create-nodegroup

Description

Create a new node group within an existing VKS cluster. By default, one node is created with an Ubuntu OS image, a 100 GiB disk, and a SURGE upgrade strategy.

Node group names must be 5–15 characters, contain only lowercase letters, digits, and hyphens, and must start and end with a letter or digit.

Use --dry-run to validate parameters (name format, disk size range, node count range) without sending a create request.

Synopsis

grn vks create-nodegroup
    --cluster-id <value>
    --name <value>
    --flavor-id <value>
    --disk-type <value>
    --ssh-key-id <value>
    [--os <value>]
    [--private-nodes <enabled|disabled>]
    [--num-nodes <value>]
    [--disk-size <value>]
    [--security-groups <value>]
    [--subnet-id <value>]
    [--labels <value>]
    [--taints <value>]
    [--tags <value>]
    [--secondary-subnets <value>]
    [--auto-scale <value>]
    [--placement-group <value>]
    [--upgrade-config <value>]
    [--enable-encryption-volume]
    [--dry-run]

Options

--cluster-id (string)

ID of the cluster to add the node group to.

Required: Yes

--name (string)

Name of the node group.

Required: Yes
Constraints: 5–15 characters; lowercase letters, digits, and hyphens; must start and end with a letter or digit.

--flavor-id (string)

Flavor (instance type) ID for the nodes.

Required: Yes

--disk-type (string)

Disk type ID for the node boot volumes (e.g. SSD, NVME).

Required: Yes

--ssh-key-id (string)

SSH key pair ID to inject into each node.

Required: Yes

--os (string)

Operating system image for the nodes.

Required: No
Default: ubuntu
Possible values: ubuntu, linux, rocky

--private-nodes (string)

Private nodes state. enabled means nodes will not have public IP addresses.

Required: No
Default: disabled
Possible values: enabled, disabled

--num-nodes (integer)

Number of nodes to create in the node group.

Required: No
Default: 1
Constraints: 0–10

--disk-size (integer)

Boot disk size in GiB.

Required: No
Default: 100
Constraints: 20–5000

--security-groups (list<string>)

Security group IDs to attach to the nodes, comma-separated.

Required: No
Syntax: secg-aaa111,secg-bbb222

--subnet-id (string)

Subnet ID for the node group. Uses the cluster subnet when not specified.

Required: No

--labels (map)

Kubernetes node labels as comma-separated key=value pairs.

Required: No
Syntax: env=prod,tier=app

--taints (list<string>)

Kubernetes node taints as comma-separated key=value:effect entries.

Required: No
Syntax: dedicated=gpu:NoSchedule

--tags (map)

Cloud tags for the node group as comma-separated key=value pairs.

Required: No
Syntax: team=platform,cost-center=42

--secondary-subnets (list<string>)

Secondary subnet IDs for the node group, comma-separated.

Required: No
Syntax: sub-aaa,sub-bbb

--auto-scale (structure)

Auto-scaling configuration for the node group. Accepts shorthand or JSON.

Required: No
Members:
- minSize (integer) — minimum number of nodes; minimum value 0
- maxSize (integer) — maximum number of nodes; minimum value 1

Shorthand syntax:

minSize=2,maxSize=10

JSON syntax:

{"minSize": 2, "maxSize": 10}

--placement-group (structure)

Placement group configuration for the node group. Accepts shorthand or JSON.

Required: No
Members:
- type (string) — NEW to create a new placement group, EXISTING to use an existing one
- placementGroupId (string) — ID of an existing placement group; used when type is EXISTING
- placementGroupName (string) — name for a new placement group; used when type is NEW

Shorthand syntax:

type=NEW,placementGroupName=pg-1

JSON syntax:

{"type": "EXISTING", "placementGroupId": "server-group-06b86747-eaf7-47dd-9e41-579c2e30bfdd"}

--upgrade-config (structure)

Upgrade strategy configuration for the node group. Accepts shorthand or JSON.

Required: No
Default: maxSurge=1,maxUnavailable=0,strategy=SURGE
Members:
- strategy (string) — upgrade strategy; currently only SURGE is supported
- maxSurge (integer) — maximum number of extra nodes added during upgrade; range 1–100
- maxUnavailable (integer) — maximum number of nodes that may be unavailable during upgrade; range 0–100

Shorthand syntax:

maxSurge=1,maxUnavailable=0,strategy=SURGE

JSON syntax:

{"maxSurge": 1, "maxUnavailable": 0, "strategy": "SURGE"}

--enable-encryption-volume (boolean)

Enable encryption for the node boot volumes.

Required: No
Default: false

--dry-run (boolean)

Validate parameters and print a report without sending the create request.

Required: No
Default: false

Global options

This command also accepts the global options (--profile, --region, --output, --query, --endpoint-url, --debug, …).

Examples

Create a basic node group:

grn vks create-nodegroup \
  --cluster-id cls-abc12345-6789-def0-1234-abcdef012345 \
  --name worker-ng \
  --flavor-id flv-4c8g \
  --disk-type SSD \
  --ssh-key-id key-abc12345-0000-0000-0000-000000000001

Create a GPU node group with taints and labels:

grn vks create-nodegroup \
  --cluster-id cls-abc12345-6789-def0-1234-abcdef012345 \
  --name gpu-ng \
  --flavor-id flv-8c32g-gpu \
  --disk-type SSD \
  --disk-size 200 \
  --num-nodes 2 \
  --ssh-key-id key-abc12345-0000-0000-0000-000000000001 \
  --labels accelerator=nvidia,tier=gpu \
  --taints dedicated=gpu:NoSchedule \
  --enable-encryption-volume

Create an auto-scaling node group with a custom upgrade config:

grn vks create-nodegroup \
  --cluster-id cls-abc12345-6789-def0-1234-abcdef012345 \
  --name auto-ng \
  --flavor-id flv-4c8g \
  --disk-type SSD \
  --ssh-key-id key-abc12345-0000-0000-0000-000000000001 \
  --auto-scale minSize=2,maxSize=10 \
  --upgrade-config '{"maxSurge":2,"maxUnavailable":1,"strategy":"SURGE"}' \
  --tags team=platform,env=prod

Validate parameters without creating (dry run):

grn vks create-nodegroup \
  --cluster-id cls-abc12345-6789-def0-1234-abcdef012345 \
  --name worker-ng \
  --flavor-id flv-4c8g \
  --disk-type SSD \
  --ssh-key-id key-abc12345-0000-0000-0000-000000000001 \
  --dry-run