Как развернуть или удалить кластер без GUI

Развёртывание и удаление клиентских кластеров выполняется через REST API платформы: запросы отправляются с токеном аутентификации в заголовке Authorization: Bearer $token. Переменная $BACKENDPOINT — адрес backend API платформы.

Авторизация

Перед управлением клиентскими кластерами авторизуйтесь в платформе и получите токен.

В примерах ниже:

$BACKENDPOINT — базовый адрес backend API платформы;
$token — токен авторизованного пользователя;
$CLUSTERNAME — имя создаваемого или удаляемого кластера;
$PROVIDERTEMPLATENAME — имя заранее созданной конфигурации провайдера.

Подготовить провайдер

Создать кластер без провайдера нельзя. Перед запросом на создание кластера в платформе должен быть создан экземпляр провайдера: именно его имя передаётся в теле запроса как identityName, providerName или name в зависимости от типа провайдера.

Получите список доступных провайдеров и их конфигураций:

curl -k -s -L --request GET $BACKENDPOINT/api/v1/platform/providers \
  --header "Authorization: Bearer $token" | jq -r

Пример фрагмента ответа:

[
  {
    "providerType": "capov",
    "configs": [
      {
        "name": "capov-name",
        "kubeVers": ["v1.34.1"],
        "allowEditCSI": false,
        "IPAddressPoolName": "",
        "IPAddressPoolAvailableCount": 0
      }
    ]
  },
  {
    "providerType": "capvs",
    "configs": [
      {
        "name": "capvs",
        "kubeVers": ["v1.34.1"],
        "allowEditCSI": false,
        "IPAddressPoolName": "name-pool",
        "IPAddressPoolAvailableCount": 1
      }
    ]
  }
]

Если подходящей конфигурации нет, создайте её через API провайдера.

VMware vSphere (`capvs`)

Команда создания провайдера VMware vSphere (capvs):

curl -k -s -L --request POST $BACKENDPOINT/api/v1/platform/providers/capvs \
  --header "Authorization: Bearer $token" \
  --data @<(cat <<EOF
{
  "IPAddressPoolName": "name-pool",
  "allowEditCSI": false,
  "cloneMode": "linkedClone",
  "connectInfo": {
    "passwd": "mypassword",
    "url": "stendcloud.aaaaaaa.bb:443",
    "user": "myusername"
  },
  "datacenter": "/StendCloud",
  "datastore": "StendCloud_Name",
  "folder": "/StendCloud/vm/StendCloud/Name/",
  "identityName": "identityname",
  "network": "/StendCloud/network/data",
  "resourcePool": "/StendCloud/host/StendCloud/Resources/Name",
  "storagePolicyName": "",
  "template": [
    { "key": "РЕД ОС-7.3.2-shturval", "value": "v1.34.1" }
  ],
  "templateNetInterface": "enp1s0"
}
EOF
) | jq -r

Пример ответа:

{
  "identityName": "identityname",
  "provider": {
    "vCenterAddress": "stendcloud.aaaaaaa.aa:443",
    "user": "",
    "passwd": "",
    "thumbprint": "A1:F1:C8:9F:04:E8:13:7A:87:39:D0:EA:10:96:40:57:FE:D0:77:3E",
    "datacenter": "/StendCloud",
    "datastore": "StendCloud_Name",
    "storagePolicyName": "",
    "cloneMode": "",
    "folder": "/StendCloud/vm/StendCloud/Name/",
    "network": "/StendCloud/network/data",
    "template": [
      { "key": "РЕД ОС-7.3.2-shturval", "value": "v1.34.1" }
    ],
    "templateNetInterface": "enp1s0",
    "resourcePool": "/StendCloud/host/StendCloud/Resources/Name",
    "allowEditCSI": false,
    "IPAddressPoolName": "name-pool"
  },
  "ready": false
}

oVirt, zVirt, HostVM, РЕД Виртуализация, ROSA Virtualization (`capov`)

Команда создания провайдера oVirt, zVirt, HostVM, РЕД Виртуализация, ROSA Virtualization (capov):

curl -k -s -L --request POST $BACKENDPOINT/api/v1/platform/providers/capov \
  --header "Authorization: Bearer $token" \
  --data @<(cat <<EOF
{
  "name": "identityname",
  "VNICProfile": "VNI",
  "allowEditCSI": false,
  "cluster": "RHVLab",
  "connection": {
    "passwd": "password",
    "tls": {
      "cert": true,
      "certs": "",
      "insecure": true
    },
    "url": "https://rhv-mg.aaaaa.bbb.cc/ovirt-engine/api",
    "user": "name"
  },
  "datacenter": "datacenter",
  "templates": [
    { "template": "AstraLinux", "kubeVersion": "v1.34.1" }
  ],
  "templateNetInterface": "eth0",
  "csistorageDomainName": "hosted_storage",
  "IPAddressPoolName": "",
  "providerSubtype": "rosavirt",
  "infraProviderName": ""
}
EOF
) | jq -r

Пример ответа:

{
  "name": "identityname",
  "cluster": "RHVLab",
  "VNICProfile": "VNI",
  "allowEditCSI": false,
  "VMType": "server",
  "templateNetInterface": "eth0",
  "connection": {
    "url": "https://rhv-mg.aaaaa.bbb.cc/ovirt-engine/api",
    "user": "name",
    "passwd": "password",
    "tls": {
      "cert": true,
      "insecure": true
    }
  },
  "templates": [
    { "template": "AstraLinux", "kubeVersion": "v1.34.1" }
  ],
  "datacenter": "datacenter",
  "csistorageDomainName": "hosted_storage",
  "IPAddressPoolName": "",
  "providerSubtype": "rosavirt",
  "infraProviderName": ""
}

Basis Dynamix (`capbd`)

Команда создания провайдера Basis Dynamix (capbd):

curl -k -s -L --request POST $BACKENDPOINT/api/v1/platform/providers/capbd \
  --header "Authorization: Bearer $token" \
  --data @<(cat <<EOF
{
  "name": "name",
  "accountName": "Shturval",
  "rgName": "dev-2.6",
  "templateNetInterface": "eth0",
  "connection": {
    "decordUrl": "https://mr4.digitalenergy.online/",
    "ssoUrl": "https://sso.digitalenergy.online",
    "appId": "хххххх",
    "appSecret": "dddddd",
    "skipSSLVerification": true
  },
  "templates": [
    { "template": "pfSASE", "kubeVersion": "v1.34.1" }
  ],
  "extNetName": "NAME",
  "vinsName": "res_group_internal",
  "haLoadBalancer": true
}
EOF
) | jq -r

Пример ответа:

{
  "name": "name",
  "accountName": "Shturval",
  "rgName": "dev-2.6",
  "templateNetInterface": "eth0",
  "connection": {
    "decordUrl": "https://mr4.digitalenergy.online/",
    "ssoUrl": "https://sso.digitalenergy.online",
    "appId": "хххххх",
    "appSecret": "dddddd",
    "skipSSLVerification": true
  },
  "templates": [
    { "template": "pfSASE", "kubeVersion": "v1.34.1" }
  ],
  "extNetName": "NAME",
  "vinsName": "res_group_internal",
  "haLoadBalancer": true
}

OpenStack, Selectel, VK Cloud (`capos`)

Команда создания провайдера OpenStack (capos):

curl -k -s -L --request POST $BACKENDPOINT/api/v1/platform/providers/capos \
  --header "Authorization: Bearer $token" \
  --data @<(cat <<EOF
{
  "name": "identityname",
  "cloudconfig": {
    "config": {
      "auth": {
        "auth_url": "https://ххххх:1111/v3/",
        "project_domain_id": "36d4ecf",
        "project_id": "c3cf4f8101",
        "username": "username@example.ru",
        "user_domain_name": "users",
        "password": "mypassword"
      },
      "auth_type": "password",
      "verify": false,
      "identity_api_version": "3",
      "interface": "public",
      "region_name": "RegionOne"
    }
  },
  "nodeCIDR": "10.0.0.0/24",
  "sshKeyName": "sshKeyName",
  "cloudName": "vkcloud",
  "templateNetInterface": "eth0",
  "externalNetworkID": "ec8c6",
  "flavors": ["STD2-16-32", "STD2-8-20"],
  "templates": [
    { "template": "ubuntu24", "kubeVersion": "v1.33.1" }
  ],
  "volumeTypes": ["high-iops"],
  "managedSecurityGroups": "true",
  "enableCSI": false,
  "disablePortSecurity": "false",
  "dns": [""],
  "allowEditCSI": false,
  "availabilityZones": ["GZ1", "PA2"],
  "providerSubtype": "vkcloud"
}
EOF
) | jq -r

Пример ответа:

{
  "name": "identityname",
  "allowEditCSI": false,
  "cloudName": "vkcloud",
  "templateNetInterface": "eth0",
  "cloudconfig": {
    "config": {
      "auth": {
        "auth_url": "https://ххххх:1111/v3/",
        "project_domain_id": "36d4ecf",
        "project_id": "c3cf4f8101",
        "username": "username@example.ru",
        "user_domain_name": "users",
        "password": "mypassword"
      },
      "auth_type": "password",
      "identity_api_version": "3",
      "verify": false
    }
  },
  "templates": [
    { "template": "ubuntu24", "kubeVersion": "v1.33.1" }
  ],
  "externalNetworkID": "ec8c6",
  "sshKeyName": "sshKeyName",
  "flavors": ["STD2-16-32", "STD2-8-20"],
  "volumeTypes": ["high-iops"],
  "nodeCIDR": "10.0.0.0/24",
  "nodeSubnetDNS": null,
  "managedSecurityGroups": "true",
  "disablePortSecurity": "false",
  "rootVolumeAvailabilityZone": "",
  "availabilityZones": ["GZ1", "PA2"],
  "providerSubtype": "vkcloud"
}

Shturval v2 (`capsmv2`)

Команда создания провайдера Shturval v2 (capsmv2):

curl -k -s -L --request POST $BACKENDPOINT/api/v1/platform/providers/capsmv2 \
  --header "Authorization: Bearer $token" \
  --data @<(cat <<EOF
{
  "name": "identityname",
  "IPAddressPoolName": "",
  "credentials": {
    "SSH": {
      "privateKey": "privateKey",
      "userName": "username"
    }
  },
  "hosts": [
    {
      "address": "10.10.10.10",
      "labelsSelector": {
        "role": "controlplane"
      },
      "credentials": null
    }
  ]
}
EOF
) | jq -r

Пример ответа:

{
  "name": "identityname",
  "hosts": [
    {
      "address": "10.10.10.10",
      "labelsSelector": {
        "role": "controlplane"
      },
      "credentials": null
    }
  ],
  "credentials": {
    "SSH": {
      "userName": "username",
      "privateKey": "privateKey"
    }
  },
  "IPAddressPoolName": "",
  "infraProviderName": ""
}

Yandex Cloud (`capyc`)

Команда создания провайдера Yandex Cloud (capyc):

curl -k -s -L --request POST $BACKENDPOINT/api/v1/platform/providers/capyc \
  --header "Authorization: Bearer $token" \
  --data @<(cat <<EOF
{
  "allowedPlatformIDs": ["standard"],
  "connectInfo": {
    "folderId": "b1gcffn",
    "key": "{\n  \"id\": \"aj2u\",\n  \"service_account_id\": \"e49pbe\",\n  \"created_at\": \"2024-12-13T13:28:15.261795772Z\",\n  \"key_algorithm\": \"RSA_2048\",\n  \"public_key\": \"-----BEGIN PUBLIC KEY-----\\nMIIUEyYcDkOn\\nqTrn6lh52A\\nfQIDB\\n-----END PUBLIC KEY-----\\n\",\n  \"private_key\": \"PLEASE DO NOT REMOVE THIS LINE! Yandex.Cloud SA Key ID <aje76qbj2u>\\n-----BEGIN PRIVATE KEY-----\\nM58VMZ0=\\n-----END PRIVATE KEY-----\\n\"\n}"
  },
  "folderID": "b1gcffn",
  "identityName": "provideryandexname",
  "loadBalancerInternal": false,
  "loadBalancerType": "ALB",
  "networkID": "en5tl",
  "subnetID": "be2",
  "templates": [
    { "template": "fd82", "kubeVersion": "v1.33.1" }
  ],
  "zoneID": "zone-c",
  "volumeTypes": ["network-ssd"],
  "allowEditCSI": false
}
EOF
) | jq -r

Пример ответа:

{
  "identityName": "provideryandexname",
  "connectInfo": {
    "key": ""
  },
  "templateNetInterface": "",
  "templates": [
    { "template": "fd82", "kubeVersion": "v1.33.1" }
  ],
  "folderID": "b1gcffn",
  "networkID": "en5tl",
  "subnetID": "be2",
  "zoneID": "zone-c",
  "volumeTypes": ["network-ssd"],
  "loadBalancerType": "ALB",
  "loadBalancerInternal": false,
  "allowedPlatformIDs": ["standard"],
  "allowEditCSI": false
}

VMware vCloud Director (`capcd`)

Команда создания провайдера VMware vCloud Director (capcd):

curl -k -s -L --request POST $BACKENDPOINT/api/v1/platform/providers/capcd \
  --header "Authorization: Bearer $token" \
  --data @<(cat <<EOF
{
  "connectInfo": {
    "username": "username",
    "password": "mypassword",
    "refreshToken": "",
    "org": "Shturval-Client-Cluster",
    "connectType": "login",
    "site": "https://example.ru"
  },
  "name": "identityname",
  "catalog": "shturval",
  "templates": [
    { "template": "Ubuntu_sthurvalv2.11.0", "kubeVersion": "v1.33.1" }
  ],
  "sizingPolicies": ["TKG"],
  "ovdc": "Shturval-Client-Cluster-v1cl4-vdc",
  "vipSubnetName": "",
  "network": "Network",
  "providerType": "vCloud Director"
}
EOF
) | jq -r

Пример ответа:

{
  "name": "identityname",
  "connectInfo": {
    "site": "https://example.ru",
    "org": "Shturval-Client-Cluster",
    "username": "username",
    "password": "mypassword",
    "refreshToken": ""
  },
  "ovdc": "Shturval-Client-Cluster-v1cl4-vdc",
  "network": "Network",
  "vipSubnet": "",
  "catalog": "shturval",
  "templates": [
    { "template": "Ubuntu_sthurvalv2.11.0", "kubeVersion": "v1.33.1" }
  ],
  "placementPolicy": "",
  "sizingPolicies": ["TKG"],
  "allowEditCSI": false,
  "IPAddressPoolName": ""
}

Провайдер можно удалить командой:

export PROVIDERNAME="provider-name"

curl -k -L --request DELETE $BACKENDPOINT/api/v1/platform/providers/{provider}/$PROVIDERNAME \
  --header "Authorization: Bearer $token"

Получить список кластеров

curl -k -s -L --request GET $BACKENDPOINT/api/v1/clusters --header "Authorization: Bearer $token" | jq -r

Пример фрагмента ответа:

{
  "items": [
    {
      "id": "shturval-new-cluster",
      "clusterName": "shturval-new-cluster",
      "clusterType": "platform",
      "annotations": [
        { "key": "shturval.tech/clustername", "value": "shturval-new-cluster" }
      ],
      "providerType": "capov",
      "providerConfigID": "capov-test",
      "apiEndpoint": "10.20.10.20:5555",
      "networks": [
        { "type": "Service CIDR", "network": "10.96.0.0/12", "netmask": "" },
        { "type": "Pod CIDR", "network": "172.16.0.0/16", "netmask": "" }
      ],
      "ingress": "*.shturval-new-cluster.ip-10-20-10-21.shturval.link",
      "argoCdIngress": "",
      "secure": false,
      "phase": "Provisioned",
      "conditions": [
        { "type": "Ready", "status": "True", "lastTransitionTime": "2024-05-27T07:44:37Z" },
        { "type": "ControlPlaneInitialized", "status": "True", "lastTransitionTime": "2024-05-27T07:44:37Z" },
        { "type": "ControlPlaneReady", "status": "True", "lastTransitionTime": "2024-05-27T07:44:37Z" },
        { "type": "InfrastructureReady", "status": "True", "lastTransitionTime": "2024-05-27T07:41:26Z" }
      ],
      "controlplaneNodesCount": 1,
      "controlplaneHealthyNodesCount": 3,
      "workersGroups": [
        { "groupName": "shturval-new-cluster-default", "workersScale": null, "workersCount": 3 },
        { "groupName": "shturval-new-cluster-infra", "workersScale": null, "workersCount": 0 }
      ],
      "UpdateReady": true
    }
  ],
  "count": 1
}

Создание кластера

Перед созданием кластера убедитесь, что для выбранного типа инфраструктуры уже создан провайдер. Запрос создания кластера ссылается на существующую конфигурацию провайдера; без неё создать кластер нельзя.

Метод и путь: POST $BACKENDPOINT/api/v1/platform/providers/clusters/{provider}

`{provider}`	Провайдер
`capvs`	VMware vSphere
`capov`	oVirt, zVirt, HostVM, РЕД Виртуализация и ROSA Virtualization
`capsmv2`	Shturval v2
`capos`	OpenStack, Selectel, VK Cloud
`capbd`	Basis Dynamix
`capyc`	Yandex Cloud
`capcd`	VMware vCloud Director

Тело запроса — JSON с секциями cluster и provider. Имя шаблона провайдера задаётся в поле identityName (vSphere, vCloud Director), providerName (oVirt, zVirt, HostVM, РЕД Виртуализация, ROSA Virtualization, OpenStack, Selectel, VK Cloud, Basis Dynamix, Yandex Cloud) или name (Shturval v2).

Поля секции cluster

Поле	Назначение
`clusterName`	Имя кластера
`apiEndpoint`	IP API-сервера Kubernetes (для capvs, capov, capsmv2, capbd; для capos, capyc, capcd может быть пустым при внешнем балансировщике)
`controlPlaneNodesCount`	Число узлов Control Plane
`ingressvip`	VIP для Ingress при `externalingresslb: false`
`ingress`	Wildcard FQDN Ingress при `externalingresslb: true` (например, `cluster.shturval.link`)
`podSubnet`, `serviceSubnet`	Подсети подов и сервисов (пример: `172.16.0.0/16`, `10.96.0.0/12`)
`shturvalVersion`	Версия платформы на кластере (например, `2.14.0`)
`enabledServices`	Список включаемых сервисов (чарты)
`disabledServices`	Список отключаемых сервисов или `["all"]` — только критически важные
`ingressControllerType`	Контроллер внешнего HTTP-трафика. `nginx` — модуль управления внешними подключениями (`shturval-ingress-controller`). `gateway` — Gateway API на Cilium. По умолчанию `nginx`. Если поле в запросе не заполнено, платформа записывает `nginx`. При `gateway` модуль `shturval-ingress-controller` устанавливается в режиме `absent`
`externalKubeAPILB`, `externalingresslb`	Внешний балансировщик для API и Ingress
`secure`	`false` — самоподписной сертификат Ingress
`acmeRootCert`, `ACMEURL`	Параметры ACME-сертификата (oVirt, zVirt, HostVM, РЕД Виртуализация, ROSA Virtualization, Shturval v2)
`cacert`, `cakey`	Промежуточный сертификат (OpenStack, Selectel, VK Cloud, Yandex Cloud — в варианте с критическими сервисами)
`mgmtCluster`	`false` для клиентского кластера
`deletePersistentVolumes`	Для oVirt, vSphere, OpenStack, Selectel, VK Cloud, vCloud Director, Yandex Cloud. `true` — удалить PV перед удалением кластера, `false` — не удалять. Если поле не передано, берётся политика профиля кластера. Подробнее: Зачистка PV при удалении кластера

Управление сервисами

Явный набор сервисов: disabledServices: [] и перечень в enabledServices.
Только критические сервисы: для любого провайдера укажите disabledServices: ["all"] и не передавайте enabledServices; остальные секции cluster и provider заполните по примеру выбранного провайдера.

При успешном создании ответ может содержать пересчитанные списки enabledServices и disabledServices, а также поле CreatedByUser / createdByUser. Списки сервисов платформа также приводит в соответствие с ingressControllerType: при gateway модуль shturval-ingress-controller устанавливается в режиме absent.

Поля секции provider

controlplaneNodeConfig / controlPlaneNodeConfig — конфигурация Control Plane (nodeConfig с groupName и roles);
workerNodeConfigs — группы default и infra с workersCount и ролями node-role.kubernetes.io/workers, node-role.kubernetes.io/infra;
enableCSI — для VMware vSphere, oVirt, zVirt, HostVM, РЕД Виртуализация, ROSA Virtualization, OpenStack, Selectel, VK Cloud, vCloud Director, Yandex Cloud (в примерах false). Поле доступно, если в экземпляре провайдера allowEditCSI — true.

Конфигурация узлов зависит от провайдера: capvsNodeConfig, capovNodeConfig, caposNodeConfig, capbdNodeConfig, capycNodeConfig, capcdNodeConfig; для Shturval v2 — селекторы controlPlaneSelector и workerSelector с метками role.

VMware vSphere (`capvs`)

Переменные и запрос с полным набором сервисов и самоподписным сертификатом (secure: false):

export CLUSTERNAME="capvs-new-cluster"
export PROVIDERTEMPLATENAME="capvs-template-name"

curl -k -s -L --request POST $BACKENDPOINT/api/v1/platform/providers/clusters/capvs \
  --header "Authorization: Bearer $token" \
  --data @<(cat <<EOF
{
  "cluster": {
    "apiEndpoint": "10.20.30.40",
    "clusterName": "$CLUSTERNAME",
    "controlPlaneNodesCount": 1,
    "disabledServices": [],
    "enabledServices": ["shturval-cert-manager-crds","shturval-vip-provider","shturval-update-crds","shturval-monitoring-crds","shturval-local-csi","shturval-policy-manager-crds","shturval-vip","shturval-node-config-crds","shturval-services","shturval-services-crds","shturval-caching-dns","shturval-cert-manager","shturval-update","shturval-networking","shturval-node-config"],
    "externalKubeAPILB": false,
    "externalingresslb": false,
    "ingress": "",
    "ingressvip": "10.20.30.42",
    "podSubnet": "172.16.0.0/16",
    "secure": false,
    "serviceSubnet": "10.96.0.0/12",
    "shturvalVersion": "2.14.0",
    "mgmtCluster": false
  },
  "provider": {
    "controlplaneNodeConfig": {
      "capvsNodeConfig": { "diskGiB": 60, "memoryMiB": 1200, "numCPUs": 4 },
      "nodeConfig": {
        "groupName": "control-plane",
        "roles": [{ "name": "node-role.kubernetes.io/control-plane" }]
      }
    },
    "enableCSI": false,
    "identityName": "$PROVIDERTEMPLATENAME",
    "workerNodeConfigs": [
      {
        "capvsNodeConfig": { "diskGiB": 60, "memoryMiB": 1200, "numCPUs": 4 },
        "nodeConfig": {
          "groupName": "default",
          "roles": [{ "name": "node-role.kubernetes.io/workers" }]
        },
        "workersCount": 3
      },
      {
        "capvsNodeConfig": { "diskGiB": 60, "memoryMiB": 1200, "numCPUs": 4 },
        "nodeConfig": {
          "groupName": "infra",
          "roles": [{ "name": "node-role.kubernetes.io/infra" }]
        },
        "workersCount": 3
      }
    ]
  }
EOF
) | jq -r

Для варианта только с критическими сервисами используйте общее правило из раздела Управление сервисами; для группы infra можно задать "workersCount": 0.

oVirt, zVirt, HostVM, РЕД Виртуализация, ROSA Virtualization (`capov`)

ACME-сертификат задаётся полями acmeRootCert и ACMEURL в секции cluster. Шаблон провайдера — providerName. Конфигурация узлов — capovNodeConfig (cpu, memory.sizeMB, osDiskSizeGB).

export CLUSTERNAME="capov-new-cluster"
export PROVIDERTEMPLATENAME="capov-template-name"

curl -k -s -L --request POST $BACKENDPOINT/api/v1/platform/providers/clusters/capov \
  --header "Authorization: Bearer $token" \
  --data @<(cat <<EOF
{
  "cluster": {
    "apiEndpoint": "10.20.30.40",
    "clusterName": "$CLUSTERNAME",
    "controlPlaneNodesCount": 1,
    "disabledServices": [],
    "enabledServices": ["shturval-cert-manager-crds","shturval-vip-provider","shturval-update-crds","shturval-monitoring-crds","shturval-local-csi","shturval-policy-manager-crds","shturval-vip","shturval-node-config-crds","shturval-services","shturval-services-crds","shturval-caching-dns","shturval-cert-manager","shturval-update","shturval-networking","shturval-node-config"],
    "externalKubeAPILB": false,
    "externalingresslb": false,
    "ingress": "",
    "ingressvip": "10.20.30.42",
    "podSubnet": "172.16.0.0/16",
    "secure": false,
    "serviceSubnet": "10.96.0.0/12",
    "shturvalVersion": "2.14.0",
    "acmeRootCert": "acmeRootCert",
    "ACMEURL": "https://issuer.ip-xx-xx-xx-xx.shturval.link:9443/acme/acme/directory",
    "mgmtCluster": false
  },
  "provider": {
    "controlplaneNodeConfig": {
      "capovNodeConfig": {
        "cpu": { "cores": 4, "sockets": 1, "threads": 1 },
        "memory": { "sizeMB": 1200 },
        "osDiskSizeGB": 60
      },
      "nodeConfig": {
        "groupName": "control-plane",
        "roles": [{ "name": "node-role.kubernetes.io/control-plane" }]
      }
    },
    "enableCSI": false,
    "providerName": "$PROVIDERTEMPLATENAME",
    "workerNodeConfigs": [
      {
        "capovNodeConfig": {
          "cpu": { "cores": 4, "sockets": 1, "threads": 1 },
          "memory": { "sizeMB": 1200 },
          "osDiskSizeGB": 60
        },
        "nodeConfig": {
          "groupName": "default",
          "roles": [{ "name": "node-role.kubernetes.io/workers" }]
        },
        "workersCount": 3
      },
      {
        "capovNodeConfig": {
          "cpu": { "cores": 4, "sockets": 1, "threads": 1 },
          "memory": { "sizeMB": 1200 },
          "osDiskSizeGB": 60
        },
        "nodeConfig": {
          "groupName": "infra",
          "roles": [{ "name": "node-role.kubernetes.io/infra" }]
        },
        "workersCount": 3
      }
    ]
  }
EOF
) | jq -r

Shturval v2 (`capsmv2`)

Запрос без флага -k у curl. Шаблон — поле name. Узлы выбираются селекторами matchLabels.role: controlplane, workers, infra.

export CLUSTERNAME="capsmv2-new-cluster"
export PROVIDERTEMPLATENAME="capsmv2-template-name"

curl -s -L --request POST $BACKENDPOINT/api/v1/platform/providers/clusters/capsmv2 \
  --header "Authorization: Bearer $token" \
  --data @<(cat <<EOF
{
  "cluster": {
    "apiEndpoint": "10.20.30.40",
    "clusterName": "$CLUSTERNAME",
    "disabledServices": [],
    "enabledServices": ["shturval-local-csi","shturval-services","shturval-update","shturval-vip","shturval-monitoring-crds","shturval-caching-dns","shturval-node-config","shturval-services-crds","shturval-cert-manager-crds","shturval-cert-manager","shturval-node-config-crds","shturval-policy-manager-crds","shturval-networking","shturval-vip-provider","shturval-update-crds"],
    "externalKubeAPILB": false,
    "externalingresslb": false,
    "ingress": "",
    "controlPlaneNodesCount": 1,
    "ingressvip": "10.20.30.42",
    "podSubnet": "172.16.0.0/16",
    "secure": false,
    "serviceSubnet": "10.96.0.0/12",
    "shturvalVersion": "2.14.0",
    "acmeRootCert": "acmeRootCert",
    "ACMEURL": "https://issuer.ip-xx-xx-xx-xx.shturval.link:9443/acme/acme/directory",
    "mgmtCluster": false
  },
  "provider": {
    "controlPlaneNodeConfig": {
      "controlPlaneSelector": {
        "matchExpressions": [],
        "matchLabels": { "role": "controlplane" }
      },
      "nodeConfig": {
        "groupName": "control-plane",
        "roles": [{ "name": "node-role.kubernetes.io/control-plane" }]
      }
    },
    "name": "$PROVIDERTEMPLATENAME",
    "workerNodeConfigs": [
      {
        "nodeConfig": {
          "groupName": "default",
          "roles": [{ "name": "node-role.kubernetes.io/workers" }]
        },
        "workersCount": 3,
        "workerSelector": {
          "matchExpressions": [],
          "matchLabels": { "role": "workers" }
        }
      },
      {
        "nodeConfig": {
          "groupName": "infra",
          "roles": [{ "name": "node-role.kubernetes.io/infra" }]
        },
        "workersCount": 0,
        "workerSelector": {
          "matchExpressions": [],
          "matchLabels": { "role": "infra" }
        }
    ]
  }
EOF
) | jq -r

OpenStack, Selectel, VK Cloud (`capos`)

externalKubeAPILB и externalingresslb — true, ingressvip — пустая строка. Промежуточный сертификат: cacert, cakey. Узлы — caposNodeConfig (openstackFlavorName, volumeType, volumeSize, availabilityZone). Для worker-групп flavor задаётся переменной $OPENSTACKFLAVORNAME.

export CLUSTERNAME="capos-new-cluster"
export PROVIDERTEMPLATENAME="capos-template-name"

curl -k -s -L --request POST $BACKENDPOINT/api/v1/platform/providers/clusters/capos \
  --header "Authorization: Bearer $token" \
  --data @<(cat <<EOF
{
  "cluster": {
    "clusterName": "$CLUSTERNAME",
    "disabledServices": [],
    "enabledServices": ["shturval-monitoring-crds","shturval-vip","shturval-services-crds","shturval-caching-dns","shturval-cert-manager","shturval-update-crds","shturval-services","shturval-update","shturval-networking","shturval-node-config-crds","shturval-node-config","shturval-cert-manager-crds","shturval-vip-provider","shturval-local-csi","shturval-policy-manager-crds"],
    "externalKubeAPILB": true,
    "externalingresslb": true,
    "ingress": "",
    "controlPlaneNodesCount": 1,
    "ingressvip": "",
    "podSubnet": "172.16.0.0/16",
    "secure": false,
    "serviceSubnet": "10.96.0.0/12",
    "shturvalVersion": "2.14.0",
    "cacert": "cacert",
    "cakey": "cakey",
    "mgmtCluster": false
  },
  "provider": {
    "controlplaneNodeConfig": {
      "caposNodeConfig": {
        "openstackFlavorName": "STD3-4-8",
        "volumeType": "ceph-hdd",
        "volumeSize": 60,
        "availabilityZone": "ME1"
      },
      "nodeConfig": {
        "groupName": "control-plane",
        "roles": [{ "name": "node-role.kubernetes.io/control-plane" }]
      }
    },
    "providerName": "$PROVIDERTEMPLATENAME",
    "enableCSI": false,
    "workerNodeConfigs": [
      {
        "nodeConfig": {
          "groupName": "default",
          "roles": [{ "name": "node-role.kubernetes.io/workers" }]
        },
        "workersCount": 3,
        "caposNodeConfig": {
          "openstackFlavorName": "$OPENSTACKFLAVORNAME",
          "volumeSize": 60,
          "volumeType": "ceph-hdd",
          "availabilityZone": "ME1"
        }
      },
      {
        "nodeConfig": {
          "groupName": "infra",
          "roles": [{ "name": "node-role.kubernetes.io/infra" }]
        },
        "workersCount": 0,
        "caposNodeConfig": {
          "openstackFlavorName": "$OPENSTACKFLAVORNAME",
          "volumeSize": 60,
          "volumeType": "ceph-hdd",
          "availabilityZone": "ME1"
        }
    ]
  }
EOF
) | jq -r

Basis Dynamix (`capbd`)

Для Control Plane в capbdNodeConfig.networkInterfaces указываются netType и networkName (переменные $NETTYPE, $NETWORKNAME). externalingresslb: true, ingressvip отличается от VMware vSphere (пример: 10.20.30.41).

export CLUSTERNAME="capbd-new-cluster"
export PROVIDERTEMPLATENAME="capbd-template-name"
export NETTYPE="net-type"
export NETWORKNAME="networkname"

curl -k -s -L --request POST $BACKENDPOINT/api/v1/platform/providers/clusters/capbd \
  --header "Authorization: Bearer $token" \
  --data @<(cat <<EOF
{
  "cluster": {
    "apiEndpoint": "10.20.30.40",
    "clusterName": "$CLUSTERNAME",
    "disabledServices": [],
    "enabledServices": ["shturval-node-config-crds","shturval-backup","shturval-services-crds","shturval-services","shturval-rollouts-crds","shturval-descheduler","shturval-vip-provider","shturval-ingress-controller","shturval-policy-manager-crds","shturval-networking","shturval-metrics-collector","shturval-monitoring-crds","shturval-scanner","shturval-caching-dns","shturval-vip","shturval-node-config","shturval-cert-expiration","shturval-update-crds","shturval-backup-crds","shturval-cd-crds","shturval-cert-manager-crds","shturval-cert-manager","shturval-local-csi","shturval-update","shturval-scanner-crds","shturval-policy-manager"],
    "externalKubeAPILB": false,
    "externalingresslb": true,
    "ingress": "",
    "controlPlaneNodesCount": 1,
    "ingressvip": "10.20.30.41",
    "podSubnet": "172.16.0.0/16",
    "secure": false,
    "serviceSubnet": "10.96.0.0/12",
    "shturvalVersion": "2.14.0",
    "mgmtCluster": false
  },
  "provider": {
    "controlplaneNodeConfig": {
      "capbdNodeConfig": {
        "cpu": 4,
        "memory": 12000,
        "networkInterfaces": [{ "netType": "$NETTYPE", "networkName": "$NETWORKNAME" }],
        "osDiskSizeGB": 60
      },
      "nodeConfig": {
        "groupName": "control-plane",
        "roles": [{ "name": "node-role.kubernetes.io/control-plane" }]
      }
    },
    "providerName": "$PROVIDERTEMPLATENAME",
    "workerNodeConfigs": [
      {
        "nodeConfig": {
          "groupName": "default",
          "roles": [{ "name": "node-role.kubernetes.io/workers" }]
        },
        "workersCount": 3,
        "capbdNodeConfig": { "cpu": 4, "memory": 12000, "osDiskSizeGB": 60 }
      },
      {
        "nodeConfig": {
          "groupName": "infra",
          "roles": [{ "name": "node-role.kubernetes.io/infra" }]
        },
        "workersCount": 3,
        "capbdNodeConfig": { "cpu": 4, "memory": 12000, "osDiskSizeGB": 60 }
      }
    ]
  }
EOF
) | jq -r

Ответ при создании может вернуть краткое описание кластера с phase: "Pending" и сгенерированным ingress вида *.capbd-new-cluster.ip-10-20-30-41.shturval.link.

Yandex Cloud (`capyc`)

externalKubeAPILB и externalingresslb — true, ingress — FQDN (например, cluster.shturval.link), ingressvip — пустая строка. Узлы — capycNodeConfig (cpuCores, memorySize, platformID, volumeType, volumeSize). enableCSI включает shturval-capy-csi; при allowEditCSI: false в экземпляре значение true отклоняется. deletePersistentVolumes — удаление PV при удалении кластера; если поле не передано, берётся политика профиля.

export CLUSTERNAME="capyc-new-cluster"
export PROVIDERTEMPLATENAME="capyc-template-name"

curl -k -s -L --request POST $BACKENDPOINT/api/v1/platform/providers/clusters/capyc \
  --header "Authorization: Bearer $token" \
  --data @<(cat <<EOF
{
  "cluster": {
    "clusterName": "$CLUSTERNAME",
    "disabledServices": [],
    "enabledServices": ["shturval-cert-manager-crds","shturval-vip-provider","shturval-update-crds","shturval-monitoring-crds","shturval-local-csi","shturval-policy-manager-crds","shturval-vip","shturval-node-config-crds","shturval-services","shturval-services-crds","shturval-caching-dns","shturval-cert-manager","shturval-update","shturval-networking","shturval-node-config"],
    "externalKubeAPILB": true,
    "externalingresslb": true,
    "ingress": "cluster.shturval.link",
    "controlPlaneNodesCount": 1,
    "ingressvip": "",
    "podSubnet": "172.16.0.0/16",
    "secure": false,
    "serviceSubnet": "10.96.0.0/12",
    "shturvalVersion": "2.14.0",
    "deletePersistentVolumes": false,
    "mgmtCluster": false
  },
  "provider": {
    "controlplaneNodeConfig": {
      "capycNodeConfig": {
        "cpuCores": 4,
        "memorySize": "800Mi",
        "platformID": "standard",
        "volumeType": "network-ssd",
        "volumeSize": "40Gi"
      },
      "nodeConfig": {
        "groupName": "control-plane",
        "roles": [{ "name": "node-role.kubernetes.io/control-plane" }]
      }
    },
    "providerName": "$PROVIDERTEMPLATENAME",
    "enableCSI": false,
    "workerNodeConfigs": [
      {
        "nodeConfig": {
          "groupName": "default",
          "roles": [{ "name": "node-role.kubernetes.io/workers" }]
        },
        "workersCount": 1,
        "capycNodeConfig": {
          "platformID": "standard",
          "volumeType": "network-ssd",
          "volumeSize": "40Gi",
          "memorySize": "800Mi",
          "cpuCores": 4
        }
      },
      {
        "nodeConfig": {
          "groupName": "infra",
          "roles": [{ "name": "node-role.kubernetes.io/infra" }]
        },
        "workersCount": 0,
        "capycNodeConfig": {
          "platformID": "standard",
          "volumeType": "network-ssd",
          "volumeSize": "40Gi",
          "memorySize": "800Mi",
          "cpuCores": 4
        }
    ]
  }
EOF
) | jq -r

VMware vCloud Director (`capcd`)

Путь API — capcd. Шаблон — identityName. Узлы — capcdNodeConfig (diskSize, sizingPolicy). Дополнительные поля провайдера: kubeAPIVIPSubnet, serviceVIPSubnet.

export CLUSTERNAME="capvcd-new-cluster"
export PROVIDERTEMPLATENAME="capvcd-template-name"

curl -k -s -L --request POST $BACKENDPOINT/api/v1/platform/providers/clusters/capcd \
  --header "Authorization: Bearer $token" \
  --data @<(cat <<EOF
{
  "cluster": {
    "apiEndpoint": "",
    "clusterName": "$CLUSTERNAME",
    "disabledServices": [],
    "enabledServices": ["shturval-cert-manager-crds","shturval-vip-provider","shturval-update-crds","shturval-monitoring-crds","shturval-local-csi","shturval-policy-manager-crds","shturval-vip","shturval-node-config-crds","shturval-services","shturval-services-crds","shturval-caching-dns","shturval-cert-manager","shturval-update","shturval-networking","shturval-node-config"],
    "externalKubeAPILB": true,
    "externalingresslb": true,
    "ingress": "cluster.shturval.link",
    "controlPlaneNodesCount": 1,
    "ingressvip": "",
    "podSubnet": "172.16.0.0/16",
    "secure": false,
    "serviceSubnet": "10.96.0.0/12",
    "shturvalVersion": "2.14.0",
    "clusterConfigProfileName": "",
    "mgmtCluster": false
  },
  "provider": {
    "enableCSI": false,
    "identityName": "$PROVIDERTEMPLATENAME",
    "kubeAPIVIPSubnet": "",
    "serviceVIPSubnet": "",
    "controlplaneNodeConfig": {
      "capcdNodeConfig": { "diskSize": "50Gi", "sizingPolicy": "TKG" },
      "nodeConfig": {
        "groupName": "control-plane",
        "roles": [{ "name": "node-role.kubernetes.io/control-plane" }]
      }
    },
    "workerNodeConfigs": [
      {
        "nodeConfig": {
          "groupName": "default",
          "roles": [{ "name": "node-role.kubernetes.io/workers" }]
        },
        "workersCount": 2,
        "capcdNodeConfig": { "diskSize": "50Gi", "sizingPolicy": "TKG" }
      },
      {
        "nodeConfig": {
          "groupName": "infra",
          "roles": [{ "name": "node-role.kubernetes.io/infra" }]
        },
        "workersCount": 0,
        "capcdNodeConfig": { "diskSize": "50Gi", "sizingPolicy": "TKG" }
      }
    ]
  }
EOF
) | jq -r

Удаление кластера

export CLUSTERNAME="clustername"

curl -k -L --request DELETE $BACKENDPOINT/api/v1/clusters/$CLUSTERNAME --header "Authorization: Bearer $token"

Изменено 2026-10-02 в 14:25

Как развернуть или удалить кластер без GUI

Авторизация #

Подготовить провайдер #

VMware vSphere (capvs) #

oVirt, zVirt, HostVM, РЕД Виртуализация, ROSA Virtualization (capov) #

Basis Dynamix (capbd) #

OpenStack, Selectel, VK Cloud (capos) #

Shturval v2 (capsmv2) #

Yandex Cloud (capyc) #

VMware vCloud Director (capcd) #

Получить список кластеров #

Создание кластера #

Поля секции cluster #

Управление сервисами #

Поля секции provider #

VMware vSphere (capvs) #

oVirt, zVirt, HostVM, РЕД Виртуализация, ROSA Virtualization (capov) #

Shturval v2 (capsmv2) #

OpenStack, Selectel, VK Cloud (capos) #

Basis Dynamix (capbd) #

Yandex Cloud (capyc) #

VMware vCloud Director (capcd) #

Удаление кластера #

Авторизация

Подготовить провайдер

VMware vSphere (`capvs`)

oVirt, zVirt, HostVM, РЕД Виртуализация, ROSA Virtualization (`capov`)

Basis Dynamix (`capbd`)

OpenStack, Selectel, VK Cloud (`capos`)

Shturval v2 (`capsmv2`)

Yandex Cloud (`capyc`)

VMware vCloud Director (`capcd`)

Получить список кластеров

Создание кластера

Поля секции cluster

Управление сервисами

Поля секции provider

VMware vSphere (`capvs`)

oVirt, zVirt, HostVM, РЕД Виртуализация, ROSA Virtualization (`capov`)

Shturval v2 (`capsmv2`)

OpenStack, Selectel, VK Cloud (`capos`)

Basis Dynamix (`capbd`)

Yandex Cloud (`capyc`)

VMware vCloud Director (`capcd`)

Удаление кластера