Would agree with this. The use case is particularly apparent if we have multiple evening tasks that use the bulk api. If we trigger all of those to run at midnight, one will run and the others will error out.

Current work around is to offset these runs with the scheduler, but we never know exactly how long those bulk runs will take.

Would love to see a queue for this instead of it simply failing.

Maybe if there was an internal queue for bulk operations that could be nice? Since it's possible to query for the current bulk operation status, the internal queue could just add bulk operations to the queue while waiting for the others to finish. Maybe this could be a configuration option on the store level or even on the action level -- some people might want to know if they're trying to do concurrent bulk ops and have it error.