Alibaba Auto Scaling

Introduction

Auto-scaling on Alibaba Cloud is the notion of automatically provisioning and de-provisioning extra ECS instances to optimise for performance and running costs. When you experience a peak in demand and trigger a scaling task either based on a schedule or a performance metric, Alibaba Cloud can automatically provision a new instance of an ECS based on a select image and add the instance to a pool of ECS compute resources to share the workload. Similarly, when this demand is no longer a concern, autoscaling can be configured to safely de-provision the extra ECS instances that are no longer in demand. This scaling out and in based on ECS node counts is also referred to as horizontal scaling.

Scaling Tasks

There are two types of tasks available in auto-scaling. Scheduled Tasks and Event-Driven Tasks.

Scheduled Tasks operate on a schedule. If you know to expect a particular increase in demand at certain points throughout the week, you can create a task to preemptively provision extra ECS instances. By using Scheduled Tasks you have the extra ECS instances up and ready before you experience the extra demand.

Event-Driven Tasks are triggered when a particular metric is met. If your average CPU load exceed 70% you can create Event-Driven Tasks to provision one or more extra ECS instances. Auto-scaling can be used in conjunction with the CloudMonitor agent installed on ECS instances to expose more metrics that can be used for Event-Driven Tasks.

Setup

Before you can start working with the Auto Scaling functionality on the Alibaba Cloud, you need to authorise the Auto Scaling service. You can do this by logging on to the Auto Scaling console (accessible through here). Once you are in the Auto Scaling console you need to click on Authorise – this will take you to the RAM console to grant this access. From the roles select the AliyunESSDefaultRole and click on Agree. When you then return to the Auto Scaling console, you need to click refresh and your console will be ready for use.

Benefits

Flexible
Intelligent scheduling to support various scenarios.
Automatically adjusts computing resources according to various scaling policies.
Supports manual scale-in and scale-out, which offer you the flexibility to control resources manually.

Automatic
Automatic monitoring of health status.
Automatically create or release ECS resources after scaling policy is triggered.
Automatically configures load balancer and RDS whitelist.

Elastic Scale-Out and Scale-In
During peak periods, automatically adds additional computing resource to the pool
When user requests decrease, Auto Scaling automatically releases ECS resources to cut down your costs

Elastic Self-Health
When an unhealthy instance has been detected, the auto-scaling service automatically replaces the instance with a new one to ensure uninterrupted service

More

For installation instructions on the CloudMonitor agent see the following guide provided by Alibaba Cloud.