All Products
Search
Document Center

Support:AIOps Solution Expert Service

Last Updated:Oct 17, 2022

1. Overview

1.1. Introduction

With the continuous development and popularization of cloud computing, more and more enterprises have deepened their understanding of the cloud and actively implemented cloud transformation. Making full use of data stored in the cloud is drawing more and more attention. In the face of complex business systems on the cloud or across multiple clouds, O&M engineers often encounter problems such as complicated technology stacks, time-consuming alert configurations, missing metrics, alert storms, and time-consuming fault location. These problems may eventually lead to serious capital losses.

Artificial intelligence for IT operations (AIOps) combines big data and machine learning to improve the efficiency of IT operations. AIOps provides capabilities such as anomaly diagnostics based on time series metrics, root cause analysis, resource orchestration, and fault self-healing. An AIOps solution often requires the following processes: 1. Perform real-time anomaly detection based on various key performance indicators (KPI). 2. Analyze the sources of multi-dimensional indicators to quickly drill down to exception dimensions and elements. 3. Identify root causes based on application topology and real-time traces. 4. Build the context of anomaly root causes in conjunction with the Configuration Management Database (CMDB) to help quickly resolve issues.

AIOps Solution Expert Service provides solution support and consultation services to satisfy the technical requirements of enterprises for AIOps. Relying on years of AI practices and the experience of Alibaba Cloud experts, AIOps Solution Expert Service uses algorithms to build models and perform intelligent analysis on monitoring metrics in real time. If business exceptions occur, alert correlation analysis and convergence are performed in real time to help you reduce the mean time to recovery (MTTR) and improve business stability. AIOps Solution Expert Service incorporates the artificial intelligence, big data, and cloud computing capabilities and supports full-stack IT operations management. It helps enterprises automate IT operations, ensure business continuity, and improve overall efficiency.

AIOps Solution Expert Service provides capabilities such as time series trend prediction, risk exception inspection, intelligent diagnostics, AI monitoring, and intelligent root cause recommendation. It addresses various problems related to IT operations, such as separate systems, traditional methods, low efficiency, and low resource utilization. It serves as a full-stack monitoring and management platform that connects the underlying infrastructure and the upper-layer applications. In addition, it achieves automated IT operations by using the scenario-oriented intelligent operations capability. The features provided by AIOps Solution Expert Service can be tailored based on the requirements of the customer.

2. Service scope

2.1. Service scope of Consultation Pack

A consultation pack contains 10 consecutive business days of AIOps solution design carried out remotely. The following services are provided:

  • Business architecture research

Investigates and analyzes the current application technology stacks and resource usage by means of survey forms and interviews, and evaluates the feasibility of deploying an AIOps solution. Determines the quantity, priorities, and strategies of business systems. Provides suggestions on cloud technology selection based on the evaluation results.

  • Intelligent fault discovery solution

Designs an intelligent fault discovery solution for the customer based on the results of research and evaluation. The solution includes the following aspects: 1. Provides unified monitoring data access for multiple accounts. 2. Designs AI algorithm capabilities based on application groups for real-time fault discovery. 3. Identifies root causes based on the analysis results. 4. Provides a real-time anomaly detection solution to ensure business stability.

  • Customized business risk inspection solution

Provides a customized business risk inspection solution for the customer based on the business resources and technical capabilities of Alibaba Cloud. The solution includes a customized architecture for business risk inspection and an implementation plan for specific business risk scenarios based on the architecture.

The service pack does not include the following services:

  • The solution design focuses on technical components. It does not involve detailed analysis on the business and does not provide separate solutions for each system.

  • The consultation and design services are provided by deploying business systems on Alibaba Cloud. Alibaba Cloud does not provide consultation services for the overall cloud architecture of the customer. If the customer has such requirements, the customer shall purchase cloud architecture consultation services separately. Alibaba Cloud shall not be responsible for code development and diagnostics.

  • The customer (Party A) shall not limit the ways in which Alibaba Cloud (Party B) provides services. Alibaba Cloud conducts investigations and provides consultation services on site or remotely to produce the final deliverables.

  • Alibaba Cloud provides only Alibaba Cloud official documentation and the documents related to the intelligent fault discovery solution and the customized business risk inspection solution.

  • Alibaba Cloud shall not be responsible for the implementation or maintenance work involved in the planning, architecture design, and use of the business systems of the customer.

  • Alibaba Cloud shall not be responsible for troubleshooting or technical support of third-party software and application systems.

  • The implementation work required after the completion of the AIOps solution is not within the scope of the consultation service pack.

2.2. Service scope of Basic Pack

  • A basic pack contains 10 consecutive business days of implementation assistance carried out remotely. Alibaba Cloud (Party B) shall provide on-site support for no more than 1 time and no more than 2 person-days each time according to the project requirements.

  • The customer can select an AIOps solution based on the consultation results. The following services are provided:

    • Assists in activating cloud resources, creating cloud accounts, formulating an intelligent fault discovery solution, and formulating a customized business risk inspection solution.

    • Provides technical support, troubleshoots implementation-related issues, and provides solutions.

    • Assists in connecting applications to the AIOps solution.

The service pack does not include the following services:

  • The service pack does not promise to provide any deliverables. The service is terminated if the service period expires.

  • Alibaba Cloud (Party B) shall not be responsible for the following implementation work of the customer (Party A): application deployment, transformation of application code, transformation of data code, and data migration. The specific implementation work is carried out by the customer. In the implementation process, Alibaba Cloud is only responsible for providing technical support and guidance, and assisting the customer in resolving issues related to the use of Alibaba Cloud products.

  • Alibaba Cloud shall not be liable for schedule delays caused by the customer.

2.3. Service scope of Standard Pack

  • A standard pack contains 10 consecutive business days of implementation assistance carried out on site.

  • The customer can select an AIOps solution based on the consultation results. The following services are provided:

    • Assists in activating cloud resources, creating cloud accounts, formulating an intelligent fault discovery solution, and formulating a customized business risk inspection solution.

    • Provides technical support, troubleshoots implementation-related issues, and provides solutions.

    • Assists in connecting applications to the AIOps solution.

    • Optional. Assists in deploying an on-premises output platform and a visualization platform.

The service pack does not include the following services:

  • The service pack does not promise to provide any deliverables. The service is terminated if the service period expires.

  • Alibaba Cloud (Party B) shall not be responsible for the following implementation work of the customer (Party A): application deployment, transformation of application code, transformation of data code, and data migration. The specific implementation work is carried out by the customer. In the implementation process, Alibaba Cloud is only responsible for providing technical support and guidance, and assisting the customer in resolving issues related to the use of Alibaba Cloud products.

  • Alibaba Cloud shall not be liable for schedule delays caused by the customer.

3. Prerequisites

  • The customer shall submit a service request at least 15 business days before the customer places an order. This way, Alibaba Cloud can evaluate the business objectives of the customer and check the feasibility of the schedule to determine whether to accept the service request.

  • The customer shall provide Alibaba Cloud with all necessary documents, information, data, diagrams, system permissions, and remote access channels in a timely manner. All such information is subject to the confidentiality clauses attached to this statement of work. The customer shall guarantee that all information disclosed or to be disclosed to Alibaba Cloud is true, accurate, and not misleading.

  • Alibaba Cloud provides consultation services by using phone calls, DingTalk, and emails. The location where Alibaba Cloud provides services is not restricted by the project.

  • In the project delivery process, Alibaba Cloud designs the AIOps solution and troubleshoots the issues that occur during implementation, and the customer deploys and tests the applications.

  • The project managers designated by the customer and Alibaba Cloud shall use mutually agreed communication methods to transfer the written information required for the project. Available communication methods include DingTalk, Internet, and email.

  • All project deliverables are in Chinese (Simplified), and the working language is Chinese. All deliverables are submitted as electronic copies in Microsoft Office formats, including PowerPoint, Word, Excel, and Visio.

  • The customer and Alibaba Cloud shall work on the project in accordance with the work plan, staffing plan, and start and end dates that are agreed upon by both parties in advance. Alibaba Cloud shall not be liable for project delays that are caused by delays in the launch of the business systems of the customer.

  • Neither party shall be liable for special, incidental, or indirect damages, or consequential economic damages (including loss of profits or discounts) in this project, even if the party has been informed of the possibility of such damages.

  • The customer shall be responsible for the O&M that is related to its business.

4. Division of responsibilities

4.1. Customer and Alibaba Cloud

  • The customer and Alibaba Cloud negotiate to confirm the business objectives and service scope of AIOps Solution Expert Service.

  • After the contract is signed, the payment shall be completed.

4.2. Division of responsibilities

The following table describes the division of responsibilities at different phases of the project.

Phase

Party A

Party B

Project preparation

1. The customer shall appoint a project manager with the required expertise and experience to communicate with Alibaba Cloud. The project manager has full authority to make decisions on all aspects of the project on behalf of the customer, and shall be directly responsible for the planning, coordination, supervision, and control of project implementation. The project manager shall also be responsible for troubleshooting and resolving the issues that occur during project implementation.

2. The project manager shall cooperate with Alibaba Cloud engineers to confirm all matters in the project preparation phase (see "3. Prerequisites" in this statement of work).

3. The customer shall prepare the office environment and make sure that all relevant personnel are authorized to enter and leave the site.

4. The customer shall communicate with the personnel to be invested in each phase and obtain the required commitment and time.

5. The customer shall make sure that relevant personnel are properly managed.

1. Alibaba Cloud shall appoint an experienced project manager to communicate with the project manager of the customer, and manage the project and project team members of Alibaba Cloud.

2. Alibaba Cloud shall propose solutions and plans for all matters in the project preparation phase (see "3. Prerequisites" in this statement of work), confirm with the project manager of the customer, and record the confirmation in writing.

Investigation of current situation

1. The customer shall organize key users to participate in research interviews according to the project plan and interview plan.

2. The customer shall provide an overview of the existing business according to the research requirements of Alibaba Cloud, such as systems, applications, data, organizational structure, and division of labor.

3. The customer shall confirm the AIOps strategy and risk control strategy provided by Alibaba Cloud.

4. The customer shall designate the person to be responsible for reviewing deliverables, providing feedback, and confirming the acceptance.

1. Alibaba Cloud shall provide an interview plan and evaluate the existing infrastructure, application architecture, and application dependencies based on the interview results.

2. Alibaba Cloud shall propose an AIOps strategy and risk control strategy based on the evaluation results, and reach an agreement with the customer.

3. Alibaba Cloud shall make sure that the final deliverables meet the acceptance criteria based on the acceptance feedback of the customer.

Solution design

1. The customer shall cooperate with Alibaba Cloud in AIOps solution design.

2. The customer shall be responsible for the overall design of relevant solutions.

3. The customer shall designate the person to be responsible for reviewing deliverables, providing feedback, and confirming the acceptance.

1. Alibaba Cloud shall design an AIOps solution based on the service scope and the business scenarios of the customer.

2. Alibaba Cloud shall make sure that the final deliverables meet the acceptance criteria based on the acceptance feedback of the customer.

Solution implementation

1. The customer shall assist Alibaba Cloud to verify the feasibility of the solution and provide necessary business input, resources, and environment for solution verification. The customer shall also cooperate with Alibaba Cloud in code transformation and solution implementation.

2. The customer shall designate the person to be responsible for reviewing deliverables, providing feedback, and confirming the acceptance.

1. Alibaba Cloud shall assist and guide the customer to activate or purchase cloud resources and complete infrastructure construction and configuration.

2. Alibaba Cloud shall provide implementation support and troubleshooting services for the AIOps solution.

3. Alibaba Cloud shall build and verify the demo based on the AIOps solution.

4. Alibaba Cloud shall provide after-sales training services according to AIOps standards.

Note: A consultation pack includes the following phases: project preparation, investigation of current situation, and solution design. A basic or standard pack includes the following phases: project preparation, investigation of current situation, solution design, and solution implementation.

5. Service catalog

The following table lists the services that are provided by AIOps Solution Expert Service.

Phase

Service

Consultation pack

Basic pack

Standard pack

Investigation of current situation

System investigation and evaluation

Supported

Solution communication and planning

Supported

Solution design

Intelligent fault discovery solution

Supported

Customized business risk inspection solution

Supported

AIOps solution implementation

Implementation of the intelligent fault discovery solution

Supported

Supported

Implementation of the customized business risk inspection solution

Supported

Supported

AIOps solution on-site deployment

On-site support for the intelligent fault discovery solution

Supported

On-site support for the customized business risk inspection solution

Supported

Note: A basic pack and a standard pack provide the same services in different ways. The services of a basic pack are provided remotely whereas the services of a standard pack are provided on site. The on-site support service can be purchased separately.

5.1. Service content

AIOps Solution Expert Service

No.

Category

Description

Deliverable

1

Business architecture research

Fully investigates the resources that are used by the customer on the cloud, the business status of the customer, and the core logic of application systems. The research includes basic resource research, business status research, and application system research.

Project Research Report

2

Design of the intelligent fault discovery solution

Builds business group units based on business data and resource group dimensions. The metrics of the business group units are analyzed in real time by using intelligent AI algorithms to help the customer quickly identify faults. Suspicious root-cause events are identified based on the fault location algorithm. Based on the intelligent AI algorithm and years of AIOps experience, Alibaba Cloud provides an automated solution to resolve faults. Alibaba Cloud designs the solution for three types of algorithm scenarios: time series prediction, root cause analysis, and historical data prediction. For more information, see 10.1. Algorithms.

Intelligent Fault Discovery Solution

3

Design of the customized business risk inspection solution

Designs a customized architecture for business risk inspection and formulates an implementation plan for specific business risk scenarios based on the architecture. To facilitate solution implementation, Alibaba Cloud also provides two demos for risk inspection scenarios that are specific to the e-commerce business. For more information, see 10.2. Business risk scenarios.

Customized Business Risk Inspection Solution

AIOps Solution Basic Pack

No.

Category

Description

Deliverable

4

Implementation of the intelligent fault discovery solution

Provides an implementation plan based on the design solution and assists in connecting applications to the AIOps solution.

Intelligent AI Detection Implementation Plan

5

Implementation of the customized business risk inspection solution

Provides an implementation plan based on the design solution and assists in connecting applications to the AIOps solution.

Implementation Plan for Customized Business Risk Inspection

AIOps Solution Standard Pack

No.

Category

Description

Deliverable

6

On-site deployment of the intelligent fault discovery solution

Provides an implementation plan based on the design solution and assists in connecting applications to the AIOps solution.

Intelligent AI Detection Implementation Plan

7

On-site deployment of the customized business risk inspection solution

Provides an implementation plan based on the design solution and assists in connecting applications to the AIOps solution.

Implementation Plan for Customized Business Risk Inspection

6. SLA

The service level agreement (SLA) of AIOps Solution Consultation Pack includes the following items:

  • Consultation services for the AIOps solution.

  • Technical support DingTalk group and on-site support within the service provision period.

  • Documents such as "Customized Business Risk Inspection Solution" and "Intelligent Fault Discovery Solution". The deliverables vary depending on the service content.

7. Service process

Time limit: The customer shall submit a service request at least 15 business days before the customer places an order.

The following figure shows the process for providing the AIOps solution consultation service:3CC2BAFE-4EEC-450E-8243-101D00D93A1C

The following figure shows the process for providing the AIOps solution implementation service.

739FE2CA-9400-4D42-AD15-4EF5653C167E

8. Acceptance criteria

8.1. List of deliverables

No.

Phase

Deliverable

Deliverable type

1

Project research

Project Research Report

Document

2

Solution design

Intelligent Fault Discovery Solution

Document

3

Customized Business Risk Inspection Solution

Document

4

Solution implementation

Implementation Plan for Intelligent Fault Discovery

Document

5

Implementation Plan for Customized Business Risk Inspection

Document

8.2. Acceptance criteria

  • Acceptance criteria

    • The solution design meets the requirements of the customer and is signed and confirmed online. For more information about solutions, see 8.1. List of deliverables.

    • Alibaba Cloud (Party B) provides an intelligent fault discovery solution and a customized business risk inspection solution. After the solutions are accepted by the customer (Party A), the first phase of work is completed. The implementation work is carried out based on the accepted solutions. Alibaba Cloud provides technical support to assist the customer in implementing the solutions. After the solutions are implemented, the customer shall complete the acceptance within 5 business days by signing the Service Acceptance Report online.

8.3. Acceptance plan

In accordance with the deliverables of each project phase described in 8.1. List of deliverables, project acceptance is based on the following acceptance plan. The customer agrees to accept the deliverables submitted by Alibaba Cloud based on the acceptance plan.

No.

Acceptance milestone

Acceptance content

Acceptance completion

1

The design and verification of the AIOps solution are completed.

All deliverables in the project preparation, research and evaluation, and solution design phases

The customer confirms the Service Acceptance Report online.

2

The AIOps solution is implemented.

All deliverables in the solution implementation phase

The customer confirms the Service Acceptance Report online.

9. Mark of completion

The project is completed after the customer confirms the acceptance.

10. Appendix

10.1. Algorithms

Type

Algorithm

Logic

Anomaly diagnostics algorithm

One-Class SVM

Algorithm learning and anomaly diagnostics based on historical data

Anomaly diagnostics algorithm

Isolation Forest (iForest)

Algorithm learning and anomaly diagnostics based on historical data

Anomaly diagnostics algorithm

Robust Covariance

Algorithm learning and anomaly diagnostics based on historical data

Anomaly diagnostics algorithm

LocalOutlierFactor

Algorithm learning and anomaly diagnostics based on historical data

Anomaly diagnostics algorithm

AutoEncoder

Algorithm learning and anomaly diagnostics based on historical data

Root cause analysis algorithm

Random Forest and PCA

Root cause analysis

Time series anomaly diagnostics algorithm

K-Sigma

Anomaly diagnostics based on real-time time series data

Time series anomaly diagnostics algorithm

ARIMA

Anomaly diagnostics based on real-time time series data

10.2. Business risk scenarios

Type

Scenario

Description

E-commerce

Issue a purchase order

Compute scenario-specific metrics by analyzing logs and quickly implement customized business scenario inspection.

E-commerce

View the product page

Compute scenario-specific metrics by analyzing logs and quickly implement customized business scenario inspection.

E-commerce

Add to shopping cart

Compute scenario-specific metrics by analyzing logs and quickly implement customized business scenario inspection.

E-commerce

Shopping cart rendering

Compute scenario-specific metrics by analyzing logs and quickly implement customized business scenario inspection.

E-commerce

Buy page rendering

Compute scenario-specific metrics by analyzing logs and quickly implement customized business scenario inspection.

E-commerce

Make a payment

Compute scenario-specific metrics by analyzing logs and quickly implement customized business scenario inspection.

E-commerce

Check the payment result

Compute scenario-specific metrics by analyzing logs and quickly implement customized business scenario inspection.

E-commerce

Cashier desk rendering

Compute scenario-specific metrics by analyzing logs and quickly implement customized business scenario inspection.

E-commerce

Issue a purchase order

Compute scenario-specific metrics by analyzing logs and quickly implement customized business scenario inspection.

10.3. Expected results

  • Intelligent fault discovery solution

The solution includes root cause analysis for multiple metrics and products. The solution provides eight real-time anomaly detection and root cause analysis algorithms and ensures the generality of the algorithms.

  • Customized business risk inspection solution

The solution provides a customized architecture for business risk inspection and two demos for risk inspection scenarios that are specific to the e-commerce business.