How Scenario‑Based AIOps Transforms IT Operations: Insights from GOPS 2023

On November 18‑19, Dingmao Technology, a leading digital operations company, participated in the fourth GOPS Global Operations Conference where chief industry expert Xu Haitao delivered a talk titled “AIOps Scenario‑Based Construction and Delivery Exploration,” engaging deeply with attendees.

AIOps Construction Challenges and Breakthrough Methods

Common difficulties such as missing or unstandardized logs, lack of business monitoring, unreadable alerts, and missing operational tools stem from deficiencies in data, algorithms, and automation. Dingmao proposes a scenario‑based construction approach to address these issues within existing resources.

Definition and Elements of AIOps Scenarios: From Standard to User Scenarios

AIOps scenarios are defined by progressing from standard scenarios to user‑specific delivery, incorporating pain points, suitability analysis, condition matching, logical execution, input data, and output results to solve user challenges.

Example: Root‑Cause Localization Scenario

For container‑cloud microservice failures, the scenario includes:

Applicable Situation : Fast root‑cause localization when applications exhibit slow response, errors, or inaccessibility.

Implementation Conditions : Presence of container management platforms, APM monitoring, container platform monitoring, basic infrastructure monitoring, and a CMDB providing full dependency information.

Process : Triggered by application alerts, the root‑cause service runs real‑time models using monitoring metrics, alerts, call chains, and dependency data to identify the most likely cause.

Input Data : Time‑series metrics, alerts, call chains, and dependency relationships of applications, containers, hosts, and network devices.

Output : Ranked list of probable root‑cause metrics and alerts.

Technical Methods for Scenario‑Based AIOps Construction

Utilizing streaming engines, batch task engines, unified search, edge nodes, AIOps operators, and algorithm management as foundational capabilities, Dingmao builds data mapping models, offline update models, and algorithm orchestration to support end‑to‑end scenario construction from data flow to visual presentation.

Planning Scenario Roadmap and Implementation

The roadmap starts with statistical, weakly‑linked data techniques to address specific pain points, then aligns with customers' IT governance evolution, gradually expanding to comprehensive, precise, and intelligent operations across all domains.

Supporting Technologies

Dingmao’s ARCANA platform provides full OPS data support, covering edge data collection, storage, and forwarding, as well as analysis, AI operator management, dashboards, reporting, API interaction, alerting, and robust security controls.

Future Outlook and Exploration

By leveraging AI to accelerate operational data support, Dingmao aims to continuously innovate through scenario‑driven technology, delivering more efficient and accurate intelligent operations services.