Hello World! Worker

Get up and going with your first worker.

First, build the hello-world worker application.

$ go build

Run the application and see the worker in action.

$ cat tsk.json | ./worker > done.json 

The screen output should be very similar to:

2017/01/01 00:00:05 Hello World!

That output represents what the worker did. In this case the worker just logged the value of "info" from the task.

Now look at the contents of 'done.json'. It should look very similar to:

{"type":"hello-world","info":"Hello World!","created":"2017-01-01T00:00:01Z","result":"complete","msg":"task complete!","started":"2017-01-01T00:00:05Z","ended":"2017-01-01T00:00:06Z"}

Finally, look at the contents of 'tsk.json'

{"type":"hello-world","info":"Hello World!","created":"2017-01-01T00:00:01Z"}

A task is a signal that tells a worker to do something. The 'info' contents tell the worker the details of what to do.

Let's look at the pieces of a task object:

type

Identifies the type of work to be done. The 'hello-world' worker knows how to do tasks with type 'hello-world'. If the task were of a different type the worker application could choose to reject the task and send it back with an error message. The worker application may choose not to reject a task with the wrong type but instead try and work on the task anyway. That's ok, the 'type' in its most basic usage is just a way to communicate 'intent'. That is, the task is intended to be worked on by a worker that knows how to complete a task of that type. In our basic example we did not enforce only working on tasks of the right type. The launcher has the ability to handle tasks of the wrong type by either rejecting the task or discarding it.

info

Tells the worker more information about the specifics of completing the task. In this simple worker the info was used as the content of the log message.

To see it in action go ahead and change the 'info' value in tsk.json and then run the application again. What output do you see?

created

Simply describes when the task was created. This is good record keeping and you will see it invaluable when understanding and debugging your task ecosystem.

result

The result value is typically either "error" or "complete". Many other job systems try to include many kinds of statuses. We feel that many statuses is unnecessary. Part of this extra simplicity is made possible because Task is a stateless system.

msg

The result message. This can be anything the worker decides to make it.

started

The datetime that the worker started work on the task.

ended

The datetime that the worker ended work on the task.