Hi team. Quick question. If I define two jobs (as shown below) how do combine them so that they can be imported into my repo.py and executed one after another?

# This is the job to run first
@job(...)
def run_first():
    dbt_seed_op()

# This is the job to run second
run_second = AssetGroup(...).build_job("run_second")

When I try the following pattern I get an error:

@job
def combined_job():
    run_second(run_first())
>>>

dagster.core.errors.DagsterInvariantViolationError: Attempted to call job 'combined_job' directly. Jobs should be invoked by using an execution API function (e.g. `job.execute_in_process`).

I could not find an example of the

job.execute_in_process

method in the docs. Does anyone know how to get around this?

You can't invoke jobs to be run within other jobs - perhaps what you want is to have

run_first

and

run_second

be graphs, and then orchestrate them within the same graph?

@graph
def run_first():
    dbt_seed_op()

@graph
def run_second():
    ...

@job
def combined_job():
    run_second(run_first())

If you need

run_first

and

run_second

to be different jobs for whatever reason, then you can set up

run_second

to be triggered by

run_first

by a sensor.