NOTE: For official information, check out the QGIS Server documentation. An example of a Flask web service with QGIS Server can be found in this repo: https://github.com/voirinprof/qgis_docker

Introduction

QGIS Server is the server version of QGIS that allows you to publish WMS, WFS, and WCS map services directly from .qgs projects without heavy development.
By combining QGIS Server with Python, you can automate, customize, and create dynamic web GIS services.

Why use QGIS Server?

  • Simplicity: Publish a QGIS project as a web service as is
  • OGC Standards compliance: WMS 1.3, WFS 2.0, etc.
  • Advanced customization via server plugins in Python
  • Same rendering capabilities as the QGIS Desktop
  • Direct integration with PostGIS, GeoTIFF, and more

Install QGIS Server

On Ubuntu / Debian:

sudo apt-get install qgis-server

Necessary services:

  • Web server (Apache, Nginx)
  • FastCGI or uwsgi to interface QGIS Server with the web server

Verify installation:

Test the local URL (e.g., with Apache):

http://localhost/qgisserver?SERVICE=WMS&REQUEST=GetCapabilities

If everything is correctly set up, you should get an XML response describing your server capabilities.

Deploy a QGIS project as a WMS service

  1. Prepare your .qgs project:
    • Set the CRS (coordinate reference system)
    • Configure project properties (layer names, min/max scales)
  2. Place the project file:
    • Save your .qgs in a folder accessible by the server (e.g., /var/www/qgis/myproject.qgs).
  3. Call your project via the URL:
http://localhost/qgisserver?MAP=/var/www/qgis/myproject.qgs&SERVICE=WMS&REQUEST=GetCapabilities

Use Python with QGIS Server

Python allows you to create custom server extensions or interact with QGIS Server. Two main approaches exist:

1. Create plugins for QGIS Server

  • Plugins are structured the same way as QGIS Desktop plugins but with a specific entry point: server instead of gui.
  • Allows:
    • Dynamic modification of layers
    • Adding new operations to the service
    • Custom data filtering

Example of server plugin architecture:

my_plugin/
├── __init__.py
├── server/
│   ├── __init__.py
│   ├── myplugin.py
├── metadata.txt

In server/myplugin.py:

from qgis.server import QgsServerFilter

class MyPluginServer(QgsServerFilter):
    def requestReady(self):
        print("Request received:", self.serverInterface().requestHandler().parameterMap())

2. Use PyQGIS to pre-process your projects

With PyQGIS (the Python API for QGIS), you can dynamically generate .qgs files, prepare styles, PostGIS layers, before publishing on QGIS Server.

Simple PyQGIS script example:

from qgis.core import QgsProject

# Load an existing project
project = QgsProject.instance()
project.read('/path/to/my_project.qgs')

# Dynamically change the title
project.title = "New server title"

# Save
project.write()

Typical Use Cases

  • Automatic generation of thematic GIS services
  • Applying data access rules via server plugins
  • Customizing GetFeatureInfo (enriched JSON response)
  • Conditional rendering based on users/sessions

Useful Resources

Best Practices

  • Optimize your .qgs projects (light styles, logical groups)
  • Use PostGIS for large spatial databases
  • Cache your services (with MapProxy or QGIS Cache) to speed up responses
  • Monitor server logs for debugging Python plugins
  • Automate project updates with PyQGIS