#Concepts

#Monitoring

#Metrics

Metrics are used to quantitatively describe the operating status of a system, and each metric consists of four basic elements:

• Metric Name: Used to identify the monitored object, such as cpu_usage
• Metric Value: Specific measurement value, such as 85.5
• Timestamp: Records the time of measurement
• Labels: Used for multidimensional data classification, such as {pod="nginx-1", namespace="default"}

#PromQL

PromQL is the query language for Prometheus, used to query and aggregate metric data from the monitoring system.

#Built-in Indicators

The platform has preset a series of commonly used monitoring metrics based on long-term operational experience. You can directly use these metrics when configuring alarm rules or creating monitoring dashboards without additional configuration.

#Exporter

The Exporter is a component for collecting monitoring data, with primary responsibilities including:

• Collecting raw monitoring data from the target system
• Transforming data into a standard time-series metric format
• Providing metric data for querying via HTTP interface

#ServiceMonitor

ServiceMonitor is used to declaratively manage monitoring configurations and primarily defines:

• The selection criteria for monitoring targets
• Configuration of metric collection interfaces
• Execution parameters for collection tasks (intervals, timeouts, etc.)

#Alarms

#Alarm Rules

Alarm rules define the specific conditions for triggering alarms:

• Alarm Expression: Describes the conditions for triggering an alarm using PromQL statements
• Alarm Threshold: Explicit boundary values for trigger
• Duration: Duration for which the conditions must be continuously met
• Alarm Level: Distinguishes the severity of alarms (e.g., P0/P1/P2)

#Alarm Policies

Alarm policies organize multiple alarm rules together for unified configuration:

• Alarm Targets: The target scope of the rules
• Notification Method: The channels for sending alarms
• Sending Interval: The time interval for repeated alarm notifications

#Notifications

#Notification Policies

Notification policies manage the rules for sending alarm messages:

• Recipients: Target users for alarm notifications
• Notification Channels: Supported message sending methods
• Notification Templates: Definition of message content format

#Notification Templates

Notification templates customize the display format of alarm messages:

• Title Template: Format of the alarm message title
• Content Template: Organization of alarm details
• Variable Replacement: Supports dynamic data filling

#Monitoring Dashboard

#Dashboard

A dashboard is a collection of multiple related panels, providing an overall view of the system status. It supports flexible layout arrangements and can organize panels in rows or columns.

#Panels

Panels are visual representations of monitoring data, supporting various display types.

#Data Sources

The configuration of monitoring data sources. Currently, only the monitoring components of the current cluster are supported as data sources, and custom data sources are not supported for now.

#Variables

Variables serve as placeholders for values and can be used in metric queries. Through the variable selector at the top of the dashboard, you can dynamically adjust query conditions, allowing chart content to update in real-time.

#Perses Monitoring Dashboard

#PersesDashboard

PersesDashboard is a monitoring dashboard resource introduced in Alauda Container Platform 4.4. It is based on the open source Perses rendering engine and is used by Operations Center > Monitor > Dashboards (Perses).

PersesDashboard coexists with the existing MonitorDashboard resource. MonitorDashboard uses the existing Grafana-style dashboard model, while PersesDashboard uses the Perses model. Existing MonitorDashboard resources can be migrated to PersesDashboard resources when you want to render them in the Perses dashboard entry.

#Panel

A panel is a single visualization unit in a Perses dashboard, such as a time series chart, bar chart, gauge chart, or table.

#Panel Group

A panel group organizes panels within a dashboard. Panel groups can be used to structure dashboards and collapse or expand related panels.

#Variable

A variable is a dashboard parameter used to dynamically filter panel queries. Perses dashboards support list variables and text variables.

#Data Source

The metric data source for Perses dashboards is fixed to the platform metric data source and is transparent to users. You do not need to select a data source when viewing or editing a Perses dashboard.

#Folder and Tag

Folders and tags are used to organize and search Perses dashboards.