Getting started

Main Page > Development > Development documentation > Getting Started

This wiki page describes getting started with Archivematica as a developer. For user and administrative manuals, please see http://www.archivematica.org.

Vital Stats

• Language: Python (primarily)
• License: AGPL
• VCS: git
• Major libraries: Django, gearman (Python API)
• Contribution guidelines
  • Coding style

Projects

Archivematica consists of several projects working together.

• Archivematica: Main repository containing the user-facing dashboard, task manager MCPServer and clients scripts for the MCPClient
• Storage Service: Responsible for moving files to Archivematica for processing, and from Archivematica into storage
• Format Policy Registry: Submodule shared between Archivematica and the Format Policy Registry (FPR) server that displays and updates FPR rules and commands

There are also several smaller repositories that support Archivematica in various ways. In general, you will not need these to develop on Archivematica.

• Development tools: Scripts to help with development. E.g. restarting services, workflow analysis
• FPR tools: All the tools, commands and rules used to populate the FPR database. Changes to the FPR should be submitted here.
• Archivematica Documentation: Documentation found at https://www.archivematica.org/en/docs/ for Archivematica
• Storage Service Documentation: Documentation found at https://www.archivematica.org/en/docs/ for the Storage Service
• Automation Tools: Scripts used to automate processing material through Archivematica
• Deployment: Ansible scripts for deploying and configuring Archivematica
• Deployment-Archivematica: Ansible playbook for Archivematica package install.
• Deployment-Archivematica-dev: Ansible playbook for Archivematica github install.
• Fixity checker: Commandline tool that assists in checking fixity for AIPs stored in Archivematica Storage Service instances.
• METS reader/writer: Library to create and parse METS files.
• agentarchives: Clients to retrieve, add, and modify records from archival management systems.
• Sample data: Data to test and show off Archivematica's processing
• History: Contains the pre-git history of Archivematica. Useful for checking the origins of code.

Installation

The recommended way to install Archivematica for development is with Ansible and Vagrant.

Ansible & Vagrant

The following instructions detail how to install and run Archivematica from source on a virtual machine.

1. Install VirtualBox, Vagrant, and Ansible with the following commands:
   • sudo apt-get install virtualbox vagrant (this is the command for Ubuntu; if you use Mac or a different Linux distribution, it may be slightly different).
     • Note: Vagrant must be at least 1.5 (it can also be downloaded from vagrantup.com). Check your version with vagrant --version.
   • sudo pip install -U ansible
2. Checkout the deployment repo:
   • git clone https://github.com/artefactual/deploy-pub.git
3. Download the Ansible roles:
   • cd deploy-pub/playbooks/archivematica
   • ansible-galaxy install -f -p roles/ -r requirements.yml
4. (Optional) Change the branch by opening the file vars-singlenode.yml and modifying the following:
   • archivematica_src_am_version: "branch-name"
   • archivematica_src_ss_version: "branch-name"
5. Create the virtual machine and provision it:
   • vagrant up (it takes a while!)
6. You can now log in to your virtual machine:
   • vagrant ssh
7. You can now access the following services in a web browser:
   • Archivematica - http://192.168.168.192
   • Archivematica Storage Service: http://192.168.168.192:8000 (user: test, pass: test)

You may also wish to do the following.

1. Provisioning (via ansible) can be re-run with vagrant to update the code on the server (for example, if new features are added):
   • vagrant provision
2. To re-deploy a new branch to the same VM, update the branch variables in vars-singlenode.yml:
   • archivematica_src_am_version: "branch-name"
   • archivematica_src_ss_version: "branch-name"
     • This will probably require resetting the Archivematica installation as well. This can be done by adding variables to vars-singlenode.yml
   • archivematica_src_reset_am_all: "true" This will reset the Archivematica database, clear ElasticSearch and clear shared directories
   • archivematica_src_reset_ss_db: "true" This will reset the Storage Service database
   • For more variables to control deployment, see the README
3. See also the FAQ below

Local install (Legacy)

Archivematica can also be installed locally on a development machine using dev-helper. See development environment setup instructions This is considered legacy, and we recommend using Vagrant and Ansible.

Alternative Vagrant projects

Community-provided alternatives have also been developed.

• https://github.com/emltech/eml-archivematica-vagrant
• https://github.com/statsbiblioteket/archivematica-vagrant

Tests

Archivematica and the related projects have a small but growing test suite. We use py.test to run our tests, which should be listed as a requirement in the development/local requirements file.

To run the tests, go to the repository root and run py.test

See below for project-specific setup or changes to running the tests.

Archivematica

Before running Archivematica tests, set the following environment variable. Archivematica does not currently have a virtualenv that needs to be activated.

#!/usr/bin/fish
set -xg PYTHONPATH /usr/share/archivematica/dashboard/:/usr/lib/archivematica/archivematicaCommon/

#!/usr/bin/bash
export PYTHONPATH=$PYTHONPATH:/usr/share/archivematica/dashboard/:/usr/lib/archivematica/archivematicaCommon/

Storage Service

Before running Storage Service tests, activate the virtualenv and set the following environment variables. The tests should be run from the storage_service directory. This is the same directory that contains manage.py. You may need to install requirements/test.txt to install testing dependencies.

#!/usr/bin/fish
set -xg PYTHONPATH (pwd)/storage_service  # This directory contains manage.py
set -xg DJANGO_SETTINGS_MODULE storage_service.settings.test
set -xg DJANGO_SECRET_KEY 'ADDKEY'

#!/usr/bin/bash
export PYTHONPATH=$(pwd)/storage_service  # This directory contains manage.py
export DJANGO_SETTINGS_MODULE=storage_service.settings.test
export DJANGO_SECRET_KEY='ADDKEY'

FAQ

How do I restart everything in Archivematica?

A default install using Ansible also installs the devtools. Run am restart-services to restart all services related to Archivematica and the storage service.

How do I restart just the dashboard?

sudo /etc/init.d/apache2 start or sudo service apache2 restart

How do I restart the storage service?

sudo service uwsgi restart

How do I restart nginx?

sudo service uwsgi restart sudo service nginx restart

How do I update or reset an ansible install?

To update an install, re-run vagrant provision. If you only want to run part of the ansible tasks, you can use ansible's tags, for example: env ANSIBLE_ARGS="--tags=amsrc-ss-code" vagrant provision More tags are documented in the ansible repo

To reset an install (delete all existing data like a fresh install) you can use ansible's role variables. For example: env ANSIBLE_ARGS="--extra-vars=archivematica_src_reset_ss_db=true" vagrant provision will reset the storage service database. More role variables are documented in the ansible repo

The other way to control a deployment is to modify the Vagrantfile and vars-singlenote.yml files directly. Tags can be provided in the Vagrantfile. For example:

# ... more above
  # Ansible provisioning
  config.vm.provision :ansible do |ansible|
    ansible.playbook = "./singlenode.yml"
    ansible.host_key_checking = false
    # ansible.verbose = "v"
    ansible.extra_vars = {
      "archivematica_src_dir" => "/srv",
      "archivematica_src_environment_type" => "development",
    }
    ansible.raw_arguments = ENV['ANSIBLE_ARGS']
    ansible.tags = ['amsrc-pipeline']
end
# ... more below

Role variables can be modified in vars-singlenote.yml. Default values are found in the Archivematica role (in roles/archivematica-src/defaults/main.yml). For example:

---

# archivematica-src role

# What to install
archivematica_src_install_devtools: "yes"
archivematica_src_install_automationtools: "yes"
# archivematica_src_install_appraisaltab: "yes"

# SS django environment variables
archivematica_src_ss_env_django_setings_module: "storage_service.settings.local"

# Branches,
archivematica_src_am_version: "qa/1.x"
archivematica_src_ss_version: "qa/0.x"
# archivematica_src_devtools_version: "master"
# archivematica_src_automationtools_version: "master"

# Reset
# archivematica_src_reset_mcpdb: "true"
# archivematica_src_reset_shareddir: "true"
# archivematica_src_reset_es: "true"
archivematica_src_reset_am_all: "true"
archivematica_src_reset_ss_db: "true"

How do I activate the Storage Service virtualenv?

source /usr/share/python/archivematica-storage-service/bin/activate

The location of the virtualenv is configured as part of the ansible install, which by default is /usr/share/python/archivematica-storage-service Sourcing the activate script should modify the prompt to display (archivematica-storage-service)vagrant@am-local. Note that if you're running fish, the activate script will not work; you may want to look in to virtualfish.

You will also want to set environment variables as described in the storage service testing section.

Documentation

Developer facing documentation can be found in the Development documentation category. Notable pages include:

• MCPServer
• MCPClient
• Storage Service
• Storage Service API
• Archivematica API