CI: Mark s3 tests parallel safe (#35895)

Closes #35856

pandas/tests/io/conftest.py

@@ -34,12 +34,13 @@ def feather_file(datapath):
 
 
 @pytest.fixture
-def s3so():
-    return dict(client_kwargs={"endpoint_url": "http://127.0.0.1:5555/"})
+def s3so(worker_id):
+    worker_id = "5" if worker_id == "master" else worker_id.lstrip("gw")
+    return dict(client_kwargs={"endpoint_url": f"http://127.0.0.1:555{worker_id}/"})
 
 
-@pytest.fixture(scope="module")
-def s3_base():
+@pytest.fixture(scope="session")
+def s3_base(worker_id):
     """
     Fixture for mocking S3 interaction.
 
@@ -61,11 +62,13 @@ def s3_base():
         # Launching moto in server mode, i.e., as a separate process
         # with an S3 endpoint on localhost
 
-        endpoint_uri = "http://127.0.0.1:5555/"
+        worker_id = "5" if worker_id == "master" else worker_id.lstrip("gw")
+        endpoint_port = f"555{worker_id}"
+        endpoint_uri = f"http://127.0.0.1:{endpoint_port}/"
 
         # pipe to null to avoid logging in terminal
         proc = subprocess.Popen(
-            shlex.split("moto_server s3 -p 5555"), stdout=subprocess.DEVNULL
+            shlex.split(f"moto_server s3 -p {endpoint_port}"), stdout=subprocess.DEVNULL
         )
 
         timeout = 5
@@ -79,7 +82,7 @@ def s3_base():
                 pass
             timeout -= 0.1
             time.sleep(0.1)
-        yield
+        yield endpoint_uri
 
         proc.terminate()
         proc.wait()
@@ -119,9 +122,8 @@ def add_tips_files(bucket_name):
                 cli.put_object(Bucket=bucket_name, Key=s3_key, Body=f)
 
     bucket = "pandas-test"
-    endpoint_uri = "http://127.0.0.1:5555/"
-    conn = boto3.resource("s3", endpoint_url=endpoint_uri)
-    cli = boto3.client("s3", endpoint_url=endpoint_uri)
+    conn = boto3.resource("s3", endpoint_url=s3_base)
+    cli = boto3.client("s3", endpoint_url=s3_base)
 
     try:
         cli.create_bucket(Bucket=bucket)
@@ -143,7 +145,7 @@ def add_tips_files(bucket_name):
     s3fs.S3FileSystem.clear_instance_cache()
     yield conn
 
-    s3 = s3fs.S3FileSystem(client_kwargs={"endpoint_url": "http://127.0.0.1:5555/"})
+    s3 = s3fs.S3FileSystem(client_kwargs={"endpoint_url": s3_base})
 
     try:
         s3.rm(bucket, recursive=True)

pandas/tests/io/json/test_compression.py

@@ -34,7 +34,7 @@ def test_read_zipped_json(datapath):
 
 
 @td.skip_if_not_us_locale
-def test_with_s3_url(compression, s3_resource):
+def test_with_s3_url(compression, s3_resource, s3so):
     # Bucket "pandas-test" created in tests/io/conftest.py
 
     df = pd.read_json('{"a": [1, 2, 3], "b": [4, 5, 6]}')
@@ -45,9 +45,7 @@ def test_with_s3_url(compression, s3_resource):
             s3_resource.Bucket("pandas-test").put_object(Key="test-1", Body=f)
 
     roundtripped_df = pd.read_json(
-        "s3://pandas-test/test-1",
-        compression=compression,
-        storage_options=dict(client_kwargs={"endpoint_url": "http://127.0.0.1:5555/"}),
+        "s3://pandas-test/test-1", compression=compression, storage_options=s3so,
     )
     tm.assert_frame_equal(df, roundtripped_df)
 

pandas/tests/io/json/test_pandas.py

@@ -1702,17 +1702,14 @@ def test_json_multiindex(self, dataframe, expected):
         result = series.to_json(orient="index")
         assert result == expected
 
-    def test_to_s3(self, s3_resource):
+    def test_to_s3(self, s3_resource, s3so):
         import time
 
         # GH 28375
         mock_bucket_name, target_file = "pandas-test", "test.json"
         df = DataFrame({"x": [1, 2, 3], "y": [2, 4, 6]})
         df.to_json(
-            f"s3://{mock_bucket_name}/{target_file}",
-            storage_options=dict(
-                client_kwargs={"endpoint_url": "http://127.0.0.1:5555/"}
-            ),
+            f"s3://{mock_bucket_name}/{target_file}", storage_options=s3so,
         )
         timeout = 5
         while True:

pandas/tests/io/test_parquet.py

@@ -158,10 +158,6 @@ def check_round_trip(
     """
     write_kwargs = write_kwargs or {"compression": None}
     read_kwargs = read_kwargs or {}
-    if isinstance(path, str) and "s3://" in path:
-        s3so = dict(client_kwargs={"endpoint_url": "http://127.0.0.1:5555/"})
-        read_kwargs["storage_options"] = s3so
-        write_kwargs["storage_options"] = s3so
 
     if expected is None:
         expected = df
@@ -555,15 +551,24 @@ def test_s3_roundtrip_explicit_fs(self, df_compat, s3_resource, pa, s3so):
             write_kwargs=kw,
         )
 
-    def test_s3_roundtrip(self, df_compat, s3_resource, pa):
+    def test_s3_roundtrip(self, df_compat, s3_resource, pa, s3so):
         if LooseVersion(pyarrow.__version__) <= LooseVersion("0.17.0"):
             pytest.skip()
         # GH #19134
-        check_round_trip(df_compat, pa, path="s3://pandas-test/pyarrow.parquet")
+        s3so = dict(storage_options=s3so)
+        check_round_trip(
+            df_compat,
+            pa,
+            path="s3://pandas-test/pyarrow.parquet",
+            read_kwargs=s3so,
+            write_kwargs=s3so,
+        )
 
     @td.skip_if_no("s3fs")
     @pytest.mark.parametrize("partition_col", [["A"], []])
-    def test_s3_roundtrip_for_dir(self, df_compat, s3_resource, pa, partition_col):
+    def test_s3_roundtrip_for_dir(
+        self, df_compat, s3_resource, pa, partition_col, s3so
+    ):
         # GH #26388
         expected_df = df_compat.copy()
 
@@ -587,7 +592,10 @@ def test_s3_roundtrip_for_dir(self, df_compat, s3_resource, pa, partition_col):
             pa,
             expected=expected_df,
             path="s3://pandas-test/parquet_dir",
-            write_kwargs={"partition_cols": partition_col, "compression": None},
+            read_kwargs=dict(storage_options=s3so),
+            write_kwargs=dict(
+                partition_cols=partition_col, compression=None, storage_options=s3so
+            ),
             check_like=True,
             repeat=1,
         )
@@ -761,9 +769,15 @@ def test_filter_row_groups(self, fp):
             result = read_parquet(path, fp, filters=[("a", "==", 0)])
         assert len(result) == 1
 
-    def test_s3_roundtrip(self, df_compat, s3_resource, fp):
+    def test_s3_roundtrip(self, df_compat, s3_resource, fp, s3so):
         # GH #19134
-        check_round_trip(df_compat, fp, path="s3://pandas-test/fastparquet.parquet")
+        check_round_trip(
+            df_compat,
+            fp,
+            path="s3://pandas-test/fastparquet.parquet",
+            read_kwargs=dict(storage_options=s3so),
+            write_kwargs=dict(compression=None, storage_options=s3so),
+        )
 
     def test_partition_cols_supported(self, fp, df_full):
         # GH #23283