Vulcan

Vulcan is a framework for enterprise collaboration platforms originally created for the DARPA Adaptive Vehicle Make (AVM) Program. There, it provded a foundation for VehicleFORGE, a platform for cyber-physical system design used to host AVM's Fast Adaptive Next-Generation (FANG) vehicle design challenges.

VehicleFORGE originated as an early fork of SourceForge's Allura platform, which is today Apache Allura. Vulcan was motivated by providing a better organization as a framework and by adding new forms of extensibility for adapting platform features and services to an enterprise.

Vulcan is comprised of a core part and a set of optional annexes. This repository provides the core part, VulcanForge.

Micro-Service Dependencies

Vulcan employs MongoDB for document persistence, Apache Solr for indexing, and Redis as a key-object store. In the current release, the following versions of these dependencies are known to be compatible:

• MongoDB 3.4.4
• Apache Solr 6.5.0
• Redis 3.2.8

Vulcan applications assume these micro-services are deployed in a manner reflecting the application's requirements. Deployments of Apache Solr are supported by assets provided in the solr_config directory.

Cloud Object Storage

Vulcan employs cloud object storage for artifact persistence, requiring an object service supporting an AWS S3-compatible API that is accessed via the Python boto package. This requirement has been satisfied in private cloud deployments using OpenStack Swift, for example.

Installation and Creating Applications

Installed alone, VulcanForge supports creating new framework applications from templates. When doing so, we recommend using a virtual environment. Below is a basic procedure.

1. Clone this repository.

2. Create and initialize a virtualenv.

   $ virtualenv vf_venv
   $ . vf_venv/bin/activate
   

3. Change directory into this repository's root directory.

4. Install VulcanForge.

   $ pip install -r requirements.txt
   $ python setup.py develop
   

5. Create an application

   $ paster create-vulcanapp -h
   
   Usage: paster create-vulcanapp [options] <package> <target directory (defaults to current)>
   Creates a new vulcan application
   
   Options:
     -h, --help            show this help message and exit
     -v, --verbose
     -n NAME, --name=NAME  Project name (defaults to name of package)
     -r, --repos           Application includes support for repositories (should
                           have vulcanrepo installed)
     -t TEMPLATE, --template=TEMPLATE
                           Path to custom application template (file path or url
   

Vulcan Applications

Vulcan applications created using create-vulcanapp will have a canonical organization that can be used to selectively extend and refine potions of the framework. Also included in the created application are two prototype configuration files: development.ini and production.ini. The first provides a commented guide to configuring the application. The second provides a set of configuration items specifically for production deployments as differences from those in development.ini.

Applications can be served using WSGI-compatible web servers; for example, Apache with mod_wsgi. For development, applications can be served using paster serve.

Release Notes

Version 2.0.3

Update chardet dependency.

Version 2.0.2

Minor Python packaging changes.

Version 2.0.1

This is a minor feature release.

• Disable gravatar support by configuration setting

Version 2.0.0

This release is compatible with the Ubuntu Xenial LTS. It is highly recommended for all application deployments, as it includes critical security fixes to previous releases.

• Support added for installing VulcanForge using pip via a requirements file.
• Markdown sanitization is now performed using the Python bleach module, which resolves vulnerabilities detected by penetration testing.
• A new core tool type catalog is added a foundation for tools that manage versioned bundles of metadata and artifacts.
• Vulcan projects now better support sub-typing and polymorphic querying.
  Project.query.get and Project.query.find continue to operate on the base type for compatibility. New query_get and query_find methods are subtype-aware.
• Solr configuration assets are now included directly in VulcanForge and are compatible with Solr 6.5.0. When upgrading Vulcan applications to Solr 6.5.0 from previous releases, Solr indexes must be rebuilt using the reindex_globals and reindex commands.
• For compatibility with MongoDB and PyMongo 3+, VulcanForge now installs the Ming ODM from a branch of its development repository pymongo-30 rather than the Ming project's PyPI packages.