Event monitoring covers cloud service faults, O&M events, and user business exceptions. It provides event statistics by service, level, name, and application group, to facilitate associated businesses and fault review. You can customize the receivers and methods of event notification to prevent key events from being ignored. The event details help you locate faults.
Cloud service events
Event monitoring provides you with a centralized platform to summarize and query system events that are generated by different types of cloud services. It enables you to track the use of cloud services.
After you classify resources into application groups, service-related system events are automatically associated with the resources of those groups. This helps you integrate monitoring information, and quickly analyze and troubleshoot problems.
Event monitoring also provides the event alert function. You can configure alerts based on the event level, notification through text messages, emails, or DingTalk Chatbot, or alert callbacks. With these configurations, you can be notified of critical events immediately after they occur and handle the events in an automated online O&M process.
Event monitoring provides you with query and alert services for cloud service faults and O&M events.
- ECS events: critical ECS system events, such as unexpected restarts or disk performance degradation that are caused by system or instance errors
- SLB events: HTTPS certificate expiration events
- OSS events: The upstream or downstream bandwidth that is used by a bucket has exceeded the throttling threshold or report threshold.
- Auto Scaling events: successful or failed scale-in and scale-out of Auto Scaling
- E-MapReduce events: cluster creation failure, timeout, and service component status
The difference between custom event monitoring and custom monitoring is as follows:
- Custom event monitoring focuses on the data of non-continuous events.
- Custom monitoring focuses on periodically collected time series data.
Alert service and automated O&M
Event monitoring provides multiple alert methods for automated O&M.
- Alert notification: Alert notification can be sent through emails, DingTalk Chatbot, or other channels.
- MNS queue: Events can be written to the MNS queue, which can then be connected to your own O&M system.
- Function Compute: Events can trigger Function Compute to process subsequent O&M logic.
- Alert callback: Alert notification is pushed to the public URL of your existing O&M system or message notification system through HTTP POST requests. You can then handle the received alert notification based on its contents.