tencent cloud

Event Overview

An event is an information carrier, which describes a change to a specific object at a specific time. An event is associated with a unique moment and a unique object. For example, an event occurs when the status of a computer changes from running to shutdown, when a program starts or stops running, or when a power failure occurs in an office building. An event records the moment when an object changes between different states.

The monitoring service helps find, locate, and solve problems to ensure the overall stability and performance of the system and services. Information and data are core drivers for OPS and monitoring.

As the main information source for monitoring, metric describes the value and trend of an object's specific variable over time in a threshold–like, periodic, and persistent manner. It shows the usage of resources and environment, while reflecting their health status. For example, the metric can show the CPU utilization of a Cloud Virtual Machine (CVM) instance and the number of slow query log of a database instance.
The metric covers most scenarios where problems are detected during monitoring. However, using metric as the only information source for monitoring has the following disadvantages:

• Information transmission delay. The threshold for a metric may or may not change after an event occurs for a time period.
• The accuracy and clarity of the transmitted information are low. It is difficult to determine and locate the real event or the cause of the problem by analyzing changes in threshold alone.
• Information carrier has limited forms. Metric is recorded in the form of threshold, and supplementary and incidental information cannot be provided as text.

Using event as a carrier of monitoring information can describe the running status of resources and underlying infrastructure and services in a more accurate and direct manner. This helps discover, locate, and solve problems more efficiently.
Advantage: events are result-oriented and can improve information accuracy, reduce delay, transmit more information, and enrich the dimension of monitoring information.

Event Center Overview

As the event information hub of Tencent Cloud, the Tencent Cloud Observability Platform (TCOP) Event Center packages important event information in the lifecycle and operation of Tencent Cloud resources and underlying infrastructure and services into a product, which constructs a complete event consumption channel and process to help customers with monitoring and OPS on the cloud.

The packaged event information provided by the Event Center is obtained from Tencent Cloud product modules and underlying infrastructure services. It is then aggregated, analyzed, condensed, and ultimately presented. Information sources are system logs and monitoring items under each module, which ensure the accuracy and value of the information passed on to customers.

Event Types

Based on information sources, causes, characteristics, and forms, events are categorized by the event center as product events and platform events, which are carried by two independent product pages and consumption processes.

• Product Event: product events are generated by resource instances and products (such as CVM instances) that are purchased and used by customers in Tencent Cloud. These events are directly or indirectly triggered by customers during use. They belong to specific resource instances. Customers can control and manage them. Resource instances affected by and associated with events can be explicitly determined.
• Platform Event: platform events are generated by the underlying infrastructure and services that support Tencent Cloud products, such as Virtual Machine Manager (VMM) that supports CVM at the virtualization layer and the underlying physical machines, networks, and storage modules. These events are generated or caused by the infrastructure and services of Tencent Cloud rather than customers. They belong to services. Customers cannot control platform events, which can only be managed by Tencent Cloud. Services or product modules affected by and associated with platform events can be determined, but the affected and associated resource instances cannot always be determined.

Event comparison

Event Center Use Case

Based on event information, the Event Center plans and provides comprehensive event consumption channels and capabilities.

• Display and tracing: records all event information related to the account. In the console, you can view, filter, search for, and trace event information.
• Alarm and notification: configures alarms for or subscriptions to monitored events and resources to promptly notify responsible personnel.
• Data pull by API: uses APIs to pull event information and connects it to Tencent's monitoring system.
• Automation: uses events as sources to trigger associated calls to support automatic OPS.

Value

• Event Center provides a wider range of resource monitoring information to offer comprehensive data support for monitoring and OPS.
• Event Center provides event information consumption channels to tap into the value of monitoring data.
• Event Center converges determination logic to efficiently and directly locate the affected resources and their causes.
• Event information can be traced and reviewed to understand important event changes in resources and platform lifecycles.
• Events can be used as triggers to automatically respond to specific resource and environment changes.