Skip to content

Latest commit

 

History

History
204 lines (146 loc) · 5.22 KB

README.md

File metadata and controls

204 lines (146 loc) · 5.22 KB

Requirements

docker, docker compose, python3 with pip

Install Docker

Follow instructions on https://docs.docker.com/engine/install/

Install Docker Compose

Follow instructions on https://docs.docker.com/compose/install/compose-plugin/

Or download and install specific version

mkdir -p ~/.docker/cli-plugins/
curl -SL https://github.com/docker/compose/releases/download/v2.5.1/docker-compose-linux-x86_64 -o ~/.docker/cli-plugins/docker-compose
chmod +x ~/.docker/cli-plugins/docker-compose

Install Python3 + pip

For Ubuntu distribution

apt-get update
apt-get install -y --no-install-recommends python3 python3-setuptools python3-pip

Gisquick CLI

Set of commands to simplify and speed up setup of new deployment(s).

Installation:

python3 -m pip install  https://github.com/gisquick/gisquick-cli/archive/master.zip

Depending on your environment, you may need to run this command with root privileges.

For development, install from source code

python3 -m pip install -e <path-to-gisquick-cli>

Check installed version

python3 -m pip show gisquick-cli | grep Version

Setup deployment

Generate configuration for Gisquick deployment

gisquick-cli create <name>

This will create a deployment directory with generated main docker-compose.yml file, environment variables files and configuration for core and selected optional services. gisquick-cli create command allow to adjust configuration with selected predefined options. For full control, you will have to modify configuration manually.

To see all Gisquick server options/environment variables, execute command:

docker run --rm gisquick/server gisquick serve --help

First start and initialization of the database

cd <name>
docker compose up -d
docker compose exec app gisquick migrate up

Note: gisquick migrate command is tiny wrapper for https://github.com/golang-migrate/migrate tool.

Progressive Web App version

For map application with PWA support, change Docker image of the web-map service in docker-compose.yml to gisquick/web-map:pwa

Create superuser/user account

docker compose exec app gisquick addsuperuser

or

docker compose exec app gisquick adduser

Superuser can manage other accounts also from admin web interface at /admin/ url.

To see all Gisquick commands, run

docker compose exec app gisquick --help

Logs and monitoring

Logs and monitoring data are accessible in Grafana web interface, available at /admin/grafana endpoint. Initial user account is admin/admin (username/password)

Setup of data sources

In Configuration > Data sources settings, add:

Deployment for server development

Build development version of server's Docker image (see README of its source code)

Use --dev-server flag for generating deployment files

gisquick-cli create --dev-server gisquick

After you made changes in source files, restart service to recompile:

docker compose restart app

Update

In most cases you can update Gisquick by fetching newer Docker images and recreating its containers

To update Gisquick services (server application and web applications), run

docker compose pull app web-map web-settings
docker compose up -d app web-map web-settings

Useful commands

Generate new secret key

tr -dc 'a-z0-9!@#%^&*(-_=+)' < /dev/urandom | head -c50

Reload Caddy server

docker compose kill -s HUP caddy

Gisquick running behind gateway server

It's easier when Gisquick is directly running on port 443 and managing it's SSL certificate. However if you don't have that option, you can deploy Gisquick on different port and configure reverse proxy on your gateway server. For most endpoints, simple reverse proxy is enough. But if your gateway server doesn't handle websocket connections automatically, you will need to adjust configuration. Also it's better to disable request buffering on project upload endpoint, otherwise upload progressbar may not work correctly.

Nginx

Example of Nginx configuration, assuming Gisquick is accessible on IP address 10.1.1.1 and port 80

location / {
    proxy_pass http://10.1.1.1;
}

location /ws/ {
    proxy_set_header Host $http_host;
    proxy_set_header X-Real-IP $remote_addr;
    proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
    proxy_set_header X-NginX-Proxy true;
    proxy_set_header X-Forwarded-Proto $scheme;

    proxy_pass http://10.1.1.1;
    proxy_redirect off;
    proxy_read_timeout 86400;

    # Websocket support
    proxy_http_version 1.1;
    proxy_set_header Upgrade $http_upgrade;
    proxy_set_header Connection "upgrade";

    # prevents 502 bad gateway error
    proxy_buffers 8 32k;
    proxy_buffer_size 64k;

    reset_timedout_connection on;
}

location /api/project/upload {
    client_max_body_size 500M;
    proxy_set_header Host $http_host;
    proxy_set_header X-Real-IP $remote_addr;
    proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
    proxy_set_header X-Forwarded-Proto $scheme;

    proxy_intercept_errors on;
    proxy_request_buffering off;
    proxy_buffering off;
    proxy_http_version 1.1;
    chunked_transfer_encoding on;
    proxy_pass http://10.1.1.1;
}