All Products
Search
Document Center

Auto Scaling:Monitoring and logging

Last Updated:Jun 18, 2024

Monitoring and logging are crucial for ensuring that your Auto Scaling resources are available and that your business operates normally and stays healthy. Auto Scaling provides a range of monitoring and log audit services, including CloudMonitor, Cloud Config, and event notification. These services assist in real-time monitoring of resource usage and instance status within your scaling group, enabling quick responses to exception alerts. You can utilize the monitoring capabilities to gather continuous monitoring data, enabling real-time monitoring of cloud resource usage and business health.

Scaling group monitoring

Scaling groups continually report their operational status to CloudMonitor in real-time. The scaling group monitoring feature provides visibility into the fluctuations of the numbers of instances in different states within a scaling group, offering insights into resource utilization over a specified period. For more information, see View instance number changes in a scaling group.

Instance monitoring in a scaling group

Scaling groups continuously monitor the performance metrics and health status of instances within the group, enabling you to stay informed about instance status and quickly address any instance exceptions to maintain high resource availability.

  • Performance metric monitoring

    Scaling groups utilize CloudMonitor to capture real-time instance performance metrics, such as CPU utilization. Scaling groups then process and aggregate the instance performance metrics to create scaling group-specific metrics. You can configure automatic resource scaling based on performance metrics to maintain the metric values within your intended range.

  • Health status check

    Auto Scaling supports the health check feature, which regularly assesses the running status of ECS instances or elastic container instances. Auto Scaling can automatically create new instances to replace any unhealthy instances.

    Auto Scaling supports three health check modes, including custom health check, instance status health check, and load balancer health check. You can select one or more health check methods based on your specific needs.

Event monitoring in a scaling group

When a scaling activity or O&M event occurs in a scaling group, Auto Scaling sends a message to the specified notification receiver, informing you of the event. For example, Auto Scaling can send a message to CloudMonitor, a Message Service (MNS) topic, or an MNS queue. After being informed about the event in CloudMonitor or MNS, you can perform manual O&M operations or trigger automatic O&M operations in response to the event. For more information, see Manage event notification rules.