The Design And Implementation Of Back-end Modules Of An Operations And Maintenance Knowledge Base System For Transwarp Products

Transwarp develops a great number of products based on big data and artificial intelligence technologies,which have been applied by many customers in a variety of industries.Typically,these products are operated and maintained by the personnel of both Transwarp and customer side.Due to the large number of product lines and the complexity of the products,the operations and maintenance personnel usually have to rely on internal systems(such as Wiki)to manage the knowledge of relevant products in order to complete corresponding operations and maintenance work.Using Wiki or other systems to provide technical documentation and user manuals is now the typical way for many IT companies to manage their operations and maintenance knowledge.In general,the contents of these knowledge are usually comprehensive introductions of the products,lacking more detailed knowledge about problems and solutions,thus it is difficult for these knowledge to provide adequate support for the personnel to work.Secondly,the systems such as Wiki have high requirements in terms of the structure and format for knowledge contents,which on one hand is not suitable for organizing a large amount of fragmented and detailed knowledge,and on the other hand reduces the motivation of the personnel to enrich the contents,as a result the problem-solving experiences are not preserved and cannot be reused.Finally,most of the current systems have some other shortcomings,such as unable to support the storage of large volumes of knowledge,unfriendly interactive ways,mechanical query functionality,difficult to obtain and share decentralized information.In response to above problems,this thesis designs and implements the back-end modules of an operations and maintenance knowledge base system for Transwarp products.Firstly,it makes use of the big data products developed by Transwarp to manage the operations and maintenance knowledge of various product details,allowing query tasks to be executed with high performance while maintaining a large volume of knowledge.Secondly,this system preserves the knowledge deriving from the problem-solving experience shared by operations and maintenance personnel,and integrates these konwledge with those extracted from Jira Issue,extending the source of the knowledge to products developers.Thirdly,by providing a series of processes such as submission,review,sharing,collection and query,the system ensures the accuracy of knowledge while facilitating the smooth flow and sharing of these knowledge within the company and among customer groups.Forthly,the system uses a full-text search engine to manage knowledge,and its query functionality is more flexible and powerful than traditional string pattern matching,which can improve the accuracy and validity of the query results.Finally,in order to improve the engagement of the operations and maintenance personnel,the system provides a variety of interactive ways,such as PC webpage,mobile lite version webpage,and We Chat chatbot based on the concept of Chat Ops.The system is based on microservices architecture to meet the requirements such as extensibility,usability and fault tolerance,and uses the Spring Cloud framework to govern microservices.This thesis highlights five back-end modules in the system that are closely related to the management of operations and maintenance knowledge from a design and implementation perspective,including the main service that responds to front-end requests and coordinates back-end services,search entrance and statistics service that receives and distributes query requests for products operations and maintenance knowledge and records historical information,problems solution management service that maintains solution knowledge with full-text search engine Search,Jira integration service that maintains integrated Jira Issue information,and We Chat chatbot implemented in Python.In addition,the system relies on a separate middleware service for docking with Jira.This system has been put into production,and basically met the needs of operations and maintenance personnel of both Transwarp and customer side with a lot of positive feedback.