I've got a partitioned asset (

group

) with a non-partitioned asset as a downstream dependency (

all_groups

). I've then got an asset reconciliation sensor on

all_groups

. When ever I launch a backfill on

group

, I get many runs kicked off by the asset reconciliation sensor. Is there a way I can make it wait for all the queued runs of

group

to complete before it launches the reconciliation run?

I think that the best option we have for this right now would be also select

all_groups

when you launch your backfill. As of Dagster's latest release,

all_groups

won't be reconciled if there's a backfill targeting it.

will do 🙂