Skip to content

What is Insight

Insight is a next-generation, cloud native observability platform that focuses on application analysis. Its out-of-the-box real-time resource monitoring, metrics, logs, events, and other data assist enterprises in analyzing application status. Furthermore, it is compatible with popular open-source components, providing alerting, multi-dimensional data visualization, fault locating, and one-click monitoring and diagnosis.

Insight implements unified collection of metrics, logs, and traces, supports multi-dimensional alerting on metrics and logs, and provides a concise and clear visual management interface.

Insight observability architecture

Key Capabilities

  • Dashboards

    Built-in dashboards based on native Grafana covering platform components, cluster resources, and core components, with support for importing custom templates.

    ▶ View dashboards

  • Infrastructure Monitoring

    Centralized multi-cluster observability across clusters, nodes, namespaces, and containers, with event query and black-box probing.

    ▶ Cluster monitoring

  • Metric & Log Query

    Metrics support basic and advanced PromQL queries; logs support keyword search, Lucene syntax, context viewing, and download.

    ▶ Metric query

  • Distributed Tracing

    Traces are collected through OpenTelemetry, providing service topology, service lists, and trace analysis for rapid fault locating.

    ▶ View traces

  • Alert Center

    Over 100 built-in alert rules, custom PromQL rules, message templates, silence and inhibition, with notifications via email, WeChat Work, DingTalk, and more.

    ▶ Alert center

  • Collection & System Config

    Manage insight-agent and metric/log collection tasks, and customize the retention duration of metrics, logs, and traces.

    ▶ Log collection and query ▶ System configuration and retention

Workflow

Using Insight usually involves two stages: deployment and integration followed by observation and alerting. Administrators plan resources and install insight-agent; business users then integrate application observability, view dashboards and query data, and configure alerts to detect anomalies in time.

flowchart LR
    A["① Resource planning<br/>Insight · insight-agent"] --> B["② Install insight-agent<br/>Online · offline"]
    B --> C["③ Integrate observability<br/>OpenTelemetry"]
    C --> D["④ Dashboards & queries<br/>Dashboards · metrics · logs"]
    D --> E["⑤ Configure alerts<br/>Rules · notifications"]

    click A href "../quickstart/res-plan/"
    click B href "../quickstart/install/install-agent/"
    click C href "../quickstart/otel/otel/"
    click D href "../user-guide/dashboard/dashboard/"
    click E href "../user-guide/alert-center/alert-policy/"
Step Role Description
① Resource planning Platform Administrator Evaluate the node and storage resources required by Insight and insight-agent
② Install insight-agent Platform Administrator Install the collection components online or offline on observed clusters, and upgrade as needed
③ Integrate observability Business User Collect traces and metrics through OpenTelemetry (non-intrusive Operator or SDK)
④ Dashboards & queries Business User Understand application and cluster status through dashboards, metric and log queries
⑤ Configure alerts Platform Administrator Create alert rules and notification configs to detect anomalies in time

Quick Start

1 Resource planning Evaluate node and storage resources Go → 2 Install insight-agent Online · offline installation Go → 3 Integrate observability Non-intrusive OpenTelemetry Go → 4 View dashboards Metrics · logs · traces Go →

Basic concepts

The basic concepts related to observability (Insight) are as follows.

# Terms Definitions
1 Target The monitored object; the system will regularly initiate capture tasks to the monitoring point to obtain metrics
2 Metric Use the open-metric format description to measure the degree of a certain attribute in a software or hardware system
3 Recording Rule A named PromQL expression, which is a new metric obtained by calculating multiple metrics to describe a more complete and complex system state
4 Dashboard Dashboard is a form of visual management, that is, the performance of data, intelligence and other conditions at a glance. It displays information through various visual perceptions with intuitive images and appropriate colors. Display the real-time situation of the platform and all performance metrics in DCE through visual graphics.
6 Service Discovery A service discovery configuration for Kubernetes environment, used to batch and automatically access monitoring points on Kubernetes
7 Exporter A service that can provide metrics, often understood as a monitoring object
8 Rule A PromQL expression whose return value is a Boolean value, which describes whether the metric or custom metric is within the threshold range, and if not, an alert event will be generated
9 Alert The record information when the alert rule is triggered, which records the alert rule, trigger time, and current system status; at the same time, it will trigger the corresponding action, such as sending an email
10 Notification The alert event information sent by the system to the user through e-mail or other channels
11 PromQL Query statements supported by the Prometheus system

More Resources

  • Feature List

    A complete list of Insight features across dashboards, infrastructure, metrics and logs, tracing, and the alert center.

    ▶ View the feature list

  • Use Cases

    Learn about unified multi-cluster data collection and observability, rapid fault locating and troubleshooting, and other typical use cases.

    ▶ View use cases

  • Benefits

    Learn about the advantages of Insight in unified collection, out-of-the-box experience, and ecosystem compatibility.

    ▶ View benefits

  • Permissions

    Learn about the user roles and permissions required for each Insight feature.

    ▶ View permissions

Download Center Learn about d.run Token Factory Learn about d.run AI OS View the Insight User Manual (PDF)

Comments