Spice Is Nice

MediaWiki [1] is the open source, highly collaborative knowledge management tool that runs the world's sixth largest website: Wikipedia. Wiki solutions such as MediaWiki make it easy for large communities to collaborate, offering common access to an information store with built-in editing tools and version control.

MediaWiki is designed to serve a vast community on a global scale, but what if you need to serve a less diverse audience of users in a local or organizational setting? Wikis are excellent tools for business because they support central storage of shared information and eliminate the knowledge bottlenecks that occur when knowledge is shared haphazardly through email and personal connections.

Several open source projects support wiki tools for business environments, including XWiki, DokuWiki, and BlueSpice (for a comparison of these projects, see [2]). BlueSpice, from Hallo Welt!, is a German MediaWiki distribution that focuses on usability in enterprise environments. BlueSpice 4.1 [3], released in early 2022, offers a large set of features. The best way to determine whether a tool such as BlueSpice is right for your company is to set it up and explore it. In this article, we'll show how to set up BlueSpice in a typical enterprise setting.

More on BlueSpice

BlueSpice comes in a free version (BlueSpice free) and a commercial version (BlueSpice pro). BlueSpice free, which is the primary focus for this article, is an extension built on top of MediaWiki. Whereas MediaWiki is very simple and austere (Figure 1), BlueSpice free offers several additional options, along with a more sophisticated look and feel (Figure 2).

Figure 1: MediaWiki has a clear but unadorned user interface.

Figure 2: Bluespice free offers several options that aren't part of Mediawiki, such as the privacy center and an a ElasticSearch engine.

Deploying BlueSpice

You have two options for getting BlueSpice up and running. The "classical" way is to set up BlueSpice on a working LAMP stack. To get BlueSpice up and running, you simply extract BlueSpice to the DocumentRoot directory of your web server, start its built-in configuration agent (in a browser), and enter relevant data (such as the database to use along with the required credentials).

However, if you don't happen to have a working LAMP stack on hand, BlueSpice also offers a Docker image of BlueSpice free 4.1 that contains all the necessary components. As this Docker image comes straight from Hallo Welt! and the image's sources are also publicly available, it's easy to verify what the image does internally, doing away with the latent danger of running a "black box" on your hardware. Preparations for getting BlueSpice up and running in Docker are a bit different from what you usually do in non-Docker environments, but overall, the Docker image should be less work than configuring a working version of BlueSpice on an existing LAMP stack.

Hardware

Although BlueSpice can cause quite a bit of CPU and RAM load, it usually isn't necessary to supply the software with huge amounts of hardware. Dedicating a physical machine will almost always be overkill. Only a huge enterprise with thousands of concurrent users accessing the tool will see BlueSpice consume dozens of CPU cores or hundreds of gigabytes of RAM. The ideal (and normal) deployment scenario is inside of a virtual machine (VM). You can use standard tools such as VMware, Proxmox, or KVM-based virtualization as a VM in one of the hyperscaling public clouds (e.g., Amazon Web Services, Azure, etc.).

However, you should pick a storage device for your BlueSpice VM that is replicated on the infrastructure level. Otherwise, once BlueSpice is up and running, there would be no way to ensure proper high availability (HA) for the service.

For an average-sized BlueSpice setup, a small machine with eight (virtual) CPU cores and 16GB of RAM should be absolutely sufficient. As the amount of stored articles in the wiki grows, RAM needs might increase. Inside the container, BlueSpice uses ElasticSearch for searching and indexing; the more data, the more resources you will need. If you are running BlueSpice inside of a VM, and the VM runs out of resources, just add more virtual resources to the system.