Евгений Штольц - Облачная экосистема

essh@kubernetes-master:~/node-cluster$ exec -l $SHELL

essh@kubernetes-master:~/node-cluster$ gcloud init

Выберем проект:

You are logged in as: [esschtolts@gmail.com].

Pick cloud project to use:

[1] agile-aleph-203917

[2] node-cluster-243923

[3] essch

[4] Create a new project

Please enter numeric choice or text value (must exactly match list

item):

Please enter a value between 1 and 4, or a value present in the list: 2

Your current project has been set to: [node-cluster-243923].

Выберем зону:

[50] europe-north1-a

Did not print [12] options.

Too many options [62]. Enter "list" at prompt to print choices fully.

Please enter numeric choice or text value (must exactly match list

item):

Please enter a value between 1 and 62, or a value present in the list: 50

essh@kubernetes-master:~/node-cluster$ PROJECT_I="node-cluster-243923"

essh@kubernetes-master:~/node-cluster$ echo $PROJECT_I

node-cluster-243923

essh@kubernetes-master:~/node-cluster$ export GOOGLE_APPLICATION_CREDENTIALS=$HOME/node-cluster/kubernetes_key.json

essh@kubernetes-master:~/node-cluster$ sudo docker-machine create –driver google –google-project $PROJECT_ID vm01

sudo export GOOGLE_APPLICATION_CREDENTIALS=$HOME/node-cluster/kubernetes_key.json docker-machine create –driver google –google-project $PROJECT_ID vm01

// https://docs.docker.com/machine/drivers/gce/

// https://github.com/docker/machine/issues/4722

essh@kubernetes-master:~/node-cluster$ gcloud config list

[compute]

region = europe-north1

zone = europe-north1-a

[core]

account = esschtolts@gmail.com

disable_usage_reporting = False

project = node-cluster-243923

Your active configuration is: [default]

Добавим копирование файла и выполнение скрипта:

essh@kubernetes-master:~/node-cluster$ cat main.tf

provider "google" {

credentials = "${file("kubernetes_key.json")}"

project = "node-cluster-243923"

region = "europe-north1"

}

resource "google_compute_address" "static-ip-address" {

name = "static-ip-address"

}

resource "google_compute_instance" "cluster" {

name = "cluster"

zone = "europe-north1-a"

machine_type = "f1-micro"

boot_disk {

initialize_params {

image = "debian-cloud/debian-9"

}

}

metadata = {

ssh-keys = "essh:${file("./node-cluster.pub")}"

}

network_interface {

network = "default"

access_config {

nat_ip = "${google_compute_address.static-ip-address.address}"

}

}

}

resource "null_resource" "cluster" {

triggers = {

cluster_instance_ids = "${join(",", google_compute_instance.cluster.*.id)}"

}

connection {

host = "${google_compute_address.static-ip-address.address}"

type = "ssh"

user = "essh"

timeout = "2m"

private_key = "${file("~/node-cluster/node-cluster")}"

# agent = "false"

}

provisioner "file" {

source = "client.js"

destination = "~/client.js"

}

provisioner "remote-exec" {

inline = [

"cd ~ && echo 1 > test.txt"

]

}

essh@kubernetes-master:~/node-cluster$ sudo ./terraform apply

google_compute_address.static-ip-address: Creating…

google_compute_address.static-ip-address: Creation complete after 5s [id=node-cluster-243923/europe-north1/static-ip-address]

google_compute_instance.cluster: Creating…

google_compute_instance.cluster: Still creating… [10s elapsed]

google_compute_instance.cluster: Creation complete after 12s [id=cluster]

null_resource.cluster: Creating…

null_resource.cluster: Provisioning with 'file'…

null_resource.cluster: Provisioning with 'remote-exec'…

null_resource.cluster (remote-exec): Connecting to remote host via SSH…

null_resource.cluster (remote-exec): Host: 35.228.82.222

null_resource.cluster (remote-exec): User: essh

null_resource.cluster (remote-exec): Password: false

null_resource.cluster (remote-exec): Private key: true

null_resource.cluster (remote-exec): Certificate: false

null_resource.cluster (remote-exec): SSH Agent: false

null_resource.cluster (remote-exec): Checking Host Key: false

null_resource.cluster (remote-exec): Connected!

null_resource.cluster: Creation complete after 7s [id=816586071607403364]

Apply complete! Resources: 3 added, 0 changed, 0 destroyed.

esschtolts@cluster:~$ ls /home/essh/

client.js test.txt

[sudo] password for essh:

google_compute_address.static-ip-address: Refreshing state… [id=node-cluster-243923/europe-north1/static-ip-address]

google_compute_instance.cluster: Refreshing state… [id=cluster]

null_resource.cluster: Refreshing state… [id=816586071607403364]

Enter a value: yes

null_resource.cluster: Destroying… [id=816586071607403364]

null_resource.cluster: Destruction complete after 0s

google_compute_instance.cluster: Destroying… [id=cluster]

google_compute_instance.cluster: Still destroying… [id=cluster, 10s elapsed]

google_compute_instance.cluster: Still destroying… [id=cluster, 20s elapsed]

google_compute_instance.cluster: Destruction complete after 27s

google_compute_address.static-ip-address: Destroying… [id=node-cluster-243923/europe-north1/static-ip-address]

google_compute_address.static-ip-address: Destruction complete after 8s

Для деплоя всего проекта можно добавить его в репозиторий, а загружать его в виртуальную машину будем через копирование установочного скрипта на эту виртуальную машину с последующим его запуском:

Переходим к Kubernetes

В минимальном варианте создание кластера из трёх нод выглядит примерно так:

essh@kubernetes-master:~/node-cluster/Kubernetes$ cat main.tf

provider "google" {

credentials = "${file("../kubernetes_key.json")}"

project = "node-cluster-243923"

region = "europe-north1"

}

resource "google_container_cluster" "node-ks" {

name = "node-ks"

location = "europe-north1-a"

initial_node_count = 3

}

essh@kubernetes-master:~/node-cluster/Kubernetes$ sudo ../terraform init

essh@kubernetes-master:~/node-cluster/Kubernetes$ sudo ../terraform apply

Кластер создался за 2:15, а после того, как я добавил europe-north1-a две дополнительные зоны europe-north1 -b, europe-north1-c и установили количество создаваемых инстансев в зоне в один, кластер создался за 3:13 секунде, потому что для более высокой доступности ноды были созданы в разных дата-центрах: europe-north1-a, europe-north1-b, europe-north1-c:

provider "google" {

credentials = "${file("../kubernetes_key.json")}"

project = "node-cluster-243923"

region = "europe-north1"

}

resource "google_container_cluster" "node-ks" {

name = "node-ks"

location = "europe-north1-a"

node_locations = ["europe-north1-b", "europe-north1-c"]

initial_node_count = 1

}

Теперь разделим наш кластер на два: управляющий кластер с Kubernetes и кластер для наших POD. Все кластера будет распределены по трём дата – центрам. Кластер для наших POD может авто масштабироваться под нагрузкой до 2 на каждой зоне (с трёх до шести в общем):