Discussion on open source solution of GIS in accordance with company

1. Preface

This week, I did some preliminary exploration of open source software and related technologies and processes involved in the open source solution for GIS, as well as some of the more mature cases that other companies have done with open source solutions. Here I will some technical key points of the preliminary research to do a summary, while the other companies open source success stories to do an analysis, and put forward the personal think the most in line with the actual company's GIS open source solutions.

2. Technical key point pre-research 2.1 spatial Data Warehousing (Postgresql+postgis)

Use PostGIS to import SHP data into PostgreSQL:

2.2 Spatial Data Editing (UDig)

Import spatial data from PostgreSQL for display and editing operations.

Creation of 2.3SLD style files

You can directly use Udig for layer mapping and production of the corresponding SLD file, and can be exported:

2.4 Using GeoServer to publish spatial data in PostGIS

Add stores Select PostGIS in GeoServer:

2.5 Publishing layer groups with GeoServer

Combine multiple separately published layer services into one layer group, and in one request you can display the layers under all layer groups:

Layer groups that contain the cell grid and community:

2.6 Using Geowebcache transduction cache

In this paper, two kinds of geowebcache are studied: one is the AGS transduction strategy, namely the pre-cut graph, and the other is the AGS dynamic graph-cutting strategy, that is, the first time cut graph.

Transduction related configurations and results:

2.7 Research on spatial data management and spatial analysis using the St_geometry function in PostGIS

The St_geometry function in PostGIS is essentially the same as in SDE, but it contains some of its own extension functions. The following functions are broadly available:

In PostgreSQL, the basic spatial elements are tested and modified, and the area and length of spatial elements are obtained:

3. Research on other companies ' mature cases

A company's Ping an XX (Security project) for the study of the case object, through the exchange with their research and development staff, I generally draw the following information:

3.1 Adoption Framework

GeoServer (Map Server) +geowebcache (tile cache service) +jboss (middleware) +postgresql (spatial database) +oracle (Business database) +udig (spatial data Editing tool).

3.2 Summary of some of the analysis points of the project

A. GeoServer released nearly 60 layers in the project, with no apparent problem of instability.

B. Spatial data queries, edits, and spatial analysis functions involved in the project are implemented using the St_geometry function, which is very efficient.

C. Geowebcache Dynamic plot strategy used in map tile cache in project. Nearly 60 layers are used as basemaps, and the first request to plot takes approximately more than 20 seconds (which is a network drain).

D. Business data and spatial data are stored separately, business data is stored in Oracle, and spatial data is stored in PostgreSQL.

E. The deployment of the project, for their research and development prior to the data are processed, and then the data has been included in the GeoServer and Space library to the field engineering staff.

3.3 Reviews of the project 3.3.1 advantages

A. Features such as plot and spatial analysis are basic, and the efficiency and stability are good. Ping An Wuhan itself is a relatively large project, experienced a relatively good practice test.

B. All spatial analysis is implemented using St_geometry+sql, without the WFS services provided by the GeoServer itself, there are more customization requirements, and it is easy to troubleshoot if errors occur. If written reasonably, it can speed up the acquisition of data. At the same time, developers who are not GIS professionals are more likely to understand.

C. Separating the spatial data from the business data, which ensures that the business data used by the original project is not changed much and the stability of the system is ensured.

3.3.2 Disadvantages

A. The implementation of the project requires the participation of research and development personnel, data warehousing and release. Then also to cooperate with the scene to decorate the environment.

3.4 Places to be confirmed

A. When using Geowebcache to cut the chart, if the Basemap map is very complex, whether it can guarantee the transduction.

B. When the Geowebcache is used for the cut chart, if the basemap is a large image (GB above), whether the transduction can be guaranteed to be non-distorted, and the first time out of the plot efficiency.

C. When publishing more than 100 layer services with GeoServer, the stability of the geoserver itself.

4. In line with the company's Open source solutions

Through the above technical key point pre-research and other company's successful case analysis, combined with the company's current overall structure, I personally proposed a self-think more consistent with the company's Open source solution:

GeoServer (map Service) + Local tile service (ArcGIS etc) +tomcat (middleware) +postgresql (spatial database) +oracle (Business database) +udig (spatial data Editing tool).

4.1 System Detailed architecture commentary 4.1.1 Basemap caching service

Companies can still use ArcGIS for transduction (the company has a genuine), which has three benefits:

A. Project implementation personnel easy to operate.

B. Easy to map.

C. can guarantee the quality and efficiency of the cut chart.

After cutting the graph, we use our own existing offline tile strategy, bypassing the adoption and configuration of Geowebcache. Of course, if you do not want to use ArcGIS transduction, want to open source, we can also study the use of Maptiler this open source tool, cut graph efficiency and effect is better.

4.1.2 Component Dynamic Plot (WMS)

Use the company's existing GeoServer-based capabilities.

4.1.3 Vector Query (WFS)

Use the company's existing GeoServer-based capabilities.

4.1.4 Spatial Analysis Function (WFS)

For spatial analysis capabilities that have been developed through GeoServer, they can be based on adoption. For the parts that are not developed, it is recommended to use the St_geometry function.

4.1.5 Spatial Database

Use Postgresql+postgis. PostGIS Support Batch storage, also support Chinese data, stability and performance are very good, easy to operate. At the same time, there are many tutorials and community forums for development.

4.1.6 Data Editor

Use Udig for spatial data editing. Udig can import data from SHP data or PostGIS directly, or you can import SHP data into PostgreSQL. It also supports style editing of data and the generation of SLD files.

4.2 Implementation of the project

A. The project of cutting diagram can be completed by itself.

B. The storage link of spatial data can also be completed by PostGIS, and the operation is as simple as catalog operation. C. The distribution of spatial data and the correlation of the basic style of the parts, as well as the generation of business data, can be achieved by modifying existing gadgets.

D. Configuration and release of complex styles can be facilitated by research and development.

--Welcome reprint, but retain the copyright, please indicate the source in obvious place: http://www.cnblogs.com/naaoveGIS/

Discussion on open source solution of GIS in accordance with company