Add status check and/or locks for APIs that are executed asynchronously (index, export)

As submitted by a user:

reExportAll ... returns a message:

curl http://localhost:8080/api/admin/metadata/reExportAll
 {"status":"WORKFLOW_IN_PROGRESS"}

If I remember correctly the reindexing also returns a similar status message. If I see something like that I expect that I can recheck the status until I get a different status message informing that it is done. However when I search for reExportAll in the manual I haven't found a call where I can retrieve the current status. Is there such an API call?

Currently, the answer is no. There is no way to check on the status of either call via API (although the progress of indexAll can be checked by looking at the server log). It is also possible to start another indexAll or exportAll job when one is already running.

A possible solution would be to change this behavior under the model already used in the publish dataset API, that is also executed asynchronously: we return the same code 202/"WORKFLOW_IN_PROGRESS", but lock the dataset for the duration of the execution. The lock prevents another attempt to publish the dataset from being started. And makes it easy to check if the task has been completed.

All our currently implemented locks are dataset locks. But this may be worth introducing a new class of application locks, perhaps?

Start by reading the reExportAll, indexAll, and exportAll API behavior described in the issue, then compare it with the asynchronous publish dataset API and its existing dataset-lock model. Check the manual and server-log status information currently available. Done should include a defined way to check completion and behavior that prevents overlapping jobs.