This docker image runs a container preinstalled with a certain release of
GrimoireLab, including all libraries and programs. By default, it runs
the GrimoireLab orchestrator - sirmordred - for producing a dashboard given
a mordred configuration file.
If you prefer, you can run GrimoireLab using docker-compose it's the simplest method to get started with. More details in the docker-compose folder.
- Git
- Docker and Docker Compose
- Recommended, 2CPUs, 8GB memory RAM, and 2GB SWAP (MacOS users can manage it with Docker client for Mac)
To produce the image, run the following command:
$ docker build -f Dockerfile -t grimoirelab/grimoirelab .
You can build a specific version of grimoirelab updating GRIMOIRELAB_RELEASE
environment variable inside the Dockerfile.
The image will run the GrimoireLab orchestrator - sirmordred- by default. You
need to define a setup.cfg and a projects.json. This repository comes with some
configuration files in default-grimoirelab-settings
directory. These files are the minimum set up to analyze git activity in this
repository. If you want to analyze a different project with a different set of
repositories and data sources, you need to modify projects.json
and setup.cfg files. See
below more information about these files format.
You also need a SortingHat server running at http://localhost:8000/identities. See Sortinghat for more information
Default configuration files assumes Elasticsearch is available in port 9200, and MariaDB or MySQL in their standard port (if they are not present, the process will fail).
Run the following command from the root of this repository to start using SirMordred:
docker run --net=host \
-v $(pwd)/default-grimoirelab-settings/projects.json:/home/grimoire/conf/projects.json \
-v $(pwd)/default-grimoirelab-settings/setup-docker.cfg:/home/grimoire/conf/setup.cfg \
-t grimoirelab/grimoirelab
Remember you can run GrimoireLab and this image using docker-compose. It's the simplest method to get started with. More details in the docker-compose folder.
You can also run any GrimoireLab tool using this image, for example, you can collect Git commits using Perceval:
$ docker run grimoirelab/grimoirelab perceval git https://github.com/chaoss/grimoirelab-perceval.git
The different files we can modify are:
- setup.cfg: SirMordred's configuration file (GrimoireLab orchestrator configuration)
- projects.json: JSON file with projects and their repositories
The configuration file is composed of several sections which allow to define the general settings such as which phases to activate (e.g., collection, enrichment) and where to store the logs, as well as the location and credentials for SortingHat and the ElasticSearch instances where the raw and enriched data is stored. Furthermore, it also includes backend sections to set up the parameters used by Perceval to access the software development tools (e.g., GitHub tokens, Gerrit username) and fetch their data.
The complete list of options and their description is available here.
We provide a complete example that works out of the box with CHAOSS data and some visualizations based on the collected data. By default, only Git collection is enabled. Take a look and configure many other options.
The projects.json aims at describing the repositories grouped by a project that will be shown on the dashboards.
The project file enables the users to list the instances of the software development tools to analyze, such as local and remote Git repositories, the URLs of GitHub and GitLab issue trackers, and the name of Slack channels. Furthermore, it also allows the users to organize these instances into nested groups, which structure is reflected in the visualization artifacts (i.e., documents and dashboards).
There are some filters, labels and a special section, take a look here for more details..