Check the scalability level of a worker node pool - Container Service for Kubernetes

When instances are out of stock or specific instance types are not supported in the zones where a node pool is deployed, the node pool may fail to be scaled out. You can evaluate the configuration availability and instance inventory sufficiency of a node pool based on its scalability level. The Container Service for Kubernetes (ACK) console provides suggestions for node pools based on their scalability levels. If the scalability level of a node pool is low, the scalability level of the node pool is displayed in the ACK console.

Terms

Scalability

The scalability level of a node pool is the sum of the scalability levels of the available resource pools in the zones where the node pool is deployed. The larger the number of available resource pools, the higher the scalability level. The scalability level of a node pool is displayed in the ACK console only when the scalability level is low (less than zero).

Resource pool

When the system scales out a node pool, the instances that are added to the node pool use the configurations of the node pool, including the zone information and instance types. Elastic Compute Service (ECS) instances of each instance type in each zone are grouped into a resource pool. The number of resource pools for a node pool depends on the number of zones and the number of instance types specified in the configurations of the node pool.

Available resource pool

Some resource pools may be unavailable. A resource pool may become unavailable due to insufficient instance inventory or constraints between resources. For example, the c6e instance family supports Enterprise SSDs (ESSDs) but not standard SSDs.

The following figure shows the node pool configurations and the corresponding resource pools. Two zones, two instance types, and one disk category are specified in the node pool configurations. There are four resource pools for the node pool. However, only one resource pool is available. The other resource pools are unavailable due to reasons such as insufficient instance inventory, unsupported zones, or disk constraints.

Check the scalability of a node pool

Log on to the ACK console. In the left-side navigation pane, click Clusters.
On the Clusters page, find the cluster that you want to manage and click its name. In the left-side pane, choose Nodes > Node Pools.
On the Node Pools page, you can check whether the scalability level of a node pool is low in the Status column. If the scalability level of a node pool is less than zero, the scalability level is evaluated as low.
Click Scalability Level: Weak to view the scalability level details. The following figure shows an example.

Causes for low scalability levels and suggestions

For more information about how to modify node pool configurations, see Create and manage a node pool. To ensure the high availability of a node pool, we recommend that you select multiple vSwitches in different zones and specify multiple instance types that have sufficient inventory.

You can view the scalability level of a node pool in the Scalability Details dialog box.

Check item	Result	Suggestion
Constraints between resources	The image does not match the instance type.	Change the instance type.
	The system disk does not match the instance type.	Change the system disk category or instance type.
	The data disk does not match the instance type.	Change the data disk category or instance type.
	The instance type does not support IPv6.	Change the instance type.
	The vSwitch does not exist.	Change the vSwitch.
Instance inventory	The instance type does not support the zone.	Change the instance type.
Instance inventory	The inventory of the instance type is insufficient.	Change the instance type.

References

For more information about how to configure nodes, see ECS specification recommendations for ACK clusters.