Concepts

In this section, you would know the core concepts of PyDolphinScheduler.

Workflow

Workflow describe the whole things except tasks and tasks dependence, which including name, schedule interval, schedule start time and end time. You would know scheduler

Workflow could be initialized in normal assign statement or in context manger.

# Initialization with assign statement
workflow = Workflow(name="my first workflow")

# Or context manger
with Workflow(name="my first workflow") as workflow:
    workflow.submit()

Workflow is the main object communicate between PyDolphinScheduler and DolphinScheduler daemon. After workflow and task is be declared, you could use submit and run notify server your definition.

If you just want to submit your definition and create workflow, without run it, you should use attribute submit. But if you want to run the workflow after you submit it, you could use attribute run.

# Just submit definition, without run it
workflow.submit()

# Both submit and run definition
workflow.run()

Schedule

We use parameter schedule determine the schedule interval of workflow, PyDolphinScheduler support seven asterisks expression, and each of the meaning of position as below

* * * * * * *
┬ ┬ ┬ ┬ ┬ ┬ ┬
│ │ │ │ │ │ │
│ │ │ │ │ │ └─── year
│ │ │ │ │ └───── day of week (0 - 7) (0 to 6 are Sunday to Saturday, or use names; 7 is Sunday, the same as 0)
│ │ │ │ └─────── month (1 - 12)
│ │ │ └───────── day of month (1 - 31)
│ │ └─────────── hour (0 - 23)
│ └───────────── min (0 - 59)
└─────────────── second (0 - 59)

Here we add some example crontab:

  • 0 0 0 * * ? *: Workflow execute every day at 00:00:00.

  • 10 2 * * * ? *: Workflow execute hourly day at ten pass two.

  • 10,11 20 0 1,2 * ? *: Workflow execute first and second day of month at 00:20:10 and 00:20:11.

Tenant

Tenant is the user who run task command in machine or in virtual machine. it could be assign by simple string.

#
workflow = Workflow(name="workflow tenant", tenant="tenant_exists")

Note

Make should tenant exists in target machine, otherwise it will raise an error when you try to run command

Execution Type

Decision which behavior to run when workflow have multiple instances. when workflow schedule interval is too short, it may cause multiple instances run at the same time. We can use this parameter to control the behavior about how to run those workflow instances. Currently we have four execution type:

  • parallel (default value): it means all instances will allow to run even though the previous instance is not finished.

  • serial_wait: it means the all instance will wait for the previous instance to finish, and all the waiting instances will be executed base on scheduling order.

  • serial_discard: it means the all instance will be discard(abandon) if the previous instance is not finished.

  • serial_priority: it means the all instance will wait for the previous instance to finish, and all the waiting instances will be executed base on workflow priority order.

Parameter execution type can be set in

  • Direct assign statement. You can pick execute type from above and direct assign to parameter execution_type.

    workflow = Workflow(
    name="workflow_name",
    execution_type="parallel"
)

  • Via environment variables, configurations file setting, for more detail about those way setting, you can see you can read Configuration section.

Tasks

Task is the minimum unit running actual job, and it is nodes of DAG, aka directed acyclic graph. You could define what you want to in the task. It have some required parameter to make uniqueness and definition.

Here we use pydolphinscheduler.tasks.Shell() as example, parameter name and command is required and must be provider. Parameter name set name to the task, and parameter command declare the command you wish to run in this task.

# We named this task as "shell", and just run command `echo shell task`
shell_task = Shell(name="shell", command="echo shell task")

If you want to see all type of tasks, you could see Tasks.

Tasks Dependence

You could define many tasks in on single Workflow. If all those task is in parallel processing, then you could leave them alone without adding any additional information. But if there have some tasks should not be run unless pre task in workflow have be done, we should set task dependence to them. Set tasks dependence have two mainly way and both of them is easy. You could use bitwise operator >> and <<, or task attribute set_downstream and set_upstream to do it.

# Set task1 as task2 upstream
task1 >> task2
# You could use attribute `set_downstream` too, is same as `task1 >> task2`
task1.set_downstream(task2)

# Set task1 as task2 downstream
task1 << task2
# It is same as attribute `set_upstream`
task1.set_upstream(task2)

# Beside, we could set dependence between task and sequence of tasks,
# we set `task1` is upstream to both `task2` and `task3`. It is useful
# for some tasks have same dependence.
task1 >> [task2, task3]

Task With Workflow

In most of data orchestration cases, you should assigned attribute workflow to task instance to decide workflow of task. You could set workflow in both normal assign or in context manger mode

# Normal assign, have to explicit declaration and pass `Workflow` instance to task
workflow = Workflow(name="my first workflow")
shell_task = Shell(name="shell", command="echo shell task", workflow=workflow)

# Context manger, `Workflow` instance workflow would implicit declaration to task
with Workflow(name="my first workflow") as workflow:
    shell_task = Shell(name="shell", command="echo shell task",

With both Workflow, Tasks and Tasks Dependence, we could build a workflow with multiple tasks.

Resource Files

During workflow running, we may need some resource files to help us run task usually. One of a common situation is that we already have some executable files locally, and we need to schedule in specific time, or add them to existing workflow by adding the new tasks. Of cause, we can upload those files to target machine and run them in shell task by reference the absolute path of file. But if we have more than one machine to run task, we have to upload those files to each of them. And it is not convenient and not flexible, because we may need to change our resource files sometimes.

The more pydolphinscheduler way is to upload those files together with workflow, and use them in task to run. For example, you have a bash script named echo-ten.sh locally, and it contains some code like this:

#!/bin/env bash
max=10
for ((i=1; i <= $max; ++i)); do
    echo "$i"
done

and you want to use it in workflow but do not want to copy-paste it to shell task parameter command. You could use this mechanism to upload it to resource center when you create workflow

# Read file content
file_name = "echo-ten.sh"

with open(file_name, "r") as f:
      content = f.read()

with Workflow(
   name="upload_and_run",
   resource_list=[
      Resource(name=file_name, content=content),
   ],
) as workflow:

And when we call workflow.run() the new file named echo-ten.sh would be uploaded to dolphinscheduler resource center.

After that we can use this file in our task by reference it by name, in this case we could use shell task to run it.

# We use `shell` task to run `echo-ten.sh` file
shell_task = Shell(
   name="run",
   command=f"bash {file_name}",
   resource_list=[
      file_name
   ],
)

During workflow running, the file would be downloaded to the task runtime working directory which mean you could execute them. We execute file by bash but reference it by name directly.

Please notice that we could also reference the resource file already in dolphinscheduler resource center, which mean we could use resource center files in task by name without upload it again. And we can upload files to resource center bare without workflow.

# Upload file to resource center
from pydolphinscheduler.core.resource import Resource

resource = Resource(name="bare-create.py", user_name="<USER-MUST-EXISTS-WITH-TENANT>", content="print('Bareh create resource')")
resource.create_or_update_resource()

After that, we could see new file named bare-create.py is be created in resource center.

Note

Both parameter resource_list in workflow and task is list of string which mean you could upload and reference multiple files. For more complex usage, you could read Upload and Use Multiple Resources.

Authentication Token

pydolphinscheduler use token as authentication when communication with dolphinscheduler server, and we have a default auth token to make it out-of-box. For security reason, we highly recommend you to change your own auth token when you deploy in production environment or test dolphinscheduler in public network. The auth token keyword in auth_token and it can be set in multiple ways which you can read Configuration section for more detail.