Overview

ZStack offers auto-scaling capabilities that let you automatically add or remove VM instances from an auto scaling group (ASG) in response to load balancing of VM instances, your business load changes, and predefined scaling policies. With the auto scaling service, you can better leverage the Cloud resources, reduce the O&M costs, and ensure smooth business operations. Currently, the auto scaling service is applicable to KVM VM instances.

Scaling Mode

Currently, the Cloud supports the following two types of scaling mode:
  1. Auto Scaling
    • Auto scaling includes elastic scale-out and elastic scale-in. For the elastic scale-out, when your businesses are growing, VM instances will be automatically added to ensure your business continuity. For the elastic scale-in, if your businesses decrease, VM instances will be automatically reduced.
    • With the ZWatch monitoring alarm, the auto scaling mode can be triggered. You can select an endpoint type as needed, including email, DingTalk, HTTP application, short message service (SMS), Microsoft Teams, and system endpoint.
    The auto scaling mode is shown in Auto Scaling.
    Figure 1. Auto Scaling


  2. Elastic Self-Health
    • In the elastic self-health mode, an auto scaling group monitors the health status of the VM instances within the group, automatically removes the unhealthy VM instances, and automatically adds additional VM instances. In this regard, healthy VM instances within the group will be ensured to be automatically adjusted not lower than the minimum specified number of VM instances.
    • Two types of health check are provided to trigger the elastic self-health, including load balancer health check and VM health check. If an auto scaling group configures the load balancing service, we recommend that you select the health check mechanism native to a load balancer.
    The elastic self-health is shown in Elastic Self-Health.
    Figure 2. Elastic Self-Health


Trigger Mechanism for Auto Scaling Group

The following illustrations show the scale-out trigger conditions:
  1. When the workload of VM instances in an auto scaling group exceeds the specified thresholds, the scale-out policy is triggered.
    • If the auto scaling group starts to calculate the cooldown time, a scale-out activity is suspended.
    • If the number of the VM instances in the auto scaling group has reached the specified maximum capacity limit, additional VM instances are no longer added.
    • The number of the VM instances in the auto scaling group must not be greater than the specified maximum number of the VM instances.
  2. When the number of the VM instances in an auto scaling group is less than the specified minimum number of the VM instances, the scale-out policy is triggered.
    • If the number of the VM instances in the auto scaling group has reached the specified maximum capacity limit, additional VM instances are no longer added.
    • The number of the VM instances in the auto scaling group must not be greater than the specified maximum number of the VM instances.
The following illustrations show the scale-in trigger conditions:
  1. When the workload of the VM instances in an auto scaling group is less than the specified thresholds, the scale-in policy is triggered.
    • If the auto scaling group starts to calculate the cooldown time, a scale-in activity is suspended.
    • If the number of the VM instances in the auto scaling group has reached the specified minimum capacity limit, VM instances are no longer removed.
    • The number of the VM instances in the auto scaling group must not be lower than the specified minimum number of the VM instances.
  2. When the number of the VM instances in an auto scaling group exceeds the specified minimum number of the VM instances, the scale-in policy is triggered.
    • If the number of the VM instances in the auto scaling group has reached the specified minimum capacity limit, VM instances are no longer removed.
    • The number of the VM instances in the auto scaling group must not be lower than the specified minimum number of VM instances.
Trigger conditions for elastic self-health: unhealthy VM instance
  • Notice that the Cloud will delete VM instances detected as unhealthy. After these VM instances are deleted, if the number of the VM instances in the auto scaling group is less than the specified minimum number of the VM instances, the scale-out policy is triggered. Hence, VM instances will be automatically added.

Typical Usage Scenario

Three types of typical usage scenario about the auto-scaling feature are introduced as follows:
  • Elastic scale-out:

    An online retailer has initiated a sales promotion such as red envelop fever and limited-time offer during Double Eleven, the Spring Festival, and other festivals and shopping sprees. As the business workload is skyrocketing, VM instances need to be automatically scaled out in time to avoid access delay and excessive resource loads.

  • Elastic scale-in:

    After the festivals and shopping sprees, the business workload of the retailer obviously declines. In that case, VM instances need to be automatically removed in time to avoid a waste of resources.

  • Elastic self-health:

    To ensure that the core businesses of the retailer are running normally, the number of the healthy VM instances must not be lower than some threshold.

Download

Already filled the basic info? Click here.

Enter at least 2 characters.
Invalid mobile number.
Enter at least 4 characters.
Invalid email address.
Wrong code. Try again. Send Code Resend Code (60s)

An email with a verification code will be sent to you. Make sure the address you provided is valid and correct.

Download

Not filled the basic info yet? Click here.

Invalid email address or mobile number.

Email Us

contact@zstack.io
ZStack certification training consulting
Enter at least 2 characters.
Invalid mobile number.
Enter at least 4 characters.
Invalid email address.
Wrong code. Try again. Send Code Resend Code (60s)

Email Us

contact@zstack.io

The download link is sent to your email address.

If you don't see it, check your spam folder, subscription folder, or AD folder. After receiving the email, click the URL to download the documentation.

The download link is sent to your email address.

If you don't see it, check your spam folder, subscription folder, or AD folder.
Or click on the URL below. (For Internet Explorer, right-click the URL and save it.)

Thank you for using ZStack products and services.

Submit successfully.

We'll connect soon.

Thank you for choosing ZStack products and services.

Back to Top