This was an experimental Python tile-server which is no longer maintained.

Documentation: http://aiovectortiler.readthedocs.org/

aiovectortiler is a Python asynchronous micro vector tile application that serves data from PostGIS. It was an asyncio-based approach that evolved out of the Utilery package created by Yohan Boniface.

aiovectortiler can serve vector tiles in json, geojson and Mapbox Vector Tile formats. It is compatible with with MVT client-side vector tile libraries such as Mapbox GL JS. MVT encoding is performed by Mapzen's mapbox-vector-tile module.

Layer configuration is by means of YAML recipe files. Here is an example recipe for OpenStreetMap data.

To use with docker, you'll need a directory containing:

• Your server_configs.yaml file;
• A layer_recipes folder.

Your directory should now look something like this:

aiovectortiler_docker_folder
 - server_configs.yaml
 - layer_recipes
    - recipe_1.yaml
    - recipe_2.yaml
    - etc.

Then run the docker image, binding the docker container port 80 to your preferred localhost port, and mapping the docker container's /configs volume to your aiovectortiler configs directory. For example:

docker run -i -t -p 80:80 -v /path/to/local/configs:/configs shongololo/aiovectortiler

If running the docker container from the command line, use an absolute path to your configs directory.

Your aiovectortiler server will now be running and available at localhost:80

For higher throughput, deploy a separate aiovectortiler instance for each available CPU core. See configs/docker-compose.yamlfile and the accompanying configs/nginx.conf file for an example load-balancing reverse-proxy configuration with caching.