Veggiecron Server: a healthier alternative to cron

Veggiecron (inspired by the unix crontab and celery) is a Python server for scheduling jobs via REST API. Standing on the shoulders of such giants as asyncio and tornado, veggiecron is able to run thousands of jobs a minute.

Contents

• Installation
• Usage
  • Start the Server
  • Register a User
  • Login (Get an Auth Token)
  • Create a Job
  • Configuration
• In-Depth Docs
  • Job Types
  • Schedule String Format

Installation

Veggiecron will not be available via pip until we reach the version 1.0 milestone For those brave enough to run the "pre-v1" product, follow these instructions:

# Ensure you have Python 3.5+
$ python --version

# Clone the repo
$ git clone https://github.com/jacobbridges/veggiecron-server.git && cd veggiecron-server

# (OPTIONAL) Create a virtual environment to run the project
$ virtualenv venv -p python3.5
$ ./venv/bin/activate

# Install the dependencies
$ pip install -r requirements.txt

Usage

Start the Server

Simple and clean:

$ python3 start.py

The server will be started on 0.0.0.0:8118.

Register a User

$ http --form POST localhost:8118/register username=test password=test
{
    "data": {
        "date_created": 1476675574.624803,
        "date_updated": 1476675574.624803,
        "username": "test"
    },
    "description": "You are registered as \"test\"! Proceed to login.",
    "id": "success"
}

Login (Get an Auth Token)

Auth tokens are required to perform any actions via REST on Veggiecron. These tokens can be retrieved from the /login endpoint, and are used in the request header under X-Auth-Token:

$ http --form POST localhost:8118/login username=test password=test
{
    "data": {
        "token": "MmIxNTYwNOM2ZmZlNjQ4YzXkNWMyM2E1ZDEzMDZlMWU6dGVzdA=="
    },
    "description": "You have successfully generated an auth token! Check the data key.",
    "id": "success"
}

Create a Job

Simple example for creating a job:

$ http --form POST localhost:8118/job X-Auth-Token:<token> name=<name> type=<type> data=<data> schedule=<schedule>
{
    "data": {
        "data": "<data>",
        "date_created": <unix-timestamp>,
        "date_updated": <unix-timestamp>,
        "name": "<name>",
        "schedule": "<schedule>",
        "type": "<type>"
    },
    "description": "Successfully created http job: \"<name>\"",
    "id": "success"
}

The above code example will raise more questions than it answers. Here is a list of relevant links you might want to check out:

• Job Types
• Schedule String Format

Configuration

High-level configurations can be found in the config.yaml file. Descriptions of each config are in the following table:

Config	Description
app.env	Application environment. Defaults to "development".
app.name	If you don't like "veggiecron-server".
app.key	Application key used to hash passwords. Be sure to generate your own!
host	Host to run the server on.
port	Port to run the server on.
db_file	Name of the SQLite3 database file.

In-Depth Docs

This section is for more detailed documentation, while the previous sections for skimming readers.

Job Types

The following job types are available:

Type	Description	Data
		Field	Type	Description
http	Job that makes HTTP requests.	url	str	URL to be requested. (must begin with "http://")
		number_of_clones	int	Number of reqeusts made on every job run.
		enable_shadows	bool	Run the job again even if a previous run is still processing.
		verb	str	HTTP method (GET, POST, DELETE, etc.)

Schedule String Format

A job's "schedule" property must follow the following syntax:

• every hour
• every x hours
• every minute
• every x minutes
• every x seconds
• every day @ 13:00
• every x days @ 7:30
• once @ <utc-timestamp>