Global workqueue throwing an exception when the database is compacting

[amaltaro]

Impact of the bug
Global workqueue

Describe the bug
When central workqueue/workqueue_inbox database is compacting, some tasks might be skipped until the compaction is over.
With that, it can cause a TypeError exception with unexpected data type returned by a given function, as can be seen in this log [1]

How to reproduce it
Run global workqueue while couchdb is compacting the workqueue database.

Expected behavior
Properly deal with this scenario, i.e., workClosed should be 0 instead of None.

Additional context and error message
[1]

2022-06-17 20:26:48,997:INFO:WorkQueueReqMgrInterface:Closing open requests
2022-06-17 20:26:49,023:INFO:WorkQueueBackend:CouchDB compacting - try again later.
2022-06-17 20:26:49,023:WARNING:WorkQueue:Backend busy or down: Can not close work at this time
2022-06-17 20:26:49,023:ERROR:WorkQueueReqMgrInterface:Generic error while closing open requests. Details: object of type 'NoneType' has no len()
Traceback (most recent call last):
  File "/data/srv/HG2206e/sw/slc7_amd64_gcc630/cms/workqueue/2.0.4.pre4/lib/python3.8/site-packages/WMCore/WorkQueue/WorkQueueReqMgrInterface.py", line 63, in __call__
    msg += "Work closed: %d, " % len(workClosed)
TypeError: object of type 'NoneType' has no len()
2022-06-17 20:26:49,024:INFO:WorkQueueReqMgrInterface:Adding new elements to open requests