kobotoolbox/kobo-install

The purpose of the script is to install KoboToolbox in minutes without messing with configuration files. It prompts the user to answer some questions to create configuration files automatically and to start docker containers based on kobo-docker.

⚠️ You must observe the following when upgrading:

…from any release older than 2.022.44 (November 2022)

If you have already installed KoboToolbox between March 2019 and November 2022, you must complete a manual upgrade process before trying to upgrade. If you do not, kobo-install will not be able to start.

…from any release older than 2.020.18 (May 2020)

Prior to release 2.020.18, KPI and KoBoCAT both shared a common Postgres database. They now each have their own. If you are upgrading an existing single-database installation, you must follow these instructions to migrate the KPI tables to a new database and adjust your configuration appropriately.

If you do not want to upgrade at this time, please use the shared-database-obsolete branch instead.

…installations made prior to March 2019

If you have already installed KoboToolbox with kobo-docker prior March 2019, you must complete a manual upgrade process before using this repository. If you do not, kobo-install will not be able to start.

Versions

Branch master is the recommended branch to use kobo-install on your production environment. Just run git checkout master before your first run.

Branch beta is a pre-release of the next version. It contains new features and bug fixes.

Other branches are for development purposes.

Usage

$kobo-install> python3 run.py

First time the command is executed, setup will be launched.
Subsequent executions will launch docker containers directly.

Rebuild configuration:
$kobo-install> python3 run.py --setup

Get info:
$kobo-install> python3 run.py --info

Get docker logs:
$kobo-install> python3 run.py --logs

Update KoBoToolbox:
$kobo-install> python3 run.py --update [branch or tag]

By default, fetch the latest version of master branch

Stop KoBoToolbox:
$kobo-install> python3 run.py --stop

Get help:
$kobo-install> python3 run.py --help

Get version:
$kobo-install> python3 run.py --version

Build kpi and kobocat (dev mode):
$kobo-install> python3 run.py --build

Run docker commands on front-end containers:
$kobo-install> python3 run.py --compose-frontend [docker-compose arguments]

Run docker commands on back-end containers:
$kobo-install> python3 run.py --compose-backend [docker-compose arguments]

Start maintenance mode:
$kobo-install> python3 run.py --maintenance

Stop maintenance mode:
$kobo-install> python3 run.py --stop-maintenance

Build the configuration

User can choose between 2 types of installations:

• Workstation: KoBoToolbox doesn't need to be accessible from anywhere except the computer where it's installed. No DNS needed
• Server: KoBoToolbox needs to be accessible from the local network or from the Internet. DNS are needed

Options

Advanced Options

¹⁾ HTTPS certificates must be installed on a Reverse Proxy. kobo-install can install one and use Let's Encrypt to generate certificates thanks to nginx-certbot project

²⁾ If AWS credentials are provided, backups are sent to configured bucket

³⁾ Custom settings are provided by PostgreSQL Configuration Tool API

⁴⁾ Redis password is optional but strongly recommended

⁵⁾ If AWS storage is selected, credentials must be provided if backups are activated

⁶⁾ WAL-E backups for PostgreSQL are only available when using AWS storage

ℹ Intercom App ID must now be configured through "Per user settings" in the Django admin interface of KPI.

Requirements

• Linux ⁵ / macOS ⁶

• Python 3.7+

• Docker & Docker Compose

• Available TCP Ports: ⁷

  1. 80 NGINX
  2. 443 NGINX (if you use kobo-install with LetsEncrypt proxy)
  3. Additional ports when expose ports advanced option has been selected
     1. 5432 PostgreSQL
     2. 6379-6380 redis
     3. 27017 MongoDB

  WARNING:

  • If you use a firewall, be sure to open traffic publicly on NGINX port, otherwise kobo-install cannot work
  • By default, additional ports are not exposed except when using multi servers configuration. If you choose to expose them, be sure to not expose them publicly (e.g. use a firewall and allow traffic between front-end and back-end containers only. NGINX port still has to stay publicly opened though).

⁵⁾ It has been tested with Ubuntu 14.04, 16.04 and 18.04, CentOS 8

⁶⁾ Docker on macOS is slow. First boot usually takes a while to be ready. You may have to answer Yes once or twice to question Wait for another 600 seconds? when prompted

⁷⁾ These are defaults but can be customized with advanced options

Tests

Tests can be run with tox.
Be sure it is installed before running the tests.

$kobo-install> sudo apt install python3-pip
$kobo-install> pip3 install tox
$kobo-install> tox

or

$kobo-install> sudo apt install tox
$kobo-install> tox