Create a Managed Kubernetes cluster on a dedicated server
You can create a Managed Kubernetes cluster with a group of nodes on the dedicated server.
Dedicated servers are used when needed:
- consistently high performance;
- increased privacy;
- reserved capacity.
In Managed Kubernetes, you can use off-the-shelf dedicated servers with a local port.
The dedicated server has already been assembled, mounted and networked privately at the L3 level with Kubernetes Control Plane using the global router.
In one project and in one pool, you can create a maximum of 10 fault-tolerant clusters and 10 basic Managed Kubernetes clusters on dedicated servers.
Configure the cluster
-
В control panels go to Cloud platform → Kubernetes.
-
Click Create a cluster.
-
Enter a name for the cluster. The name will appear in the names of the cluster objects: node groups, nodes, balancers, networks, and disks. For example, if the cluster name is
kelsie
then the name of the node group will bekelsie-node-gdc8q
and the boot diskkelsie-node-gdc8q-volume
. -
Select region and pool In the cluster, the master nodes will be located in the cluster. Once a cluster is created, the pool cannot be changed.
-
Select a version of Kubernetes 1.28.
-
Select cluster type:
- fault-tolerant — Control Plane is placed on three master nodes that run on different hosts in different segments of the same pool. If one of the three master nodes is unavailable, Control Plane continues to run;
- basic — Control Plane is hosted on a single master node that runs on a single host on a single pool segment. If the master node is unavailable, Control Plane will not run.
Once a cluster is created, the cluster type cannot be changed.
-
Optional: to make the cluster accessible over a private network and inaccessible from the Internet, check the checkbox Private kube API. By default, the cluster is created in a public network and it is automatically assigned a public IP-address of kube API, accessible from the Internet. After cluster creation the type of access to kube API cannot be changed.
-
Click Continue.
Configure the node group
-
In the field Server type select Dedicated server.
-
Select pool The nodes must be located in the same node in the group. The work nodes must be located in the same accessibility zone with master nodes. Once a cluster is created, the pool cannot be changed.
-
Click Select configuration and select the configuration of worker nodes in the group:
3.1 Select tariff plan.
3.2 Select a ready-made dedicated server configuration.
3.3. Press Select.
Once the cluster is created, the node configuration cannot be changed.
-
Specify the number of working nodes in the group.
-
Optional: to add node group tags in the field Tags click Add. Enter the key and the label value. Press Add. Once a cluster is created, you cannot create new labels, modify existing labels, or delete labels.
-
Optional: to add node group teints in the field Taints click Add. Enter the key and the value of theint. Select the effect:
- NoSchedule — new pods will not be added and existing pods will continue to run;
- PreferNoSchedule — new pods will be added if there are no other available slots in the cluster;
- NoExecute — running pods without tolerations will be removed.
Click Add.
Once a cluster is created, you cannot create new teints, modify existing teints, or delete teints.
-
Click Continue.
set up automation
-
Select service start time cluster — the time when automatic cluster maintenance actions will start.
-
Optional: to enable audit logs and check the box Audit logs. After the cluster is created Set up integration with the log storage and analysis system.
-
Check the price of a cluster on a dedicated server.
-
Click Create. Cluster creation takes up to 60 minutes, during which time the cluster will be in the status
CREATING
. The cluster will be ready for operation when it moves to statusACTIVE
.A private network will be automatically created
cluster_name-network
, private subnet, VLAN and global router<cluster_name>-router
wherecluster_name
— cluster name. The CIDR subnets of the dedicated server and the CIDR subnets of the cloud infrastructure are assigned automatically.