Make work order-specific queues blocking (Run Concurrency)

Make work order-specific queues blocking (Enforce order/sequence);

Via the workflow settings panel, we want to enable users to the able to:

1. Specify that for a given workflow, work order 3 should not be processed until work order 2 reaches state = success || ignored
2. Allow users to specify that for a given workflow, work order 7 should not be processed until work order 4 reaches state = success || ignored if both meet the same key criterion. (This feature is required by Kafka trigger as well to enforce sequence in the how messages are processed by the consumer group)

Select from runs
	where state = available
	and no other workorders for this workflow with dataclip.body.arbitrary_key is yet
        to have state  = [success|skipped]

3. Allow users to specify concurrency per-arbitrary-dataclip-key-per-workflow.

So if we have 10 runs enqueued (3 with the same arbitrary-data clip-value: dataclip.body.xyx = “myguy!”) 
then we’d do 8-at-a-time, but keep the 2 other dataclip.body.xyx = “myguy!” containing runs BLOCKED
 until the first was done.

4. Allow users to mark work orders in the ignored state

update work order set state = ‘ignored’ where wo.id = this_wo_id

5. Allow users to clear (set to ignore, never process) pending runs/workorders in their queue

update runs set state = ‘skipped’ where runs.state = ‘available’
update work order state if the skipped run is only run for that workorder

6. Maybe make it easier to find work orders to retry that include skipped runs?

Start with the workflow settings panel, the run/work-order queue behavior, and the Kafka trigger or consumer-group entry point named in the issue. The issue names no files or tests; done would require agreeing on and implementing the six listed ordering, concurrency, ignored-state, clearing, and retry behaviors.