Skip to content

morgangallant/scheduler

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

16 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Deploy on Railway

A simple job scheduler backed by Postgres used in production at https://operand.ai. Setup needs two environment variables, SECRET and ENDPOINT. The secret is attached to incoming/outgoing requests within the Scheduler-Secret HTTP header, and is used both by the scheduler to verify that incoming request is legit, as well as the end-application to verify the request is coming from the scheduler. The endpoint is simply the URL to send HTTP requests to.

This scheduler also has support for CRON-type expressions. This allows the application to specify a set of id's to run on a schedule. You should probably just make a request on application start with all the IDs and the relevant CRON expressions. This will tear-down the world and re-start the CRON server with the new values. This is how you should do it, especially in auto-deployment environments. When a cron job fires, the application will get a POST with the "cron_id" set. You should check for this in the message and respond appropriately.

This is used in production at Operand.

Example Usage:

Scheduling a Job

POST https://scheduler.yourcompany.com/insert
Headers: Scheduler-Secret=XXXXXX

{
    "timestamp": "2021-04-15T14:17:00-07:00",
    "body": {
        "foo": "bar"
    }
}

Response:
{"id":"cknjdu2k300153zugmucamxxo"}

Cancelling a Scheduled Job (note: if the job doesn't exist, this will fail silently.)

POST https://scheduler.yourcompany.com/delete
Headers: Scheduler-Secret=XXXXXX

{
    "id":"cknjdu2k300153zugmucamxxo"
}

Response: None

Configuring CRON Jobs

POST https://scheduler.yourcompany.com/cron
Headers: Scheduler-Secret=XXXXXX

{
    "jobs":[
        {
            "id":"hello_world",
            "spec": "30 * * * * *"
        },
        {
            "id":"hello_world_2",
            "spec": "0 * * * * *"
        }
    ]
}

Response: None

When a scheduled job, or a CRON job fires, your application (located at endpoint) will get the following message in a POST:

{
  "id": "<job id>", // will be empty if cron_id non-empty
  "cron_id": "<cron id>", // will be empty if id non-empty
  "body": { // Omitted for CRON jobs.
    ... // Your stuff here
  }
}

And that's it! Feel free to file issues or do pull requests.

About

A simple job scheduler backed by Postgres.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published