CI: specified bucket does not exist in TestParquetPyArrow.test_s3_roundtrip_explicit_fs

[charlesdong1991]

might be related to #35655

================================== FAILURES ===================================
______________ TestParquetPyArrow.test_s3_roundtrip_explicit_fs _______________
[gw2] win32 -- Python 3.8.5 C:\Miniconda\envs\pandas-dev\python.exe

self = <pandas.tests.io.test_parquet.TestParquetPyArrow object at 0x000002B24268BF10>
df_compat =    A    B
0  1  foo
1  2  foo
2  3  foo
s3_resource = s3.ServiceResource(), pa = 'pyarrow'
s3so = {'client_kwargs': {'endpoint_url': 'http://127.0.0.1:5555/'}}

    def test_s3_roundtrip_explicit_fs(self, df_compat, s3_resource, pa, s3so):
        s3fs = pytest.importorskip("s3fs")
        if LooseVersion(pyarrow.__version__) <= LooseVersion("0.17.0"):
            pytest.skip()
        s3 = s3fs.S3FileSystem(**s3so)
        kw = dict(filesystem=s3)
>       check_round_trip(
            df_compat,
            pa,
            path="pandas-test/pyarrow.parquet",
            read_kwargs=kw,
            write_kwargs=kw,
        )

        service_id = self._service_model.service_id.hyphenize()
        handler, event_response = self.meta.events.emit_until_response(
            'before-call.{service_id}.{operation_name}'.format(
                service_id=service_id,
                operation_name=operation_name),
            model=operation_model, params=request_dict,
            request_signer=self._request_signer, context=request_context)
    
        if event_response is not None:
            http, parsed_response = event_response
        else:
            http, parsed_response = self._make_request(
                operation_model, request_dict, request_context)
    
        self.meta.events.emit(
            'after-call.{service_id}.{operation_name}'.format(
                service_id=service_id,
                operation_name=operation_name),
            http_response=http, parsed=parsed_response,
            model=operation_model, context=request_context
        )
    
        if http.status_code >= 300:
            error_code = parsed_response.get("Error", {}).get("Code")
            error_class = self.exceptions.from_code(error_code)
>           raise error_class(parsed_response, operation_name)
E           botocore.errorfactory.NoSuchBucket: An error occurred (NoSuchBucket) when calling the PutObject operation: The specified bucket does not exist

[alimcmaster1]

Seeing this in a few more builds https://dev.azure.com/pandas-dev/pandas/_build/results?buildId=40825&view=logs&j=4b4d380c-92b5-58cf-5b80-2c6c9b1d5805&t=8eab1181-1335-58ed-7ed5-a271d9d86005

Ref: #35029

[TomAugspurger]

@martindurant will you have time to look into this? I'm unsure if it's related to your recent moto changes.

[martindurant]

Are any tests being run in parallel here? That was suggested elsewhere as a possibility, and would for certain cause races.

[TomAugspurger]

Yes, I think that is in parallel. I'll try putting up a PR marking them as single to see if that fixes things.

[martindurant]

The fixture could be changed to produce a random bucket for each test function, which would take some work to get the right kwargs to each s3 call.

[TomAugspurger]

I tried that briefly, but it's a bit hard to do the synchronization since we need to start the moto server exactly once, and it needs to live until all the workers are done.

Probably best to just mark these as single.