| The informatization of large-scale pig breeding and production in my country has been initially completed,and a large number of indicators will be generated in the process of pig breeding and production to provide enterprises with decision-making.In view of the increasing number of live pigs in breeding companies,large-scale historical business data already exists,and hundreds of thousands to one million operations occur on the farm every day,making it difficult for on-site staff to manually maintain huge production data,which affects the interior of the pig farm.The personnel check the pig business and production status.To this end,this paper designs and develops a high-performance parallel computing offline reporting system based on distributed technology.This article first takes the pig production data of the report system as an entry point to sort out the business and indicators of the report system.Then,based on the actual task requirements,design and implement the pig breeding report system based on the distributed computing framework.Aiming at the problem of time-sharing reading causing excessive database IO and causing a serious burden on the database,the system proposes a method to separate data reading and writing and logic processing;use hash storage and producer-consumer mode in report calculation to be efficient Algorithms to improve the speed of data search and the performance of parallel computing;a new Redis interface operation algorithm is designed for any type of data reading and writing problems in Redis,which makes the program operate on Redis more concise and efficient;for each step of report calculation,There is a corresponding message log,and the messages generated by important node information are directly sent to the designated work group through the Ding Talk group management robot message push,which is convenient for staff and managers to view and maintain.At present,the pig breeding report system has been developed and put into actual production development projects.Compared with the existing algorithm framework,the memory allocation of the proposed system development scheme is more reasonable,the program execution speed is faster,and the system implementation effect is better,which can well meet the needs of enterprises. |