Skip to content

Latest commit

 

History

History
50 lines (33 loc) · 2.43 KB

README.md

File metadata and controls

50 lines (33 loc) · 2.43 KB

cron-metrics

Implementation of a monitoring system of interval cron tasks in Python. Useful for metrics tasks that need to run every n minutes.

Implementation requiring Plan and Crab.

  • Plan to create the crontab.
  • Crab to provide a dashboard to monitor the tasks.

####Installation 0. Install sqlite3

  1. pip install -r requirements.txt
  2. Crab specific steps:

2.1 Need to add a crabd.ini file in either ~/.crab or /etc/crab.

2.2 recreate crabdb.db:

  `% sqlite3 crab.db < doc/schema.txt`

2.3 Download jquery and Font Awesome and install them on the res folder (specified in crabd.ini).

2.4 Port 8000 needs to be accesible on the machine (if using other port, need to change it in crabd.ini)

  1. Modify config.yml (example included in config.yml.example). 3.1. Change the plan user to the user in charge of running the crontab. 3.2. Change the path in the plan section to the path where your modules folder will be located.

####Usage

  1. To add a task:
  • Create a folder on the /modules folder. That task should have one entrypoint (one file run as __main__). (preferably named module_main.py.

  • Add that task on the task section of config.yml.

    • the name of the task should be the name of the folder you created for it.
    • task_specs should include the plan dsl for setting up a cron.
    • If you need to import some task specific variables, you can either use the environment paramenter on the task config, or use the utils.get_task_config(task_name) that returns an Env object with the task config as attributes.
  • Use python start_cron.py write to create the crontab with all the tasks, and start the crab dashboard. You can go to localhost:8000 to see it in action.

  • Use python start_cron.py clear to empy the crontab, clean the /logs folder and kill crabd

####Screenshots

By using the utils.crab_task context manager, every task gets sent to crab, along with their stdout and stderr.

You can see the history of each task.