peer.key",
]
}
}
# Secure copy bootkube assets to ONE controller and start bootkube to perform
# one-time self-hosted cluster bootstrapping.
resource "null_resource" "bootkube-start" {
depends_on = [
"null_resource.copy-controller-secrets",
"module.workers",
"google_dns_record_set.apiserver",
]
connection {
type = "ssh"
host = "${element(local.controllers_ipv4_public, 0)}"
user = "fedora"
timeout = "15m"
}
provisioner "file" {
source = "${var.asset_dir}"
destination = "$HOME/assets"
}
provisioner "remote-exec" {
inline = [
"while [ ! -f /var/lib/cloud/instance/boot-finished ]; do sleep 4; done",
"sudo mv $HOME/assets /var/bootkube",
"sudo systemctl start bootkube",
]
}
}

View File

@ -0,0 +1,104 @@
variable "cluster_name" {
type = "string"
description = "Unique cluster name (prepended to dns_zone)"
}
# Google Cloud
variable "region" {
type = "string"
description = "Google Cloud Region (e.g. us-central1, see `gcloud compute regions list`)"
}
variable "dns_zone" {
type = "string"
description = "Google Cloud DNS Zone (e.g. google-cloud.example.com)"
}
variable "dns_zone_name" {
type = "string"
description = "Google Cloud DNS Zone name (e.g. example-zone)"
}
# instances
variable "controller_count" {
type = "string"
default = "1"
description = "Number of controllers (i.e. masters)"
}
variable "worker_count" {
type = "string"
default = "1"
description = "Number of workers"
}
variable "controller_type" {
type = "string"
default = "n1-standard-1"
description = "Machine type for controllers (see `gcloud compute machine-types list`)"
}
variable "worker_type" {
type = "string"
default = "n1-standard-1"
description = "Machine type for controllers (see `gcloud compute machine-types list`)"
}
variable "os_image" {
type = "string"
description = "Custom Fedora Atomic image"
}
variable "disk_size" {
type = "string"
default = "40"
description = "Size of the disk in GB"
}
variable "worker_preemptible" {
type = "string"
default = "false"
description = "If enabled, Compute Engine will terminate workers randomly within 24 hours"
}
# configuration
variable "ssh_authorized_key" {
type = "string"
description = "SSH public key for user 'fedora'"
}
variable "asset_dir" {
description = "Path to a directory where generated assets should be placed (contains secrets)"
type = "string"
}
variable "networking" {
description = "Choice of networking provider (flannel or calico)"
type = "string"
default = "calico"
}
variable "pod_cidr" {
description = "CIDR IPv4 range to assign Kubernetes pods"
type = "string"
default = "10.2.0.0/16"
}
variable "service_cidr" {
description = <<EOD
CIDR IPv4 range to assign Kubernetes services.
The 1st IP will be reserved for kube_apiserver, the 10th IP will be reserved for kube-dns.
EOD
type = "string"
default = "10.3.0.0/16"
}
variable "cluster_domain_suffix" {
description = "Queries for domains with the suffix will be answered by kube-dns. Default is cluster.local (e.g. foo.default.svc.cluster.local) "
type = "string"
default = "cluster.local"
}

View File

@ -0,0 +1,20 @@
module "workers" {
source = "workers"
name = "${var.cluster_name}"
cluster_name = "${var.cluster_name}"
# GCE
region = "${var.region}"
network = "${google_compute_network.network.name}"
count = "${var.worker_count}"
machine_type = "${var.worker_type}"
os_image = "${var.os_image}"
disk_size = "${var.disk_size}"
preemptible = "${var.worker_preemptible}"
# configuration
kubeconfig = "${module.bootkube.kubeconfig}"
ssh_authorized_key = "${var.ssh_authorized_key}"
service_cidr = "${var.service_cidr}"
cluster_domain_suffix = "${var.cluster_domain_suffix}"
}

View File

@ -0,0 +1,72 @@
#cloud-config
write_files:
- path: /etc/systemd/system/cloud-metadata.service
content: |
[Unit]
Description=Cloud metadata agent
[Service]
Type=oneshot
Environment=OUTPUT=/run/metadata/cloud
ExecStart=/usr/bin/mkdir -p /run/metadata
ExecStart=/usr/bin/bash -c 'echo "HOSTNAME_OVERRIDE=$(curl\
-H "Metadata-Flavor: Google"\
--url http://metadata.google.internal/computeMetadata/v1/instance/hostname\
--retry 10)" > $${OUTPUT}'
[Install]
WantedBy=multi-user.target
- path: /etc/systemd/system/kubelet.service.d/10-typhoon.conf
content: |
[Unit]
Requires=cloud-metadata.service
After=cloud-metadata.service
Wants=rpc-statd.service
[Service]
ExecStartPre=/bin/mkdir -p /opt/cni/bin
ExecStartPre=/bin/mkdir -p /etc/kubernetes/manifests
ExecStartPre=/bin/mkdir -p /etc/kubernetes/cni/net.d
ExecStartPre=/bin/mkdir -p /var/lib/cni
ExecStartPre=/bin/mkdir -p /var/lib/kubelet/volumeplugins
ExecStartPre=/usr/bin/bash -c "grep 'certificate-authority-data' /etc/kubernetes/kubeconfig | awk '{print $2}' | base64 -d > /etc/kubernetes/ca.crt"
Restart=always
RestartSec=10
- path: /etc/kubernetes/kubelet.conf
content: |
ARGS="--allow-privileged \
--anonymous-auth=false \
--client-ca-file=/etc/kubernetes/ca.crt \
--cluster_dns=${k8s_dns_service_ip} \
--cluster_domain=${cluster_domain_suffix} \
--cni-conf-dir=/etc/kubernetes/cni/net.d \
--exit-on-lock-contention \
--kubeconfig=/etc/kubernetes/kubeconfig \
--lock-file=/var/run/lock/kubelet.lock \
--network-plugin=cni \
--node-labels=node-role.kubernetes.io/node \
--pod-manifest-path=/etc/kubernetes/manifests \
--volume-plugin-dir=/var/lib/kubelet/volumeplugins"
- path: /etc/kubernetes/kubeconfig
permissions: '0644'
content: |
${kubeconfig}
- path: /etc/selinux/config
owner: root:root
permissions: '0644'
content: |
SELINUX=permissive
SELINUXTYPE=targeted
bootcmd:
- [setenforce, Permissive]
- [systemctl, disable, firewalld, --now]
runcmd:
- [systemctl, daemon-reload]
- [systemctl, enable, cloud-metadata.service]
- "atomic install --system --name=kubelet quay.io/dghubble/kubelet:8767d4433f7c5a38c55edf6e682efb53fcd06f61"
- [systemctl, start, --no-block, kubelet.service]
users:
- default
- name: fedora
gecos: Fedora Admin
sudo: ALL=(ALL) NOPASSWD:ALL
groups: wheel,adm,systemd-journal,docker
ssh-authorized-keys:
- "${ssh_authorized_key}"

View File

@ -0,0 +1,45 @@
# Static IPv4 address for the Network Load Balancer
resource "google_compute_address" "ingress-ip" {
name = "${var.name}-ingress-ip"
}
# Network Load Balancer (i.e. forwarding rules)
resource "google_compute_forwarding_rule" "worker-http-lb" {
name = "${var.name}-worker-http-rule"
ip_address = "${google_compute_address.ingress-ip.address}"
port_range = "80"
target = "${google_compute_target_pool.workers.self_link}"
}
resource "google_compute_forwarding_rule" "worker-https-lb" {
name = "${var.name}-worker-https-rule"
ip_address = "${google_compute_address.ingress-ip.address}"
port_range = "443"
target = "${google_compute_target_pool.workers.self_link}"
}
# Network Load Balancer target pool of instances.
resource "google_compute_target_pool" "workers" {
name = "${var.name}-worker-pool"
health_checks = [
"${google_compute_http_health_check.ingress.name}",
]
session_affinity = "NONE"
}
# Ingress HTTP Health Check
resource "google_compute_http_health_check" "ingress" {
name = "${var.name}-ingress-health"
description = "Health check Ingress controller health host port"
timeout_sec = 5
check_interval_sec = 5
healthy_threshold = 2
unhealthy_threshold = 4
port = 10254
request_path = "/healthz"
}

View File

@ -0,0 +1,3 @@
output "ingress_static_ip" {
value = "${google_compute_address.ingress-ip.address}"
}

View File

@ -0,0 +1,94 @@
variable "name" {
type = "string"
description = "Unique name for the worker pool"
}
variable "cluster_name" {
type = "string"
description = "Must be set to `cluster_name of cluster`"
}
# Google Cloud
variable "region" {
type = "string"
description = "Must be set to `region` of cluster"
}
variable "network" {
type = "string"
description = "Must be set to `network_name` output by cluster"
}
# instances
variable "count" {
type = "string"
default = "1"
description = "Number of worker compute instances the instance group should manage"
}
variable "machine_type" {
type = "string"
default = "n1-standard-1"
description = "Machine type for compute instances (e.g. gcloud compute machine-types list)"
}
variable "os_image" {
type = "string"
description = "Custom Fedora Atomic image"
}
variable "disk_size" {
type = "string"
default = "40"
description = "Size of the disk in GB"
}
variable "preemptible" {
type = "string"
default = "false"
description = "If enabled, Compute Engine will terminate instances randomly within 24 hours"
}
# configuration
variable "kubeconfig" {
type = "string"
description = "Must be set to `kubeconfig` output by cluster"
}
variable "ssh_authorized_key" {
type = "string"
description = "SSH public key for user 'fedora'"
}
variable "service_cidr" {
description = <<EOD
CIDR IPv4 range to assign Kubernetes services.
The 1st IP will be reserved for kube_apiserver, the 10th IP will be reserved for kube-dns.
EOD
type = "string"
default = "10.3.0.0/16"
}
variable "cluster_domain_suffix" {
description = "Queries for domains with the suffix will be answered by kube-dns. Default is cluster.local (e.g. foo.default.svc.cluster.local) "
type = "string"
default = "cluster.local"
}
# unofficial, undocumented, unsupported, temporary
variable "accelerator_type" {
type = "string"
default = ""
description = "Google Compute Engine accelerator type (e.g. nvidia-tesla-k80, see gcloud compute accelerator-types list)"
}
variable "accelerator_count" {
type = "string"
default = "0"
description = "Number of compute engine accelerators"
}

View File

@ -0,0 +1,72 @@
# Regional managed instance group maintains a homogeneous set of workers that
# span the zones in the region.
resource "google_compute_region_instance_group_manager" "workers" {
name = "${var.name}-worker-group"
description = "Compute instance group of ${var.name} workers"
# instance name prefix for instances in the group
base_instance_name = "${var.name}-worker"
instance_template = "${google_compute_instance_template.worker.self_link}"
region = "${var.region}"
target_size = "${var.count}"
# target pool to which instances in the group should be added
target_pools = [
"${google_compute_target_pool.workers.self_link}",
]
}
resource "google_compute_instance_template" "worker" {
name_prefix = "${var.name}-worker-"
description = "Worker Instance template"
machine_type = "${var.machine_type}"
metadata {
user-data = "${data.template_file.worker-cloudinit.rendered}"
}
scheduling {
automatic_restart = "${var.preemptible ? false : true}"
preemptible = "${var.preemptible}"
}
disk {
auto_delete = true
boot = true
source_image = "${var.os_image}"
disk_size_gb = "${var.disk_size}"
}
network_interface {
network = "${var.network}"
# Ephemeral external IP
access_config = {}
}
can_ip_forward = true
tags = ["worker", "${var.cluster_name}-worker", "${var.name}-worker"]
guest_accelerator {
count = "${var.accelerator_count}"
type = "${var.accelerator_type}"
}
lifecycle {
# To update an Instance Template, Terraform should replace the existing resource
create_before_destroy = true
}
}
# Worker Cloud-Init
data "template_file" "worker-cloudinit" {
template = "${file("${path.module}/cloudinit/worker.yaml.tmpl")}"
vars = {
kubeconfig = "${indent(6, var.kubeconfig)}"
ssh_authorized_key = "${var.ssh_authorized_key}"
k8s_dns_service_ip = "${cidrhost(var.service_cidr, 10)}"
cluster_domain_suffix = "${var.cluster_domain_suffix}"
}
}