This topic describes how to configure alert rules for ApsaraDB for Redis in the Cloud Monitor console. Alerts are raised immediately after the specified thresholds are exceeded or events are detected. This way, you can handle anomalies at the earliest opportunity.

Description

Cloud Monitor is a service that can be used to monitor Alibaba Cloud resources and Internet applications. The service is an all-in-one enterprise-class monitoring service that can be used out-of-the-box. You can create alert rules and specify the metrics based on which the alerts are triggered. When the alert rules of a specified metric are triggered, alerts are raised and sent to contacts in an alert contact group.

Note Cloud Monitor sends alerts to contacts in contact groups. Before you add an alert contact to an alert group, you must create the alert contact or alert group. For more information, see Create an alert contact or alert group.

The following table describes the alert types that are supported by Cloud Monitor.

Type Service Metric Description Configuration
Threshold-based alerts
  • Standard instances
  • Cluster instances
  • Read/Write splitting instances
Metrics vary based on the instance type. Cluster instances and read /write splitting instances support metrics for data nodes and proxy nodes. For more information, see the metrics in the Cloud Monitor console. When the value of a metric exceeds the specified threshold, for example, the CPU utilization of an instance is greater than 90%, the system automatically sends an alert. This helps you understand and handle anomalies at the earliest opportunity.

In most cases, workloads are sensitive to fluctuations in the CPU utilization, memory capacity, and network traffic of the ApsaraDB for Redis instance. We recommend that you specify alert thresholds for the following key metrics:

  • CPU utilization: greater than 60%.
  • Memory usage: greater than 80%.
  • Inbound bandwidth usage and outbound bandwidth usage: greater than 80%.
Create a threshold-triggered alert rule
Event-based alerts N/A
  • InstanceMaintenance (automatic O&M events)
  • Exceptions
  • Maintenance
If an ApsaraDB for Redis instance fails, performs a master-replica failover, or runs an automatic O&M task, the system automatically sends an alert. This allows you to resolve these issues at the earliest opportunity. Create an event-triggered alert rule