乐趣区

关于腾讯云:Terraform系列二腾讯云CVM进一步相关玩法

jiezi

背景：

紧接 Terraform 系列一腾讯云 CVM 相干简略创立。筹备围绕着 cvm 先相熟一下根本的流程。比方：系统盘扩容，挂载数据盘，帐号密钥 ssh-key，绑定公网 ip. 钻研一下官网文档体验一下！

Terraform 系列二腾讯云 CVM 进一步相干玩法

1. 对于硬盘的操作

参照：https://registry.terraform.io/providers/tencentcloudstack/tencentcloud/latest/docs/resources/instance
默认的系统盘是 50G , 设置 system_disk_size = 100。批改系统盘为 100G，并减少了数据盘data_disks 配置（50G）

1. 批改 cvm.tf 配置文件

[root@zhangpeng terraform]# cat cvm.tf


resource "tencentcloud_instance" "cvm_almalinux" {
  instance_name = "cvm-almalinux"
  availability_zone = "ap-beijing-2"
  image_id = "img-q95tlc25"
  instance_type = "S2.MEDIUM2"
  system_disk_type = "CLOUD_PREMIUM"
  system_disk_size  = 100
  hostname = "cvm-almalinux"   
  data_disks {
    data_disk_type = "CLOUD_PREMIUM"
    data_disk_size = 50
    encrypt        = false
  }
  security_groups = ["${tencentcloud_security_group.sg_bj.id}"
  ]

  vpc_id = "${tencentcloud_vpc.vpc_bj.id}"
  subnet_id = "${tencentcloud_subnet.subnet_bj_02.id}"
  internet_max_bandwidth_out = 10
  count = 1
}

当然了我这里看文档的时候看到了 hostname 配置这里也增加了 hostname！恩 反正当初还没有公网 Ip 一步一步操作！

2. terraform plan

3. terraform apply


4. 验证

登陆腾讯云后盾找到对应 cvm 查看系统盘与数据盘

目测是失常的然而认真看了一眼信息，系统盘更换 服务器明码也从新进行了初始化 ：

故：到这里系统盘的扩容与数据盘的挂载实现了。然而没有能确认系统盘是间接扩容还是进行了系统盘硬盘的间接更换！这个前面再去钻研！

2. 创立公网 ip 并绑定 cvm

公网 ip 关上腾讯云 控制台 云服务器有个 公网 IP的选项看了下 url 叫 eip? 官网搜寻一下 eip：https://registry.terraform.io/providers/tencentcloudstack/tencentcloud/latest/docs/resources/eip? 然而没有看到这里有配置的阿 …. 网上搜寻了一下发现能够通过allocate_public_ip = true 开启 公网 IP！参照：http://www.panooo.com/Terraform_On_TencentCloud！
先这样搞一下吧！前面钻研一下如何新建一个 EIP 而后绑定 CVM？具体步骤如下：

1. 批改 cvm.tf

cat cvm.tf

resource "tencentcloud_instance" "cvm_almalinux" {
  instance_name = "cvm-almalinux"
  availability_zone = "ap-beijing-2"
  image_id = "img-q95tlc25"
  instance_type = "S2.MEDIUM2"
  system_disk_type = "CLOUD_PREMIUM"
  system_disk_size  = 100
  hostname = "cvm-almalinux"   
  allocate_public_ip = true
  data_disks {
    data_disk_type = "CLOUD_PREMIUM"
    data_disk_size = 50
    encrypt        = false
  }
  security_groups = ["${tencentcloud_security_group.sg_bj.id}"
  ]

  vpc_id = "${tencentcloud_vpc.vpc_bj.id}"
  subnet_id = "${tencentcloud_subnet.subnet_bj_02.id}"
  internet_max_bandwidth_out = 10
  count = 1
}

减少了 allocate_public_ip = true。另外也明确了 internet_max_bandwidth_out = 10 是限度带宽的配置

2. terraform plan

3. terraform apply


4. 验证

登陆后盾验证的确有了公网 ip 了

然而 ….. 这鬼货色每次都是新建吗？又收到了服务器创立，生成明码的短信提醒 ……. 请看上面的分析测试！

3.cvm 重建失去的论断

在下面步骤中 cvm 领有了公网的 Ip。ssh 登陆服务器先看一下：
系统盘 数据盘创立胜利，主机名 hostname 也设置胜利了！
先轻易生成一个文件，而后更改 cvm. tf 相干配置。确认一下在什么环境下 cvm 会重建！

touch zhangpeng.txt

1. 批改一下公网 ip 进口带宽测试一下？

将 i nternet_max_bandwidth_out = 10批改为internet_max_bandwidth_out = 15
cat cvm.tf

resource "tencentcloud_instance" "cvm_almalinux" {
  instance_name = "cvm-almalinux"
  availability_zone = "ap-beijing-2"
  image_id = "img-q95tlc25"
  instance_type = "S2.MEDIUM2"
  system_disk_type = "CLOUD_PREMIUM"
  system_disk_size  = 100
  hostname = "cvm-almalinux"   
  allocate_public_ip = true
  data_disks {
    data_disk_type = "CLOUD_PREMIUM"
    data_disk_size = 50
    encrypt        = false
  }

  security_groups = ["${tencentcloud_security_group.sg_bj.id}"
  ]

  vpc_id = "${tencentcloud_vpc.vpc_bj.id}"
  subnet_id = "${tencentcloud_subnet.subnet_bj_02.id}"
  internet_max_bandwidth_out = 15
  count = 1
}

仍旧是 terraform plan and terraform apply




未收到 CVM 重建信息, 原明码失常登陆。登陆服务器查看 zhangpeng.txt 存在！所以确认批改带宽配置不会触发 cvm 重建！

2. 批改系统盘与数据盘大小

两个的测试都放在一起了，首先是批改数据盘的大小：
data_disk_size = 50 批改为 data_disk_size = 100
cat cvm.tf

resource "tencentcloud_instance" "cvm_almalinux" {
  instance_name = "cvm-almalinux"
  availability_zone = "ap-beijing-2"
  image_id = "img-q95tlc25"
  instance_type = "S2.MEDIUM2"
  system_disk_type = "CLOUD_PREMIUM"
  system_disk_size  = 100
  hostname = "cvm-almalinux"   
  allocate_public_ip = true
  data_disks {
    data_disk_type = "CLOUD_PREMIUM"
    data_disk_size = 100
    encrypt        = false
  }

  security_groups = ["${tencentcloud_security_group.sg_bj.id}"
  ]

  vpc_id = "${tencentcloud_vpc.vpc_bj.id}"
  subnet_id = "${tencentcloud_subnet.subnet_bj_02.id}"
  internet_max_bandwidth_out = 15
  count = 1
}


仍旧是 terraform plan and terraform apply



服务没有重建。数据盘扩容胜利，zhangpeng.txt 还在

而后再试一下批改系统盘：
system_disk_size = 100 批改为 system_disk_size = 150
cat cvm.tf

resource "tencentcloud_instance" "cvm_almalinux" {
  instance_name = "cvm-almalinux"
  availability_zone = "ap-beijing-2"
  image_id = "img-q95tlc25"
  instance_type = "S2.MEDIUM2"
  system_disk_type = "CLOUD_PREMIUM"
  system_disk_size  = 150
  hostname = "cvm-almalinux"   
  allocate_public_ip = true
  data_disks {
    data_disk_type = "CLOUD_PREMIUM"
    data_disk_size = 100
    encrypt        = false
  }

  security_groups = ["${tencentcloud_security_group.sg_bj.id}"
  ]

  vpc_id = "${tencentcloud_vpc.vpc_bj.id}"
  subnet_id = "${tencentcloud_subnet.subnet_bj_02.id}"
  internet_max_bandwidth_out = 15
  count = 1
}

仍旧是 terraform plan and terraform apply


仍然没有重建 CVM，什么起因呢？这里的所有操作都是针对与 已有的配置 进行批改大小，没有 新增或者删除 。那就试一下 持续减少一块数据盘 吧！

3. 减少一块新的数据盘

cat cvm.tf

resource "tencentcloud_instance" "cvm_almalinux" {
  instance_name = "cvm-almalinux"
  availability_zone = "ap-beijing-2"
  image_id = "img-q95tlc25"
  instance_type = "S2.MEDIUM2"
  system_disk_type = "CLOUD_PREMIUM"
  system_disk_size  = 150
  hostname = "cvm-almalinux"   
  allocate_public_ip = true
  data_disks {
    data_disk_type = "CLOUD_PREMIUM"
    data_disk_size = 100
    encrypt        = false
  }
  data_disks {
    data_disk_type = "CLOUD_PREMIUM"
    data_disk_size = 50
    encrypt        = false
  }

  security_groups = ["${tencentcloud_security_group.sg_bj.id}"
  ]

  vpc_id = "${tencentcloud_vpc.vpc_bj.id}"
  subnet_id = "${tencentcloud_subnet.subnet_bj_02.id}"
  internet_max_bandwidth_out = 15
  count = 1
}


仍旧是terraform plan and terraform apply


目测带 replaced 的都会重建 ……

4. 论断

貌似在进行新增或者删除相干配置的时候都会重建？找泽阳大佬确认了一下有什么形式能够防止。貌似是我了解错了：写在 cvm.tf 这里貌似算是批改了 cvm 初始化，应该最好把数据盘 负载平衡独自创立，而后将其绑定到对应 cvm!

4. 特别强调

1. terraform destroy

正好顺便体验一下删除配置而后从新创立一下利用：

terraform destroy

2. 独自创立 vpc subset route and cvm

放弃其余配置文件 (vpc subset route and cvm) 不变，批改 cvm.tf 如下：
cat cvm.tf

resource "tencentcloud_instance" "cvm_almalinux" {
  instance_name = "cvm-almalinux"
  availability_zone = "ap-beijing-2"
  image_id = "img-q95tlc25"
  instance_type = "S2.MEDIUM2"
  system_disk_type = "CLOUD_PREMIUM"
  system_disk_size  = 50
  hostname = "cvm-almalinux"   

  security_groups = ["${tencentcloud_security_group.sg_bj.id}"
  ]
  lifecycle {create_before_destroy = false}

  vpc_id = "${tencentcloud_vpc.vpc_bj.id}"
  subnet_id = "${tencentcloud_subnet.subnet_bj_02.id}"
}

3. terraform plan and terraform apply


4. 独自减少一个 eip 并绑定

1. 创立 eip 公网 ip

参照：https://registry.terraform.io/providers/tencentcloudstack/tencentcloud/latest/docs/resources/eip
[root@zhangpeng terraform]# cat eip.tf

resource "tencentcloud_eip" "cvm_almalinux_eip" {
  name                       = "cvm_almalinux_eip"
  internet_max_bandwidth_out = 10
  internet_service_provider  = "BGP"
  type                       = "EIP"
  internet_charge_type       = "TRAFFIC_POSTPAID_BY_HOUR"
}

2. eip 绑定 cvm

参照：https://registry.terraform.io/providers/tencentcloudstack/tencentcloud/latest/docs/resources/eip_association
[root@zhangpeng terraform]# cat eip_association.tf


resource "tencentcloud_eip_association" "cvm_almalinux_association" {eip_id      = "${tencentcloud_eip.cvm_almalinux_eip.id}"
  instance_id = "${tencentcloud_instance.cvm_almalinux.id}"
}

3. terraform plan and terraform apply

这里截图就疏忽了！间接看后果！！


不晓得带宽为什么显示 0 呢这里？
而后 ssh 登陆服务器测试一下：

[root@zhangpeng terraform]# ssh root@xxx.xxx.xxx.xxx
kex_exchange_identification: Connection closed by remote host
[root@zhangpeng terraform]# ssh root@xxx.xxx.xxx.xxx
ssh: connect to host root@xxx.xxx.xxx.xxx port 22: Connection timed out
[root@zhangpeng terraform]# ssh root@xxx.xxx.xxx.xxx

不出所料 带宽没有设置失效！

可是我这里应该三设置胜利了阿 …… 先手动设置一下验证一下独自设置 EIP 绑定 CVM！

持续 ssh 登陆：

登陆胜利没有重建 …… 当然了这里也体验到了还是 allocate_public_ip = true 的形式简略！

5. 持续体验一下独自创立数据盘绑定 cvm

1. 创立数据盘

参照：https://registry.terraform.io/providers/tencentcloudstack/tencentcloud/latest/docs/resources/cbs_storage
[root@zhangpeng terraform]# cat cbs.tf

resource "tencentcloud_cbs_storage" "cvm_almalinux_storage" {
  storage_name      = "cvm_almalinux"
  storage_type      = "CLOUD_PREMIUM"
  storage_size      = 100
  availability_zone = "ap-beijing-2"
  project_id        = 0
  encrypt           = false

  tags = {abc = "tf"}
}

2. 数据盘绑定 cvm

参照：https://registry.terraform.io/providers/tencentcloudstack/tencentcloud/latest/docs/resources/cbs_storage_attachment
[root@zhangpeng terraform]# cat cbs_attachment.tf


resource "tencentcloud_cbs_storage_attachment" "cvm_almalinux_attachment" {storage_id  = "${tencentcloud_cbs_storage.cvm_almalinux_storage.id}"
  instance_id = "${tencentcloud_instance.cvm_almalinux.id}"
}

3. terraform plan and terraform apply


恩服务器没有重建 ….. 登陆服务器查看 disk 数据盘

5. 进一步的体验绑定 ssh-key 密钥形式登陆服务器

有了后面的失败案例，当初筹备独自创立一个密钥文件而后绑定 CVM
参照：https://registry.terraform.io/providers/tencentcloudstack/tencentcloud/latest/docs/resources/key_pair

1. 创立 key_pair

resource "tencentcloud_key_pair" "ssh-key" {
  key_name   = "ssh-key"
  public_key = "ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAAAgQDjd8fTnp7Dcuj4mLaQxf9Zs/ORgUL9fQxRCNKkPgP1paTy1I513maMX126i36Lxxl3+FUB52oVbo/FgwlIfX8hyCnv8MCxqnuSDozf1CD0/wRYHcTWAtgHQHBPCC2nJtod6cVC3kB18KeV4U7zsxmwFeBIxojMOOmcOBuh7+trRw=="
}

留神：我这里应用了我本地环境的 id_rsa.pub! 以上为官网例子

2. cvm 减少 key_pair 配置

减少 key_name 配置！
[root@zhangpeng terraform]# cat cvm.tf

resource "tencentcloud_instance" "cvm_almalinux" {
  instance_name = "cvm-almalinux"
  availability_zone = "ap-beijing-2"
  image_id = "img-q95tlc25"
  instance_type = "S2.MEDIUM2"
  system_disk_type = "CLOUD_PREMIUM"
  system_disk_size  = 50
  hostname = "cvm-almalinux"   

  security_groups = ["${tencentcloud_security_group.sg_bj.id}"
  ]
  lifecycle {create_before_destroy = false}

  key_name= "${tencentcloud_key_pair.ssh_key.id}"
  vpc_id = "${tencentcloud_vpc.vpc_bj.id}"
  subnet_id = "${tencentcloud_subnet.subnet_bj_02.id}"
}

3. terraform plan and terraform apply


ssh 登陆验证：因为我的 ssh-key 是 zhangpeng 用户的故 root 用户登陆失败！切换到 zhangpeng 用户 ssh 免密登陆胜利！

cvm 也没有重建 …… 初步目标达到！

题外话：

总结一下：

  1. 公网 ip 还是在创立 CVM 的时候间接设置allocate_public_ip = true 比拟不便
  2. 数据盘的增加 还有如果须要额定公网 ip 的绑定。能够 独自创立组件，而后参照 attachment相干将其 绑定到 cvm。

  3. ssh-key 的绑定服务器不会重建

    下一步的打算

  4. 配置文件如何治理的更优雅？
  5. 应用 Terraform 在 cvm 中装置软件治理 CVM
  6. Terraform 体验治理其余利用

注：错别字请原谅 ……rocky 中文输入法太残害了 …… 硬盘的扩容也测试了具体过程就不写了！

退出移动版