A Distributed Cluster Management Method for Map Cache Slices Based on Quadtree Hierarchical Grid

Abstract

The invention discloses a distributed cluster management method for map cache slices based on a quadtree layered grid, which uses a quadtree layered grid algorithm to realize a distributed cluster of map cache slices, and distributes map cache slices based on spatial geographic regions. Cluster management ensures that the spatial and geographical areas that users really care about get higher performance and improves the rationality of cluster resource utilization. The downtime of a cluster grouping node will only affect the spatial geographic area corresponding to the cluster grouping, and will not affect other spatial geographic areas, thereby improving the robustness of the cluster.

Description

technical field [0001] The invention belongs to the technical field of map hierarchical cache slicing, and in particular relates to the design of a distributed cluster management method for map cache slicing based on a quadtree layered grid. Background technique [0002] Map browsing is the basic function of all GIS systems. According to the realization principle, it can be divided into real-time rendering map and hierarchical cache map. Compared with real-time rendering map, hierarchical cache map supports fast response and high concurrency of map browsing. It is a rendering display with big data The preferred method when measuring GIS maps. [0003] Map hierarchical cache slice is a function that renders GIS maps into small pictures of the same size and specifications according to different preset scale levels, and stores them in the server. When the client browses the map, it directly queries the corresponding small pictures on the server side to realize the map browsing ...

AI Technical Summary

Problems solved by technology

[0005] (1) The method of local files on the server has the following disadvantages: the access speed of small pictures is slow; a large number of disk fragments are generated, wasting disk space; too many small picture files are troublesome to backup and manage; there is a risk of single node failure

[0006] (2) The method of database storage has the following disadvantages: there is a single node failure; distributed cluster management is not realized, and the efficiency is low

[0007] (3) In the traditional database cluster storage method, clusters are formed by means of hash functions or data table horizontal segmentation, without considering the spatial location distribution of small pictures, resulting in random distribution of small pictures in the same spatial geographical area on the cluster nodes, This leads to the problem of fragmentation of small images, and finally brings the following two disadvantages: firstly, small images in key spatial geographic areas are treated equally with non-focused spatial geographic areas, wasting server resources; secondly, a node in the cluster is down or Unstable, it will cause unavailability or instability of the entire cluster, and the cluster robustness is too poor

Images