tencent cloud

Auto Scaling

Product Introduction
Overview
Strengths
Scenarios
Service Limits
Access Management
Purchase Guide
Getting Started
Creating a Scaling Plan in 5 Minutes
Step 1:Creating a Launch Configuration
Step 2:Creating a Scaling Group
Step 3:Creating a Scaling Policy
Tutorial
Creating Web Services
Creating High-performance Computing Cluster
Creating Servers for Sending Requests
Configuring For High Availability Services
Launch Configuration
Launch Configuration Overview
Creating a Launch Configuration
Viewing Launch Configuration List
Modifying Launch Configurations
Exception Detection
Scaling Groups
Scaling Group Overview
Creating a Scaling Group
Viewing Scaling Group List
Modifying Scaling Groups
Modifying Bound Instances
Adding CLBs
Delete scaling group
Creating Notifications
Creating Scaling Activity Notifications
Specifying Notification Recipients
Expanding and Reducing Capacity
Auto Scaling Lifecycle Hook
Managing Scheduled Actions
Managing an Alarm-triggered Policy
Instance Health Check
Expanding Capacity Manually
Scale-in Process
Viewing Scaling Activities
Suspending and Resuming Scaling
Scale-in Removal Protection
Scaling Activity Cancelled
Scaling Activity Failed
Cooldown Period
Monitoring
Monitor Alarm Metrics
Running Status Check
Scheduled Task
Scheduled Task Execution Time Description
Troubleshooting
Scaling Group Doesn't Work
Scaling Activity Failed
Use Cases
Setting a Fixed Outbound IP Address for Scaling Groups
Deploying Elastic Web Application Service
Enabling Services Upon CVM Start-up
Customer Cases
iCarbonX
API Documentation
History
Introduction
API Category
Making API Requests
Scaling Group APIs
Scaling Group Instances APIs
Launch Configuration APIs
Scheduled Action APIs
Alarm Trigger Policy APIs
Notification APIs
Lifecycle Hook APIs
Instance Refresh APIs
Other APIs
Data Types
Error Codes
FAQs
Contact Us
Glossary

Monitor Alarm Metrics

PDF
Modo Foco
Tamanho da Fonte
Última atualização: 2025-04-03 16:49:43
By using Tencent Cloud Cloud Monitoring, you can retrieve statistical data based on an ordered set of time series data (called metrics). You can use these metrics to verify whether your system is running as expected. If the thresholds are reached, scaling will be triggered.

AS monitoring metrics

AS currently supports the following metrics:
CPU utilization
Memory utilization
Private network bandwidth in
Public network bandwidth in
Private network bandwidth out
Public network bandwidth out
TCP established connection count
Each metric supports the following values:
Maximum value
Minimum value
Average value

Metrics aggregation method

Tencent Cloud Auto Scaling monitors CVM clusters, and collects multiple monitoring metrics for multiple CVMs across multiple time periods. This data is aggregated and used to inform operations based on the policies that you configure.
It collects one sample per minute for each of CVM based on the set monitoring items. If the sample value meets the configured rules for multiple periods consecutively (the number of periods can be customized), the alarm scaling activity will be triggered.
For example: A scaling group has 5 CVM instances and the defined alarm scaling policy is “Maximum/minimum/average CPU utilization of more than 50% within a 5 minutes period for 3 times consecutively.” Auto Scaling collects the data and determines whether the policy for the alarm scaling rule is met. The steps are as follows:
1. Collects one value from each CVM per minute. In one sampling period (the current period is 5 minutes) it takes 25 CPU utilization values.
2. According to the configured maximum/minimum/average values, it determines whether the policy for the alarm scaling rule is met.
Maximum value: If the maximum value in these 25 values exceeds the threshold (50%), this period meets the alarm scaling rule.
Minimum value: If the minimum value in these 25 values exceeds the threshold (50%), this period meets the alarm scaling rule.
Average value: If the average value of these 25 values exceeds the threshold (50%), this period meets the alarm scaling rule.
3. If the rule is met for 3 consecutive periods (a total of 15 minutes, with each 5-minute period considered the current period), the scaling action is triggered.

Ajuda e Suporte

Esta página foi útil?

comentários