For the case of hivemind, we're always getting the documents and processing them. In case of running a DAG tasks in parallel, the data could be mixed up and for a single document, we might end up having duplicate or no nodes for it. We should ensure that a DAG won't be run in parallel.
For the case of hivemind, we're always getting the documents and processing them. In case of running a DAG tasks in parallel, the data could be mixed up and for a single document, we might end up having duplicate or no nodes for it. We should ensure that a DAG won't be run in parallel.