Skip to content

Monitor Troubleshooting

Introduction

This document will guide you through troubleshooting common monitor issues:

  • The monitor does not generate events
  • The monitor generates events but does not send notification messages

Prerequisites

  • You need an account with permissions to create monitors in Guance
  • You need backend management permissions for Guance

Troubleshooting Steps

Step One: Create a Test Monitor

  • Log in to the Guance console
  • Select "Monitoring" - "Create" - "Threshold Detection"

  • Detection Metrics:

M::`cpu`:(COUNT(`usage_guest`)) BY `host`
  • Trigger Condition: Critical Result > 0

  • Event Title: test

  • Event Content: test

  • Alert Strategy: Choose the strategy you have set up

  • Save

Step Two: Obtain the Automatic Trigger Configuration ID

  • Log in to the Guance console and select "Monitoring" - open browser debugging mode

  • Click on the monitor name link

  • Obtain the automatic trigger configuration ID

Step Three: Log in to the Func Platform to Query Issues

  • Launcher to get the Func login address

  • Log in to your func platform

  • "Manage" - "Automatic Trigger Configurations" - "Select Show All" - "Enter the ID from Step Two" - "Confirm Search" - "Recent Executions"

  • You can view detailed error information via "Recent Executions"

Step Four: Check Alert Notification Logs for Troubleshooting

Note

It is recommended to execute kubectl rollout restart -n middleware deploy message-desk-worker message-desk, then proceed with testing and troubleshooting.

The message-desk-worker service is the alert notification module of Guance, responsible for sending notifications such as DingTalk bot notifications, email notifications, Lark bot notifications, etc.

  • namespace: middleware
  • deployment: message-desk-worker
  • log path: /logdata/bussiness.log

Feedback

Is this page helpful? ×