Design And Implementation Of Index Resource Scheduling And Updating System

Nowadays, the O2O scene, as one of the most popular directions of the Internet, developing rapidly. More and more offline businesses provide customers with service by the platform of 020. As the counts of business increasing, people cannot find the businesses which they interest in in an effective way. Search engine provides a convenient service that looking for businesses for the costumes. The search engine in O2O, with its own unique feature, has to provide a service that has to ensure the integrity of information of results. How to ensure that service has became a grave subject.The thesis introduces resource scheduling and index updating system of vertical search engine in scene of O2O. This system can not only ensure the consistency of index across multiple machine room, but also be taken to ensure that pull index work is completed in a relatively short period of time in terms of timeliness, while handling amount of one hundred million data. Moreover, the system can be automated to complete the service offline, the index automatically pull, service monitoring work. The system ensures stable operation of online services, ensures the provision of a reliable search service. The system is completed by Python +Lua, which the Lua scripting language provide support for BS services, because Lua can be seamlessly embedded in C++. The Functional requirements of system is abstracted to be hierarchical, at the same time, it is abstracted to be several modules on business. The designation ensure better scalability and performance of the system stability. The system can not only guarantee the integrity of the index reconstruction, but also can guarantee the index's pull and push in a relatively short time.The system ensures the rational allocation of resources of vertical search engines effectively and timeliness and consistency of index updates. The system has been developed to adapt 020 scenarios. It has been applied to the actual production environment, and has a positive impact on the overall retrieval services. The system provides an automated management solution without affecting the performance of the retrieval system.