YandexClusterConfiguration
Version: deckhouse.io/v1
Describes the configuration of a cloud cluster in Yandex Cloud.
Used by the cloud provider if a cluster’s control plane is hosted in the cloud.
Run the following command to change the configuration in a running cluster:
kubectl -n d8-system exec -ti svc/deckhouse-leader -c deckhouse -- deckhouse-controller edit provider-cluster-configuration
Example:
apiVersion: deckhouse.io/v1
kind: YandexClusterConfiguration
layout: Standard
nodeNetworkCIDR: 127.0.0.1/8
labels:
label-2: b
sshPublicKey: "<SSH_PUBLIC_KEY>"
masterNodeGroup:
replicas: 1
instanceClass:
cores: 4
memory: 8192
imageID: fd8nb7ecsbvj76dfaa8b
nodeGroups:
- name: worker
replicas: 1
zones:
- ru-central1-a
instanceClass:
cores: 4
memory: 8192
imageID: fd8nb7ecsbvj76dfaa8b
coreFraction: 50
externalIPAddresses:
- 198.51.100.5
- Auto
provider:
cloudID: "<CLOUD_ID>"
folderID: "<FOLDER_ID>"
serviceAccountJSON: |
{
"id": "id",
"service_account_id": "service_account_id",
"key_algorithm": "RSA_2048",
"public_key": "-----BEGIN PUBLIC KEY-----\nMIIwID....AQAB\n-----END PUBLIC KEY-----\n",
"private_key": "-----BEGIN PRIVATE KEY-----\nMIIE....1ZPJeBLt+\n-----END PRIVATE KEY-----\n"
}
- apiVersion
Required value
Allowed values:
deckhouse.io/v1
,deckhouse.io/v1alpha1
- dhcpOptions
A list of DHCP parameters to use for all subnets.
Note that setting dhcpOptions may lead to problems.
- dhcpOptions.domainName
The name of the search domain.
- dhcpOptions.domainNameServers
A list of recursive DNS addresses.
- Element of the array
Pattern:
^[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}$
- existingNetworkID
The ID of the existing VPC Network.
- existingZoneToSubnetIDMap
One or more pre-existing subnets mapped to respective zone.
Warning! Deckhouse will create a route table that must be manually attached to these subnets.
Pattern:
^[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}/[0-9]{1,2}$
Example:
existingZoneToSubnetIDMap: ru-central1-a: e2lu8r1tbbtryhdpa9ro ru-central1-b: e2lu8r1tbbtryhdpa9ro ru-central1-d: e2lu8r1tbbtryhdpa9ro
- kind
Required value
Allowed values:
YandexClusterConfiguration
- labels
Labels to attach to resources created in the Yandex Cloud.
Note that you have to re-create all the machines to add new labels if labels were modified in the running cluster.
- layout
Required value
The way resources are located in the cloud.
Read more about possible provider layouts.
Allowed values:
Standard
,WithoutNAT
,WithNATInstance
- masterNodeGroup
Required value
The definition of the master’s NodeGroup.
Caution! After changing the parameters of the section, you need to run
dhctl converge
for the changes to take effect.- masterNodeGroup.instanceClass
Required value
Partial contents of the fields of the YandexInstanceClass.
- masterNodeGroup.instanceClass.additionalLabels
Additional labels.
Example:
project: cms-production severity: critical
- masterNodeGroup.instanceClass.cores
Required value
Amount of CPU cores to provision on a Yandex Compute Instance.
- masterNodeGroup.instanceClass.diskSizeGB
Yandex Compute Instance disk size in gibibytes.
Default:
50
Example:
diskSizeGB: 0
- masterNodeGroup.instanceClass.diskType
Yandex Compute Instance disk type.
Default:
"network-ssd"
Allowed values:
network-ssd
,network-ssd-io-m3
,network-ssd-nonreplicated
Example:
diskType: network-ssd-io-m3
- masterNodeGroup.instanceClass.etcdDiskSizeGb
etcd disk size in gibibytes.
If this parameter is changed, each etcd disk must be manually expanded in the Yandex Cloud interface (the reason for this behavior is in the issue).
Default:
10
Example:
etcdDiskSizeGb: 10
- masterNodeGroup.instanceClass.externalIPAddresses
A list of external addresses.
If
externalSubnetID
is not set, you have to use either reserved public IP addresses or theAuto
constant.If
externalSubnetID
is set, you must select specific unallocated IP addresses from the specified subnet.The number of array elements must correspond to the number of nodes in the group (the value of the
replicas
parameter). If the valueAuto
is used (automatic public IP address reservation), then in the list, you still need to specify as many elements with the valueAuto
as there are nodes in the group.The array elements must be listed in the same order as the zones in which the nodes will be created. For example:
- <RU-CENTRAL1-A-IP-ADDRESS> - <RU-CENTRAL1-B-IP-ADDRESS> - <RU-CENTRAL1-D-IP-ADDRESS>
- Element of the array
Pattern:
^([0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3})|(Auto)$
- masterNodeGroup.instanceClass.externalSubnetIDDeprecated
If specified, an additional network interface will be added to the node (the latter will use it as a default route).
- masterNodeGroup.instanceClass.externalSubnetIDs
If specified, an additional network interface will be added to the node (the latter will use it as a default route).
Also, a route for the node’s internal interface will be added (it will cover the entire
nodeNetworkCIDR
subnet).The array elements must be listed in the same order as the zones in which the nodes will be created. For example:
- <RU-CENTRAL1-A-SUBNET-ID> - <RU-CENTRAL1-B-SUBNET-ID> - <RU-CENTRAL1-D-SUBNET-ID>
- masterNodeGroup.instanceClass.imageID
Required value
Image ID to use while provisioning Yandex Compute Instances.
Example:
imageID: fd8nb7ecsbvj76dfaa8b
- masterNodeGroup.instanceClass.memory
Required value
Amount of primary memory in MB provision on a Yandex Compute Instance.
Example:
memory: 8192
- masterNodeGroup.instanceClass.networkType
Network type.
Default:
"Standard"
Allowed values:
Standard
,SoftwareAccelerated
- masterNodeGroup.instanceClass.platform
The type of virtual machine to create.
Default:
"standard-v2"
- masterNodeGroup.replicas
Required value
The number of master nodes to create. It is important to have an odd number of masters to ensure a quorum.
Allowed values:
1 <= X
- masterNodeGroup.zones
Not required value.
A limited set of zones in which nodes can be created.
- Element of the array
Allowed values:
ru-central1-a
,ru-central1-b
,ru-central1-d
- nodeGroups
An array of additional NodeGroups for creating static nodes (e.g., for dedicated front nodes or gateways).
- nodeGroups.instanceClass
Required value
Partial contents of the fields of the YandexInstanceClass.
- nodeGroups.instanceClass.additionalLabels
Additional labels.
Example:
project: cms-production severity: critical
- nodeGroups.instanceClass.coreFraction
Percent of reserved CPU capacity on a Yandex Compute Instance. Details…
Default:
100
Allowed values:
5
,20
,50
,100
Example:
coreFraction: 20
- nodeGroups.instanceClass.cores
Required value
Amount of CPU cores to provision on a Yandex Compute Instance.
- nodeGroups.instanceClass.diskSizeGB
Yandex Compute Instance disk size in gibibytes.
Default:
50
Example:
diskSizeGB: 0
- nodeGroups.instanceClass.diskType
Yandex Compute Instance disk type.
Default:
"network-ssd"
Allowed values:
network-ssd
,network-ssd-io-m3
,network-ssd-nonreplicated
Example:
diskType: network-ssd-io-m3
- nodeGroups.instanceClass.externalIPAddresses
A list of external addresses.
If
externalSubnetID
is not set, you have to use either reserved public IP addresses or theAuto
constant.If
externalSubnetID
is set, you must select specific unallocated IP addresses from the specified subnet.The number of array elements must correspond to the number of nodes in the group (the value of the
replicas
parameter). If the valueAuto
is used (automatic public IP address reservation), then in the list, you still need to specify as many elements with the valueAuto
as there are nodes in the group.The array elements must be listed in the same order as the zones in which the nodes will be created. For example:
- <RU-CENTRAL1-A-IP-ADDRESS> - <RU-CENTRAL1-B-IP-ADDRESS> - <RU-CENTRAL1-D-IP-ADDRESS>
- Element of the array
Pattern:
^([0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3})|(Auto)$
- nodeGroups.instanceClass.externalSubnetIDDeprecated
If specified, an additional network interface will be added to the node (the latter will use it as a default route).
- nodeGroups.instanceClass.externalSubnetIDs
If specified, an additional network interface will be added to the node (the latter will use it as a default route).
Also, a route for the node’s internal interface will be added (it will cover the entire
nodeNetworkCIDR
subnet).The array elements must be listed in the same order as the zones in which the nodes will be created. For example:
- <RU-CENTRAL1-A-SUBNET-ID> - <RU-CENTRAL1-B-SUBNET-ID> - <RU-CENTRAL1-D-SUBNET-ID>
- nodeGroups.instanceClass.imageID
Required value
Image ID to use while provisioning Yandex Compute Instances.
Example:
imageID: fd8nb7ecsbvj76dfaa8b
- nodeGroups.instanceClass.memory
Required value
Amount of primary memory in MB provision on a Yandex Compute Instance.
Example:
memory: 8192
- nodeGroups.instanceClass.networkType
Network type.
Default:
"Standard"
Allowed values:
Standard
,SoftwareAccelerated
- nodeGroups.instanceClass.platform
The type of virtual machine to create.
Default:
"standard-v2"
- nodeGroups.name
Required value
The name of the NodeGroup to use for generating node names.
- nodeGroups.nodeTemplate
Parameters of Node objects in Kubernetes to add after registering the node.
- nodeGroups.nodeTemplate.annotations
The same as the
metadata.annotations
standard field.Example:
annotations: ai.fleet.com/discombobulate: "true"
- nodeGroups.nodeTemplate.labels
A list of labels to attach to cluster resources.
The same as the
metadata.labels
standard field.Note that you have to re-create all the machines to add new tags if tags were modified in the running cluster.
Example:
labels: environment: production app: warp-drive-ai
- nodeGroups.nodeTemplate.taints
The same as the
.spec.taints
field of the Node object.Caution! Only the
effect
,key
,values
fields are available.Example:
taints: - effect: NoExecute key: ship-class value: frigate
- nodeGroups.nodeTemplate.taints.effect
Allowed values:
NoSchedule
,PreferNoSchedule
,NoExecute
- nodeGroups.nodeTemplate.taints.key
- nodeGroups.nodeTemplate.taints.value
- nodeGroups.replicas
Required value
The number of nodes to create.
- nodeGroups.zones
A limited set of zones in which nodes can be created.
- Element of the array
Allowed values:
ru-central1-a
,ru-central1-b
,ru-central1-d
- nodeNetworkCIDR
Required value
This subnet will be split into three equal parts.
They will serve as a basis for subnets in three Yandex Cloud zones.
- provider.cloudID
Required value
The cloud ID.
- provider.folderID
Required value
ID of the directory.
- provider.serviceAccountJSON
Required value
A key to the Service Account in the JSON format.
You can get it by using
[yc iam key create](environment.html)
command.Pattern:
^[ \t\n]*\{(.|\n)*\}[ \t\n]*$
Example:
serviceAccountJSON: | { "id": "...", "service_account_id": "...", "created_at": "2022-08-04T05:38:34.756137618Z", "key_algorithm": "RSA_2048", "public_key": "-----BEGIN PUBLIC KEY-----...-----END PUBLIC KEY-----\n", "private_key": "-----BEGIN PRIVATE KEY-----...-----END PRIVATE KEY-----\n" }
- sshPublicKey
Required value
A public key for accessing nodes.
- withNATInstance.exporterAPIKey
API-key for cloud metrics exporter.
- If parameter is empty, cloud metrics exporter will not be deployed in the cluster.
- If parameter is
Auto
, Deckhouse will create service account with themonitoring.viewer
role and create API-key manually. Provider service account should have theadmin
role. - Any other value is considered a valid API-key. See this instruction for creating API-key.
Service account should have
monitoring.viewer
role.
Default:
""
- withNATInstance.externalSubnetID
If specified, an additional network interface will be added to the node (the node will use it as a default route).
- withNATInstance.internalSubnetID
ID of a subnet for the internal interface.
- withNATInstance.natInstanceExternalAddress
A reserved external IP address (or
externalSubnetID
address if specified).Pattern:
^[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}$
- withNATInstance.natInstanceInternalAddressDeprecated
Consider using automatically generated address instead.
Pattern:
^[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}$
- withNATInstance.natInstanceResources
Computing resources that are allocated to the NAT instance. If not specified, the default values will be used.
- withNATInstance.natInstanceResources.cores
Amount of CPU cores to provision on the NAT instance.
Default:
2
- withNATInstance.natInstanceResources.memory
Amount of primary memory in MB provision on the NAT instance.
Default:
2048
- zones
The globally restricted set of zones that this cloud provider works with.
- Element of the array
Allowed values:
ru-central1-a
,ru-central1-b
,ru-central1-d