Defining schedules

tip

You can scaffold schedules from the command line by running dg scaffold defs dagster.schedule <path/to/schedule_file.py>. For more information, see the dg CLI docs.

Defining basic schedules

The following examples demonstrate how to define some basic schedules.

Using ScheduleDefinition

This example demonstrates how to define a schedule using ScheduleDefinition that will run all assets every day at midnight.

src/<project_name>/defs/assets.py

basic_schedule = dg.ScheduleDefinition(
    name="basic_schedule", cron_schedule="0 0 * * *", target="*"
)

note

The cron_schedule argument accepts standard cron expressions. If your croniter dependency's version is >= 1.0.12, the argument will also accept the following:

• @daily
• @hourly
• @monthly

Using @schedule

This example demonstrates how to define a schedule using @dg.schedule, which provides more flexibility than ScheduleDefinition. For example, you can configure job behavior based on its scheduled run time or emit log messages.

@schedule(target="*", cron_schedule="0 0 * * *")
def basic_schedule(): ...
  # things the schedule does, like returning a RunRequest or SkipReason

note

The cron_schedule argument accepts standard cron expressions. If your croniter dependency's version is >= 1.0.12, the argument will also accept the following:

• @daily
• @hourly
• @monthly

Emitting log messages from schedule evaluation

This example demonstrates how to emit log messages from a schedule during its evaluation function. These logs will be visible in the UI when you inspect a tick in the schedule's tick history.

src/<project_name>/defs/schedules.py

@dg.schedule(target="*", cron_schedule="* * * * *")
def logs_then_skips(context):
    context.log.info("Logging from a dg.schedule!")
    return dg.SkipReason("Nothing to do")

note

Schedule logs are stored in your Dagster instance's compute log storage. You should ensure that your compute log storage is configured to view your schedule logs.