API
The Gcore Customer Portal is being updated. Screenshots may not show the current version.
Edge Cloud
Edge Cloud
Overview
Overview
Activate and manage you eSIMGetting online in China
Troubleshooting
Account types and permissionsPackagesQuality of Service
Billing
Quotas
Create a project
User rolesAdd a userManage accessView user actions
Cost reports
About Resource reservation
Types of Virtual MachinesCreateCustomize initial setupCheck the statusEnable root user
Connect via Customer Portal
Configure and manage SSH keysConvert an SSH key to PPK formatEstablish SSH connection to a Virtual Machine
OverviewConfigure
OverviewCreate and configure
Monitor
Take a snapshot of your file systemSet up automatic snapshots
SSH connection
About Bare Metal serversCreate a Bare Metal serverConnect to your Bare Metal server via SSH
SSH connection
About Advanced DDoS Protection for Bare MetalActivate Advanced DDoS Protection for Bare Metal
OverviewUpload
NetworkSubnetworkRouter
Floating IP addressReserved IP addressVirtual IP addressDifference between virtual, floating, and reserved IP addressesAllowed address pair
Create Load BalancerAdd TLS certificates to a Load BalancerConfigure mutual TLS authenticationManage Load BalancersLoggingAnnotations
Firewall
OverviewConfigure
OverviewBare Metal support
Create
SSH connectionkubectl
Secure cluster with OpenID Connect
OverviewChange limitsCreate and configure a poolAdvanced autoscaler settingsConfigure GPU autoscaling for Kubernetes
Add users with limited rights to a Kubernetes cluster
Monitor load on a specific node
OverviewInstall nginx ControllerUse nginx Controller
Create a PVC and bind it to a podConfigure NFS CSI driver for Kubernetes
Annotations for Load Balancers
Manage PostgreSQL serversBackup and restore
Upload a PKCS12 file
Function as a ServiceAuthenticate to functions with API keysFunction examples
Create a containerDeploy via GitHub actionsManage containers
Create a Container RegistryManage Container Registries
OverviewConfigure
FilebeatFluent BitFluentdLogstash
Prometheus exporterView logs in Grafana with OpenSearch plugin
Terraform
Set up a mailing server
API
Chosen image
Home/Edge Cloud/Managed Kubernetes/Clusters/Autoscaling/Overview

About autoscaling

What is autoscaling?

Autoscaling is a system that adjusts the number of nodes in a pool based on changes in load. If the existing nodes cannot handle the load, a new one will be deployed. Conversely, if some nodes are idle without any load, they will be removed. 

How does autoscaling work?

You set the minimum and maximum number of nodes, and autoscaling will ensure that the number of nodes remains within these limits. It will not create more nodes than the specified maximum, and it will not remove any nodes if the minimum has been reached. 

  • If a new pod (application container group) needs to run and the existing nodes don’t have enough resources to do it, autoscaling will deploy a new node. The amount of free resources is determined by "resources.requests".
  • If some nodes are running without any pods for 20 minutes straight, the system will remove them.     
  • If you manually delete nodes with running pods, reducing the current number of nodes below the level set by autoscaling, the system will reassess the load on the pool after 20 minutes. If the current nodes are unable to handle the required pods, more nodes will be added. If the nodes are sufficient, no additional nodes will be added.

Change autoscaling settings

1. In the Cloud menu, go to Kubernetes, find the required cluster and click its name.

Screenshot_2023-02-27_at_13.37_2.png

2. Go to the Pools section, click ⋯ next to the pool and select Change autoscaling limits.

Screenshot_2023-02-27_at_20.42_1.jpg

3. In the new window, specify the maximum and minimum number of nodes and click Save.

New autoscaling settings have been applied.

Disable autoscaling

When you turn off autoscaling, the number of nodes in the pool will remain constant.

1. In the Cloud menu, go to Kubernetes, find the required cluster and click its name.

Screenshot_2023-02-27_at_13.37_2.png

2. Go to the Pools section, click ⋯ next to the pool and select Change autoscaling limits.

Screenshot_2023-02-27_at_20.42_1.jpg

3. In the new window, enter the same number of nodes for minimum and maximum and click Save.

Autoscaling has been disabled. The pool will now maintain the specified number of nodes all the time, regardless of load.

Was this article helpful?

Not a Gcore user yet?

Discover our offerings, including virtual instances starting from 3.7 euro/mo, bare metal servers, AI Infrastructure, load balancers, Managed Kubernetes, Function as a Service, and Centralized Logging solutions.

Go to the product page
Edit on GitHub