Random GSI missing error "ValueError: Table <tablename> has no index: <index_name>"

[armujahid]

We randomly notice this error in AWS Lambda logs (and occasionally in local testing) where PynamoDB complains about a missing index that is not missing. Note that the index is a GSI. The stacktrace is attached.

    |   File "/opt/python/our_repo/models/resources/base.py", line 443, in list
    |     pynamodb_results = list(
    |                        ^^^^^
    |   File "/opt/python/pynamodb/pagination.py", line 189, in __next__
    |     self._get_next_page()
    |   File "/opt/python/pynamodb/pagination.py", line 175, in _get_next_page
    |     page = next(self.page_iter)
    |            ^^^^^^^^^^^^^^^^^^^^
    |   File "/opt/python/pynamodb/pagination.py", line 110, in __next__
    |     page = self._operation(*self._args, **self._kwargs)
    |            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    |   File "/opt/python/pynamodb/connection/table.py", line 256, in query
    |     return self.connection.query(
    |            ^^^^^^^^^^^^^^^^^^^^^^
    |   File "/opt/python/pynamodb/connection/base.py", line 1203, in query
    |     raise ValueError("Table {} has no index: {}".format(table_name, index_name))
    | ValueError: Table <tablename> has no index: <index_name>

PynamoDB Version: 6.1.0

Any clue why this might be happening? I am trying to create a minimal, complete, and verifiable example (MCVE), but so far I am not successful.