Research On Data Storage Strategies For Internet Of Things

The Internet of Things(Internet of Things,IoT)is a huge network that connects massive smart sensing devices.In recent years,the Internet of Things technology has been widely used in more and more industries.With the expansion of application fields and scale,the number of sensing devices in the Internet of Things is also increasing,Sensor equipment continuously collects data and sends it to the data center,which constitutes big data of the Internet of Things.There are two main types of IoT data,one is structured data collected by numeric sensors,and the other is unstructured data such as pictures and videos collected by non-numeric sensors such as cameras.Massive IoT big data is a huge challenge for storage systems.In order to solve the above problems,here proposes a hybrid technology IoT data storage strategy to provide solutions for data storage problems with similar needs in a heterogeneous big data environment.For this reason,the following work has been done in this thesis:First,through literature research,understand the main characteristics of IoT data.According to the literature survey,it summarizes the four characteristics of the Internet of Things data: massiveness,real-timeness,timing and relatively high degree of structure,and some periodicity.On the basis of summarizing the data characteristics of the Internet of Things,Proposed four requirements for data storage in the Internet of Things: high concurrency,real-time,reliability and openness.At the same time,it also reviewed the main technical directions of data storage.Secondly,on the basis of literature research,put forward the storage strategy of structured data and unstructured data.For structured data storage,considering that MySQL database is the current mainstream open source relational database,it has the advantages of high market share,free,high availability and good performance.This article is based on the MySQL database,from the perspective of database structure design,high availability,high performance,and scalability,according to the characteristics of the Internet of Things data,designed a corresponding storage strategy.For unstructured data storage such as pictures,traditional file storage systems have more deficiencies in terms of concurrency,performance,and scalability.Since most of the image data is a single small file with a small volume,the FastDFS distributed file system is used to store such unstructured data.For the convenience of reading files,Redis is combined to complete the index mapping relationship and read cache function of the stored files.Finally,the hybrid storage strategy proposed in this paper is implemented in actual engineering projects.And in the experimental environment,built a test environment and plan,and test the performance of system stress,the test results show that the established goals have been achieved.