Test cases of Upload reorg with PSC using reindex queue

[yuanzhou]

Previously the PROD VM was overloaded due to PSC issued lots of reindex calls and update calls as part of the Upload reorg. And we had to disable the reindex on Upload reorg then to avoid enity-api from being overloaded. Then batch reindex all the datasets.

• Set Upload to Reorganize, reindex=false
• Update Upload with dataset_uuids, reindex=false, Gateway timeout
• Update Upload with Error, reindex=true

Following the initial release of reindex job queue, work with PSC team to test Upload reorg with a large number (500 - 700 ish) of associated datasets and use ?reindex-priority=3 to test out the performance improvements.