I have two dagster jobs (running ML training and predictions). Is it possible to now create a graph that links the two dagster jobs? I want to have it also in a way that if one job fails, the subsequent one also does not get executed.

commenting to follow the discussion

you can link the jobs using a run status sensor. The idea is that you could write the sensor to run when the first job succeeds, and the sensor would launch a run of the second job https://docs.dagster.io/concepts/partitions-schedules-sensors/sensors#run-status-sensors

I will have a look!

yeah, you can use the

monitored_jobs

argument to the

run_status_sensor

decorator to specify the jobs to watch

Ah ok, but if I have multiple instances of the same job running, would it react to all of them? Can I tie it to a particular instance of the job?