ACP-Cloud1 Exam Dumps : ACP Cloud Computing Certification

Alibaba Cloud Server Load Balancer (SLB) is designed to distribute incoming network traffic across multiple ECS instances within Alibaba Cloud, enhancing fault tolerance and application reliability. Backend servers for SLB must be Alibaba Cloud ECS instances, as SLB is specifically optimized to work within the Alibaba Cloud ecosystem. While it does not support instances from other cloud providers, it ensures integration and performance with Alibaba Cloud’s ECS instances.

According to the Alibaba Cloud Academy, Auto Scaling supports three scaling modes: manual mode, scheduled task, and cloud monitor and event-triggered task. Manual mode requires manual intervention to adjust the number of ECS instances, which is not feasible and with minimum operation efforts. Scheduled task allows you to set the time and number of ECS instances to be added or removed in advance, which is suitable for predictable traffic changes. Cloud monitor and event-triggered task allows you to set the monitoring metrics and thresholds for scaling actions, which is suitable for dynamic traffic changes. Healthy check mode is not a scaling mode, but a feature of SLB to ensure the availability of ECS instances behind the load balancer. References: : ACP Cloud Computing Certification Preparation Course - Auto Scaling, : ACP Cloud Computing Certification Preparation Course - Server Load Balancer

 To execute a task at a specific time such as removing 1 ECS instance every night at 00:00, you need to perform two operations: create a scaling rule and create a scheduled task. A scaling rule is a set of instructions that defines how Auto Scaling scales computing resources in response to changes in business load. A scheduled task is a type of scaling task that can execute a specified scaling rule at a specified time. By creating a scaling rule that removes 1 ECS instance from the scaling group, and creating a scheduled task that executes this scaling rule every night at 00:00, you can achieve the desired result. You do not need to create a new scaling group, because you already have one scaling group created. You also do not need to create an event-triggered task, because this type of scaling task is triggered by events such as Cloud Monitor alarms or API calls, not by a specific time. References: 1, 2, and 3.