PERF03-03 Applying Auto Scaling

Risk level
Medium

Key strategies

For scalable workloads like stateless applications, consider compute services with auto scaling to dynamically adjust resources as needed. Auto scaling ensures sufficient resources during peak hours and prevents over-allocation during off-peak hours. While both VM and container auto-scaling adjust application capacity dynamically, containers offer faster response times and better resource efficiency compared to VMs. Use Auto Scaling (AS) in VM scenarios and use Cluster AutoScaling (CA) and Horizontal Pod Autoscaling (HPA) in container scenarios.

The following describes the auto scaling policies applicable to containers:

CCE supports auto scaling for workloads and nodes.

Workload scaling: auto scaling at the scheduling layer to change the scheduling capacity of workloads. For example, you can use HPA, a scaling component at the scheduling layer, to adjust the number of pods used for an application. Adjusting the number of pods changes the scheduling capacity occupied by the current workload, thereby enabling scaling at the scheduling layer.
Node scaling: auto scaling at the resource layer. When the planned cluster nodes cannot allow workload scheduling, ECS resources are provided to support scheduling.

Workload scaling and node scaling can work separately or together.

For details, see Using HPA and CA for Auto Scaling of Workloads and Nodes.

The following table introduces the workload scaling components

Type	Component	Description	Reference
HPA	Kubernetes Metrics Server	It is a built-in component of Kubernetes and enables horizontal scaling of pods. It is based on Kubernetes HPA, but it also has a cooldown time window and scaling thresholds for applications.	HPA Policy
CustomedHPA	CCE Advanced HPA	An enhanced auto scaling feature, used for auto scaling of Deployments based on metrics (CPU usage and memory usage) or at a periodic interval (a specific time point every day, every week, every month, or every year).	CustomedHPA Policy
Prometheus	Prometheus Cloud Native Cluster Monitoring	An open-source system monitoring and alarm framework, which collects public metrics (CPU usage and memory usage) of kubelet in the Kubernetes cluster.	N/A
CronHPA	CCE Advanced HPA	CronHPA can scale in or out a cluster at a fixed time. It can work with HPA policies to periodically adjust the HPA scaling scope, implementing workload scaling in complex scenarios.	CronHPA Policy

Node scaling components are described as follows:

Component	Description	Scenario	Reference
CCE Cluster Autoscaler	An open source Kubernetes component for horizontal scaling of nodes, which is optimized by CCE in scheduling, auto scaling, and costs.	Online services, deep learning, and large-scale computing with limited resource budgets	Node Scaling
CCE Cloud Bursting Engine for CCI	Used to extend Kubernetes APIs to serverless container platforms (such as CCI), which means you no longer have to worry about node resources.	Online traffic surge, CI/CD, big data, and more	Elastic Scaling of CCE Pods to CCI