You can freely host your own instance of Open-Elevation. There are two main options: Docker or native. We recommend using docker to ensure that your environment matches the development environment
First things first, clone the repository and cd onto its directory
git clone http://github.com/Jorl17/open-elevation
cd open-elevation
An image of Open-Elevation is available at DockerHub. You can use this image as the basis for your Open-Elevation installation.
The Docker image roots itself at /code/ and expects that all GeoTIFF datafiles be located at /code/data/, which you should mount using a volume. You can also change this directory by changing the configuration file
If you want HTTPS, an SSL certificate and key can be mounted at /code/certs/. See below.
Open-Elevation doesn't come with any data of its own, but it offers a set of scripts to get the whole SRTM 250m dataset.
If you wish to host the whole world, just run
mkdir data # Create the target folder for the dataset
docker run -t -i -v $(pwd)/data:/code/data openelevation/open-elevation /code/create-dataset.sh
Or if you're on Windows Powershell:
mkdir data # Create the target folder for the dataset
docker run -t -i -v ${pwd}/data:/code/data openelevation/open-elevation /code/create-dataset.sh
The above command should have downloaded the entire SRTM dataset and split it into multiple smaller files in the data directory. Be aware that this directory may be over 20 GB in size after the process is completed!
If you don't want to use the whole world, or wish to use another data source, you can provide your own dataset in GeoTIFF format, compatible with the SRTM dataset. Simply drop the files for the regions you desire in the data directory. You are advised to split these files in smaller chunks to make Open-Elevation less memory-hungry (the largest file has to fit in memory). The create-tiles.sh is capable of doing this, and you can see it working in create-dataset.sh. Since you are using docker, you should always run the commands within the container. For example:
docker run -t -i -v $(pwd)/data:/code/data openelevation/open-elevation /code/create-tiles.sh /code/data/SRTM_NE_250m.tif 10 10
The above example command splits SRTM_NE_250m.tif into 10 by 10 files inside the /code/data directory, which is mapped to $(pwd)/data.
Now that you've got your data, you're ready to run Open-Elevation! Simply run
docker run -t -i -v $(pwd)/data:/code/data -p 80:8080 openelevation/open-elevation
This command:
- Maps
$(pwd)/data(your data directory) to/code/datawithin the container - Exposes port 80 to forward to the container's port 8080
- Runs the default command, which is the server at port 8080
You should now be able to go to http://localhost for all your open-route needs.
Before starting, the server checks for an SSL certificate and key files at the certs/subdirectory (this can be changed in the config file).
If found, the server boots using SSL/HTTPS (and only that). File names should be /code/certs/cert.crt and /code/certs/cert.key.
The following command will mount also /code/certs and run the server on port 443:
docker run -t -i -v $(pwd)/data:/code/data -v $(pwd)/certs:/code/certs -p 443:8080 openelevation/open-elevation
In order for Open-Elevation to work, you need GDAL and libspatialindex. For the full process to work you also need a version of unrar.
The setup for gdal depends on the distro and may even change among distro versions, thus being outside the scope of this documentation. Please follow the documentation found in GDAL's homepage.
The following are instructions for Ubuntu/Debian compatible distros, and similar ones might apply to your particular setup. Again, make sure you also install GDAL. Please consider using a virtualenv instead of using the default python installation for the following commands.
apt-get update -y
apt-get install -y libspatialindex-dev unrar-free bc
pip install -r requirements.txt
If all goes well, you now have the required dependencies to run Open-Elevation.
Open-Elevation doesn't come with any data of its own, but it offers a set of scripts to get the whole SRTM 250m dataset.
If you wish to host the whole world, just run
./create-dataset.sh
Assuming you have wget and unrarinstalled, the above command should have downloaded the entire SRTM dataset and split it into multiple smaller files in a data directory. Be aware that this directory may be over 20 GB in size after the process is completed!
If you don't want to use the whole world, you can provide your own dataset in GeoTIFF format, compatible with the SRTM dataset. Simply drop the files for the regions you desire in the data directory. You are advised to split these files in smaller chunks so as to make Open-Elevation less memory-hungry (the largest file has to fit in memory). The create-tiles.sh is capable of doing this, and you can see it working in create-dataset.sh.
After this process, you should now have your dataset in the tiles subdirectory. You're good to go!
Note that the server expects all GeoTIFF files to be in a data folder (not tiles). You can also change this directory by changing the configuration file
Now that you've got your data, you're ready to run Open-Elevation! Simply run
python server.py
And you should be good to go!
Before starting, the server checks for an SSL certificate and key files at the certs/subdirectory (this can be changed in the config file).
If found, the server boots using SSL/HTTPS (and only that). File names should be /code/certs/cert.crt and /code/certs/cert.key.
Have you found any problems? Open an issue or submit your own pull request!