kingarthur 0.2.5

Distributed job queue platform for scheduling Perceval jobs

Arthur

King Arthur commands his loyal knight Perceval on the quest to fetch data from software repositories.

Arthur is a distributed job queue platform that schedules and executes Perceval. The platform is composed by two components: arthurd, the server that schedules the jobs and one or more instances of arthurw, the work horses that will run each Perceval job.

The repositories whose data will be fetched are added to the platform using a REST API. Then, the server transforms these repositories into Perceval jobs and schedules them between its job queues.

Workers are waiting for new jobs checking these queues. Workers only execute a job at a time. When a new job arrives, an idle worker will take and run it. Once a job is finished, if the result is successful, the server will re-schedule it to retrieve new data.

By default, items fetched by each job will be published using a Redis queue. Additionally, they can be written to an Elastic Search index.

Requirements

• Python >= 3.7
• Redis (>= 2.3 and < 3.0) database will also be needed to schedule and execute Perceval jobs.

You will also need some other libraries for running the tool, you can find the whole list of dependencies in pyproject.toml file.

Installation

There are several ways to install Arthur on your system: packages or source code using Poetry or pip.

PyPI

Arthur can be installed using pip, a tool for installing Python packages. To do it, run the next command:

$ pip install kingarthur

Source code

To install from the source code you will need to clone the repository first:

$ git clone https://github.com/chaoss/grimoirelab-kingarthur
$ cd grimoirelab-kingarthur

Then use pip or Poetry to install the package along with its dependencies.

Pip

To install the package from local directory run the following command:

$ pip install .

In case you are a developer, you should install kingarthur in editable mode:

$ pip install -e .

Poetry

We use poetry for dependency management and packaging. You can install it following its documentation. Once you have installed it, you can install kingarthur and the dependencies in a project isolated environment using:

$ poetry install

To spaw a new shell within the virtual environment use:

$ poetry shell

Usage

arthurd

usage: arthurd [-c <file>] [-g] [-h <host>] [-p <port>] [-d <database>]
               [--es-index <index>] [--log-path <path>] [--archive-path <cpath>]
               [--no-archive] [--no-daemon] | --help

King Arthur commands his loyal knight Perceval on the quest
to retrieve data from software repositories.

This command runs an Arthur daemon that waits for HTTP requests
on port 8080. Repositories to analyze are added using an REST API.
Repositories are transformed into Perceval jobs that will be
scheduled and run using a distributed job queue platform.

optional arguments:
  -?, --help            show this help message and exit
  -c FILE, --config FILE
                        set configuration file
  -g, --debug           set debug mode on
  -h, --host            set the host name or IP address on which to listen for connections
  -p, --port            set listening TCP port (default: 8080)
  -d, --database        URL database connection (default: 'redis://localhost/8')
  -s, --sync            work in synchronous mode (without workers)
  --es-index            output ElasticSearch server index
  --log-path            path where logs are stored
  --archive-path        path to archive manager directory
  --no-archive          do not archive fetched raw data
  --no-daemon           do not run arthur in daemon mode

arthurd configuration file

To run arthurd using a configuration file:

$ arthurd [-c <file>]

Where <file> is the path to an ini file which uses the same parameters as in command line, but replacing underscores by hyphens. This configuration file has the following structure:

[arthur]
archive_path=/tmp/.arthur/archive
debug=True
log_path=/tmp/logs/arthurd
no_archive=True
sync_mode=True

[connection]
host=127.0.0.1
port=8080

[elasticsearch]
es_index=http://localhost:9200/items

[redis]
database=redis://localhost/8

arthurw

usage: arthurw [-g] [-d <database>] [--burst] [<queue1>...<queueN>] | --help

King Arthur's worker. It will run Perceval jobs on the quest
to retrieve data from software repositories.

positional arguments:
   queues               list of queues this worker will listen for
                        ('create' and 'update', by default)

optional arguments:
  -?, --help            show this help message and exit
  -g, --debug           set debug mode on
  -d, --database        URL database connection (default: 'redis://localhost/8')
  -b, --burst           Run in burst mode (quit after all work is done)

How to run it

The first step is to run a Redis server that will be used for communicating Arthur's components. Moreover, an Elastic Search server can be used to store the items generated by jobs. Please refer to their documentation to know how to install and run them both.

To run Arthur server:

$ arthurd -g -d redis://localhost/8 --es-index http://localhost:9200/items --log-path /tmp/logs/arthud --no-archive

To run a worker:

$ arthurw -d redis://localhost/8

Adding tasks

To add tasks to Arthur, create a JSON object containing the tasks needed to fetch data from a set of repositories. Each task will run a Perceval backend, thus, backend parameters will also needed for each task.

$ cat tasks.json
{
    "tasks": [
        {
            "task_id": "arthur.git",
            "backend": "git",
            "backend_args": {
                "gitpath": "/tmp/git/arthur.git/",
                "uri": "https://github.com/chaoss/grimoirelab-kingarthur.git",
                "from_date": "2015-03-01"
            },
            "category": "commit",
            "scheduler": {
                "delay": 10
            }
        },
        {
            "task_id": "bugzilla_mozilla",
            "backend": "bugzillarest",
            "backend_args": {
                "url": "https://bugzilla.mozilla.org/",
                "from_date": "2016-09-19"
            },
            "category": "bug",
            "archive": {
                "fetch_from_archive": true,
                "archived_after": "2018-02-26 09:00"
            },
            "scheduler": {
                "delay": 60,
                "max_retries": 5
            }
        }
    ]
}

Then, send this JSON stream to the server calling add method.

$ curl -H "Content-Type: application/json" --data @tasks.json http://127.0.0.1:8080/add

For this example, items will be stored in the items index on the Elastic Search server (http://localhost:9200/items).

Listing tasks

The list of tasks currently scheduled can be obtained using the method tasks.

$ curl http://127.0.0.1:8080/tasks

{
    "tasks": [
        {
            "backend_args": {
                "from_date": "2015-03-01T00:00:00+00:00",
                "uri": "https://github.com/chaoss/grimoirelab-kingarthur.git",
                "gitpath": "/tmp/santi/"
            },
            "backend": "git",
            "category": "commit",
            "created_on": 1480531707.810326,
            "task_id": "arthur.git",
            "scheduler": {
                "max_retries": 3,
                "delay": 10
            }
        }
    ]
}

Removing tasks

Scheduled tasks can also be removed calling to the server using the remove method. A JSON stream must be provided setting the identifiers of the tasks to be removed.

$ cat tasks_to_remove.json

{
    "tasks": [
        {
            "task_id": "bugzilla_mozilla"
        },
        {
            "task_id": "arthur.git"
        }
    ]
}

$ curl -H "Content-Type: application/json" --data @tasks_to_remove.json http://127.0.0.1:8080/remove

License

Licensed under GNU General Public License (GPL), version 3 or later.