Add support for worker pools on google-cloud

* Set defaults for internal worker module's count,
machine_type, and os_image
* Allow "pools" of homogeneous workers to be created
using the google-cloud/kubernetes/workers module

CHANGES.md

@@ -24,6 +24,7 @@ Notable changes between versions.
 
 #### Google Cloud
 
+* Add support for "worker pools" - groups of homogeneous workers joined to an existing cluster ([#148](https://github.com/poseidon/typhoon/pull/148))
 * Add kubelet `--volume-plugin-dir` flag to allow flexvolume plugins ([#142](https://github.com/poseidon/typhoon/pull/142))
 * Add `kubeconfig` variable to `controllers` and `workers` submodules ([#147](https://github.com/poseidon/typhoon/pull/147))
 * Remove `kubeconfig_*` variables from `controllers` and `workers` submodules ([#147](https://github.com/poseidon/typhoon/pull/147))

docs/advanced/overview.md

@@ -0,0 +1,6 @@
+# Advanced
+
+Typhoon clusters offer several advanced features for skilled users.
+
+* [Customization](customization.md)
+* [Worker Pools](worker-pools.md)

docs/advanced/worker-pools.md

@@ -0,0 +1,71 @@
+# Worker Pools
+
+Typhoon can create "worker pools", groups of homogeneous workers that are part of an existing cluster. For example, you may wish to augment a Kubernetes cluster with groups of workers with a different machine type, larger disks, or preemptibility.
+
+## Google Cloud
+
+Create a cluster following the Google Cloud [tutorial](../google-cloud.md#cluster). Then define a worker pool using the internal `workers` Terraform module.
+
+```tf
+module "yavin-worker-pool" {
+  source = "git::https://github.com/poseidon/typhoon//google-cloud/container-linux/kubernetes/workers?ref=v1.9.4"
+
+  # Google Cloud
+  network      = "${module.google-cloud-yavin.network_name}"
+  region       = "us-central1"
+  count        = 2
+  machine_type = "n1-standard-16"
+  preemptible  = true
+
+  cluster_name = "yavin-16x"
+  ssh_authorized_key = "${var.ssh_authorized_key}"
+
+  kubeconfig = "${module.google-cloud-yavin.kubeconfig}"
+}
+```
+
+Apply the change.
+
+```
+terraform apply
+```
+
+Verify a managed instance group of workers joins the cluster within a few minutes.
+
+```
+$ kubectl get nodes
+NAME                                             STATUS   AGE    VERSION
+yavin-controller-0.c.example-com.internal        Ready    6m     v1.9.3
+yavin-worker-jrbf.c.example-com.internal         Ready    5m     v1.9.3
+yavin-worker-mzdm.c.example-com.internal         Ready    5m     v1.9.3
+yavin-16x-worker-jrbf.c.example-com.internal     Ready    3m     v1.9.3
+yavin-16x-worker-mzdm.c.example-com.internal     Ready    3m     v1.9.3
+```
+
+### Variables
+
+The Google Cloud internal `workers` module supports a number of [variables](https://github.com/poseidon/typhoon/blob/master/google-cloud/container-linux/kubernetes/workers/variables.tf).
+
+#### Required
+
+| Name | Description | Example |
+|:-----|:------------|:--------|
+| cluster_name | Unique name | "yavin-worker-pool" |
+| region | Must match region of cluster | "us-central1" |
+| network | Must match network name output by cluster | "${module.cluster.network_name}" |
+| ssh_authorized_key | SSH public key for ~/.ssh_authorized_keys | "ssh-rsa AAAAB3NZ..." |
+
+#### Optional
+
+| Name | Description | Default | Example |
+|:-----|:------------|:--------|:--------|
+| count | Number of workers | 1 | 3 |
+| machine_type | Machine type for compute instances | "n1-standard-1" | See below |
+| os_image | OS image for compute instances | "coreos-stable" | "coreos-alpha" |
+| disk_size | Size of the disk in GB | 40 | 100 |
+| preemptible | If enabled, Compute Engine will terminate instances randomly within 24 hours | false | true |
+| service_cidr | Must match service_cidr of cluster | "10.3.0.0/16" | "10.3.0.0/24" |
+| cluster_domain_suffix | Must match domain suffix of cluster | "cluster.local" | "k8s.example.com" |
+
+Check the list of valid [machine types](https://cloud.google.com/compute/docs/machine-types).
+

docs/google-cloud.md

@@ -100,11 +100,11 @@ module "google-cloud-yavin" {
   source = "git::https://github.com/poseidon/typhoon//google-cloud/container-linux/kubernetes?ref=v1.9.3"
   
   providers = {
-    google = "google.default"
+    google   = "google.default"
-    local = "local.default"
+    local    = "local.default"
-    null = "null.default"
+    null     = "null.default"
     template = "template.default"
-    tls = "tls.default"
+    tls      = "tls.default"
   }
 
   # Google Cloud
@@ -257,7 +257,7 @@ resource "google_dns_managed_zone" "zone-for-clusters" {
 | machine_type | Machine type for compute instances | "n1-standard-1" | See below |
 | controller_count | Number of controllers (i.e. masters) | 1 | 1 |
 | worker_count | Number of workers | 1 | 3 |
-| worker_preemptible | If enabled, Compute Engine will terminate controllers randomly within 24 hours | false | true |
+| worker_preemptible | If enabled, Compute Engine will terminate workers randomly within 24 hours | false | true |
 | networking | Choice of networking provider | "calico" | "calico" or "flannel" |
 | pod_cidr | CIDR range to assign to Kubernetes pods | "10.2.0.0/16" | "10.22.0.0/16" |
 | service_cidr | CIDR range to assign to Kubernetes services | "10.3.0.0/16" | "10.3.0.0/24" |

google-cloud/container-linux/kubernetes/cluster.tf

@@ -1,7 +1,6 @@
 module "controllers" {
-  source             = "controllers"
+  source       = "controllers"
-  cluster_name       = "${var.cluster_name}"
+  cluster_name = "${var.cluster_name}"
-  ssh_authorized_key = "${var.ssh_authorized_key}"
 
   # GCE
   network       = "${google_compute_network.network.name}"
@@ -14,15 +13,15 @@ module "controllers" {
 
   # configuration
   networking            = "${var.networking}"
+  kubeconfig            = "${module.bootkube.kubeconfig}"
+  ssh_authorized_key    = "${var.ssh_authorized_key}"
   service_cidr          = "${var.service_cidr}"
   cluster_domain_suffix = "${var.cluster_domain_suffix}"
-  kubeconfig            = "${module.bootkube.kubeconfig}"
 }
 
 module "workers" {
-  source             = "workers"
+  source       = "workers"
-  cluster_name       = "${var.cluster_name}"
+  cluster_name = "${var.cluster_name}"
-  ssh_authorized_key = "${var.ssh_authorized_key}"
 
   # GCE
   network      = "${google_compute_network.network.name}"
@@ -33,7 +32,8 @@ module "workers" {
   preemptible  = "${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}"
-  kubeconfig            = "${module.bootkube.kubeconfig}"
 }

google-cloud/container-linux/kubernetes/outputs.tf

@@ -13,3 +13,7 @@ output "network_name" {
 output "network_self_link" {
   value = "${google_compute_network.network.self_link}"
 }
+
+output "kubeconfig" {
+  value = "${module.bootkube.kubeconfig}"
+}

google-cloud/container-linux/kubernetes/workers/variables.tf

@@ -17,6 +17,7 @@ variable "network" {
 
 variable "count" {
   type        = "string"
+  default     = "1"
   description = "Number of worker compute instances the instance group should manage"
 }
 
@@ -27,11 +28,13 @@ variable "region" {
 
 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"
+  default     = "coreos-stable"
   description = "OS image from which to initialize the disk (e.g. gcloud compute images list)"
 }
 

mkdocs.yml

@@ -58,4 +58,6 @@ pages:
     - 'Performance': 'topics/performance.md'
   - 'FAQ': 'faq.md'
   - 'Advanced':
+    - 'Overview': 'advanced/overview.md'
     - 'Customization': 'advanced/customization.md'
+    - 'Worker Pools': 'advanced/worker-pools.md'