前言
Terraform 是一种开源基础设施及代码(IaC)的工具,可提供统一的 CLI(命令行接口)工作流来治理数百个云服务,将云 API 编码为申明性的配置文件进行治理。
本文创立一个治理 AWS Lightsail 实例的例子来入门 Terraform 的应用。
装置 Terraform CLI
要应用 Terramform,首先要在本地零碎装置 Terraform 命令行工具。HashiCorp 提供了预编译好的二进制散发包,能够通过(https://www.terraform.com/dow…) 间接下载相应平台的二进制包,解压后放到相应的执行门路。也能够通过一些软件包管理工具装置,例如在 Linux/OS X 上通过 LinuxBrew/HomeBrew 进行装置,在 Windows 上通过 Chocolatey 进行装置。
这里咱们示例在 Linux 上是应用 LinuxBrew 进行装置
> brew install terraform
装置实现后,能够查看其版本
❯ terraform -version
Terraform v1.0.11
on linux_amd64
应用 -help
查看其可用命令,装置胜利后,咱们就能够应用 Terraform 来创立相应的基础设施我的项目了。
AWS 账号筹备
本文将通过创立一个治理 AWS Lightsial 实例的我的项目来尝试 Terraform,因而须要一个 AWS 账号,以及在本地环境装置和配置好 AWS CLI 工具的拜访凭据。
装置和配置 AWS CLI,请参考其文档 (https://docs.aws.amazon.com/c…)。配置实现之后,能够在本地命令行终端拜访相应的 AWS 资源。
创立并初始化 Terraform 我的项目
Terraform 在本地通过文件夹来治理一个基础设施我的项目的申明性代码,例如咱们在本地创立一个文件夹
> mkdir mylightsail
> cd mylightsail/
进入文件夹后,创立一个以 .tf 作为后缀的文件,例如 main.tf
> touch main.tf
而后应用编辑器关上文件进行编辑,写入以下代码块
terraform {
required_providers {
aws = {
source = "hashicorp/aws"
version = "~> 3.65"
}
}
}
# Configure the AWS Provider
provider "aws" {region = "ap-southeast-1"}
其中 terraform/required_providers
块定义了该我的项目须要的 Provider,Terraform 是通过不同的 Provider 来治理相应的基础设施资源的,可到 (https://registry.terraform.io) 来查找须要的 Providers,例如 GCP,Azure 以及阿里云等的 Providers。这里因为咱们要治理的资源是 AWS Lightsail 实例,所以应用了 Hashicorp 官网提供的 hashicorp/aws。
provider "aws"
局部配置了该 Provider 的一些可选项,例如这里配置了区域为 ap-southeast-1,因而请确保下面配置的 AWS 拜访凭据可能操作该区域的资源。
也就是这里咱们定义了咱们须要应用的 Provider 及其相应的选项配置,接下来咱们须要应用 terraform init
命令来初始化我的项目
> terraform init
Initializing provider plugins...
...
Terraform has been successfully initialized!
初始化将会从 (https://registry.terraform.io/) 下载相应的 Provider 插件到 .terramorm/providers/ 目录,供接下來的命令应用。
同时,会生成一个 .terraform.lock.hcl 的文件来记录应用的具体 Provider 版本,其性能相似于 NPM 的 package-lock 文件,可将其提交到代码版本治理仓库,其余合作的成员就能够放弃应用雷同的插件版本。
创立基础设施资源
实现我的项目的初始化之后,咱们就能够编写须要创立的资源申明性配置,能够间接将相应的配置写入 main.tf 文件,也能够另外创立新的以 .tf
作为后缀的文件,这里咱们创立一个新的名为 resources.tf 的文件,并写入咱们须要的资源的定义
## LightSail Resources
resource "aws_lightsail_static_ip" "Example-sig-ip" {name = "Example-EIP"}
resource "aws_lightsail_instance" "Example-sig" {
name = "Example-Sig"
availability_zone = "ap-southeast-1c"
blueprint_id = "ubuntu_20_04"
bundle_id = "nano_2_0"
key_pair_name = "LightsailDefaultKeyPair"
tags = {Example = ""}
}
resource "aws_lightsail_static_ip_attachment" "Example-sig-ip-attache" {
static_ip_name = aws_lightsail_static_ip.Example-sig-ip.id
instance_name = aws_lightsail_instance.Example-sig.id
}
resource "aws_lightsail_instance_public_ports" "Example-sig-public-ports" {
instance_name = aws_lightsail_instance.Example-sig.name
port_info {
protocol = "tcp"
from_port = 0
to_port = 65535
cidrs = ["0.0.0.0/0"]
}
port_info {
protocol = "udp"
from_port = 0
to_port = 65535
cidrs = ["0.0.0.0/0"]
}
}
定义资源的格局为 resource "[provider_resource _type]" "resource_name"
,第一个参数为相应 Provider 反对的资源类型名称,第二个参数为本人定义的资源名称(可用于其余资源援用)。例如,咱们首先定义了一个 Lightsail 的动态 IP 资源,其中参数 name
指定了 AWS 资源的名称。
下面的定义中,咱们申明了以下资源
- 一个 Lightsail 动态 IP 地址
- 一个 Lightsail 计算实例,并绑定名为
LightsailDefaultKeyPair
的 SSH 密钥 - 动态 IP 地址和计算实例的绑定
- 实例的凋谢的网络端口组(相似于 AWS EC2 实例的平安组定义)
保留文件之后,咱们能够应用 terraform fmt
命令来格式化文件格式,以及 terraform validate
来查看是否有语法错误。
定义好咱们想要的资源之后,咱们先通过命令 terraform plan
命令来执行打算,查看具体的执行更改(plan 不会实际操作相应的资源)
❯ terraform plan
Terraform used the selected providers to generate the following execution plan. Resource actions are indicated with the following symbols:
+ create
Terraform will perform the following actions:
# aws_lightsail_instance.Example-sig will be created
+ resource "aws_lightsail_instance" "Example-sig" {+ arn = (known after apply)
+ availability_zone = "ap-southeast-1c"
+ blueprint_id = "ubuntu_20_04"
+ bundle_id = "nano_2_0"
+ cpu_count = (known after apply)
+ created_at = (known after apply)
+ id = (known after apply)
+ ipv6_address = (known after apply)
+ ipv6_addresses = (known after apply)
+ is_static_ip = (known after apply)
+ key_pair_name = "LightsailDefaultKeyPair"
+ name = "Example-Sig"
+ private_ip_address = (known after apply)
+ public_ip_address = (known after apply)
+ ram_size = (known after apply)
+ tags = {+ "Example" = ""}
+ tags_all = {+ "Example" = (known after apply)
}
+ username = (known after apply)
}
# aws_lightsail_instance_public_ports.Example-sig-public-ports will be created
+ resource "aws_lightsail_instance_public_ports" "Example-sig-public-ports" {+ id = (known after apply)
+ instance_name = "Example-Sig"
+ port_info {
+ cidrs = [+ "0.0.0.0/0",]
+ from_port = 0
+ protocol = "tcp"
+ to_port = 65535
}
+ port_info {
+ cidrs = [+ "0.0.0.0/0",]
+ from_port = 0
+ protocol = "udp"
+ to_port = 65535
}
}
# aws_lightsail_static_ip.Example-sig-ip will be created
+ resource "aws_lightsail_static_ip" "Example-sig-ip" {+ arn = (known after apply)
+ id = (known after apply)
+ ip_address = (known after apply)
+ name = "Example-EIP"
+ support_code = (known after apply)
}
# aws_lightsail_static_ip_attachment.Example-sig-ip-attache will be created
+ resource "aws_lightsail_static_ip_attachment" "Example-sig-ip-attache" {+ id = (known after apply)
+ instance_name = (known after apply)
+ ip_address = (known after apply)
+ static_ip_name = (known after apply)
}
Plan: 4 to add, 0 to change, 0 to destroy.
+
示意将要减少的资源,(know after apply)的意思是要在具体只想(apply)之后,AWS 依据定义创立相应资源之后才会返回的具体值。接下来就能够应用 terraform apply
来具体执行操作了,执行胜利之后,会生成 .terraform/terraform.state 文件来记录执行后的资源状态,也能够通过命令 terraform show
命令来查看
❯ terraform show
# aws_lightsail_instance.Example-sig:
resource "aws_lightsail_instance" "Example-sig" {
arn = "arn:aws:lightsail:ap-southeast-1:090767794770:Instance/21cb0ea5-e814-4307-8606-01348d98be15"
availability_zone = "ap-southeast-1c"
blueprint_id = "ubuntu_20_04"
bundle_id = "nano_2_0"
cpu_count = 1
created_at = "2021-11-08T05:49:05Z"
id = "Example-Sig"
ipv6_address = "2406:da18:8ae:4b02:1f2:4ff1:daa1:6a8c"
ipv6_addresses = ["2406:da18:8ae:4b02:1f2:4ff1:daa1:6a8c",]
is_static_ip = true
key_pair_name = "LightsailDefaultKeyPair"
name = "Example-Sig"
private_ip_address = "172.26.45.249"
public_ip_address = "54.220.33.133"
ram_size = 0.5
tags = {"Example" = ""}
tags_all = {"Example" = ""}
username = "ubuntu"
}
# aws_lightsail_instance_public_ports.Example-sig-public-ports:
resource "aws_lightsail_instance_public_ports" "Example-sig-public-ports" {
id = "Example-Sig-987241840"
instance_name = "Example-Sig"
port_info {
cidrs = ["0.0.0.0/0",]
from_port = 0
protocol = "tcp"
to_port = 65535
}
port_info {
cidrs = ["0.0.0.0/0",]
from_port = 0
protocol = "udp"
to_port = 65535
}
}
# aws_lightsail_static_ip.Example-sig-ip:
resource "aws_lightsail_static_ip" "Example-sig-ip" {
arn = "arn:aws:lightsail:ap-southeast-1:090767794770:StaticIp/3f0298e0-efeb-4429-9574-156fef12a48f"
id = "Example-EIP"
ip_address = "54.220.33.133"
name = "Example-EIP"
support_code = "313963776615/54.220.33.133"
}
# aws_lightsail_static_ip_attachment.Example-sig-ip-attache:
resource "aws_lightsail_static_ip_attachment" "exmaple-sig-ip-attache" {
id = "Example-EIP"
instance_name = "Example-Sig"
ip_address = "54.220.33.133"
static_ip_name = "Example-EIP"
}
之后咱们就能够通过更新资源定义文件,而后执行 plan & apply
来更新云上的资源了,如果资源不须要了,咱们也能够通过 terraform destroy
一条命令来销毁所有资源。
总结
本文以治理简略的 AWS Lightsail 资源为例,展现了如何应用 Terraform 来治理云上的资源。通过一申明性的代码来定义咱们须要的资源,同时资源定义代码能够通过版本管理工具进行版本化治理,进一步实现 IaC 的工作形式。
这么咱们应用的是 Terraform 的本地执行模式,除了在本地执行之外,Terraform 还提供了执行后端(Backend)的性能,就能够将执行放到云上的一些执行环境,资源状态的治理也将会保护在 Backend 端,不便实现 CI/CD Pipeline,以及 GitOps 的形式。其中 Hashicorp 的 Terraform Cloud 就是一个 Backend。下一篇文章将示例将该我的项目放到 Terraform Cloud 下来执行。