db: Remove use of 'bind' arguments

Resolve the following RemovedIn20Warning warnings:

  The MetaData.bind argument is deprecated and will be removed in
  SQLAlchemy 2.0.

  The ``bind`` argument for schema methods that invoke SQL against an
  engine or connection will be required in SQLAlchemy 2.0.

We must retain a couple of workarounds to keep sqlalchemy-migrate happy.
This shouldn't be an issue since this library is not compatible with
SQLAlchemy 2.x and will have to be removed (along with the legacy
migration files themselves) when we eventually support 2.x.

Change-Id: I946b4c7926ba2583b84ab95a1fe7aedc6923d9f8
Signed-off-by: Stephen Finucane <[email protected]>

Author: stephenfin

nova/db/api/legacy_migrations/versions/067_train.py

@@ -27,6 +27,9 @@ def InetSmall():
 
 def upgrade(migrate_engine):
     meta = sa.MetaData()
+    # NOTE(stephenfin): This is not compatible with SQLAlchemy 2.0 but neither
+    # is sqlalchemy-migrate which requires this. We'll remove these migrations
+    # when dropping SQLAlchemy < 2.x support
     meta.bind = migrate_engine
 
     cell_mappings = sa.Table('cell_mappings', meta,

nova/db/main/api.py

@@ -4233,8 +4233,9 @@ def _get_fk_stmts(metadata, conn, table, column, records):
     return inserts, deletes
 
 
-def _archive_deleted_rows_for_table(metadata, tablename, max_rows, before,
-                                    task_log):
+def _archive_deleted_rows_for_table(
+    metadata, engine, tablename, max_rows, before, task_log,
+):
     """Move up to max_rows rows from one tables to the corresponding
     shadow table.
 
@@ -4249,7 +4250,7 @@ def _archive_deleted_rows_for_table(metadata, tablename, max_rows, before,
         - number of extra rows archived (due to FK constraints)
           dict of {tablename: rows_archived}
     """
-    conn = metadata.bind.connect()
+    conn = engine.connect()
     # NOTE(tdurakov): table metadata should be received
     # from models, not db tables. Default value specified by SoftDeleteMixin
     # is known only by models, not DB layer.
@@ -4382,8 +4383,9 @@ def archive_deleted_rows(context=None, max_rows=None, before=None,
     table_to_rows_archived = collections.defaultdict(int)
     deleted_instance_uuids = []
     total_rows_archived = 0
-    meta = sa.MetaData(get_engine(use_slave=True, context=context))
-    meta.reflect()
+    meta = sa.MetaData()
+    engine = get_engine(use_slave=True, context=context)
+    meta.reflect(bind=engine)
     # Get the sorted list of tables in order of foreign key dependency.
     # Process the parent tables and find their dependent records in order to
     # archive the related records in a single database transactions. The goal
@@ -4409,7 +4411,7 @@ def archive_deleted_rows(context=None, max_rows=None, before=None,
 
         rows_archived, _deleted_instance_uuids, extras = (
             _archive_deleted_rows_for_table(
-                meta, tablename,
+                meta, engine, tablename,
                 max_rows=max_rows - total_rows_archived,
                 before=before,
                 task_log=task_log))
@@ -4437,8 +4439,7 @@ def purge_shadow_tables(context, before_date, status_fn=None):
     engine = get_engine(context=context)
     conn = engine.connect()
     metadata = sa.MetaData()
-    metadata.bind = engine
-    metadata.reflect()
+    metadata.reflect(bind=engine)
     total_deleted = 0
 
     if status_fn is None:

nova/db/main/legacy_migrations/versions/402_train.py

@@ -45,10 +45,13 @@ def process(element, compiler, **kw):
 
 
 def _create_shadow_tables(migrate_engine):
-    meta = sa.MetaData(migrate_engine)
+    meta = sa.MetaData()
     meta.reflect(migrate_engine)
     table_names = list(meta.tables.keys())
 
+    # NOTE(stephenfin): This is not compatible with SQLAlchemy 2.0 but neither
+    # is sqlalchemy-migrate which requires this. We'll remove these migrations
+    # when dropping SQLAlchemy < 2.x support
     meta.bind = migrate_engine
 
     for table_name in table_names:
@@ -184,6 +187,9 @@ def _create_shadow_tables(migrate_engine):
 
 def upgrade(migrate_engine):
     meta = sa.MetaData()
+    # NOTE(stephenfin): This is not compatible with SQLAlchemy 2.0 but neither
+    # is sqlalchemy-migrate which requires this. We'll remove these migrations
+    # when dropping SQLAlchemy < 2.x support
     meta.bind = migrate_engine
 
     agent_builds = sa.Table('agent_builds', meta,

nova/db/main/migrations/versions/8f2f1571d55b_initial_version.py

@@ -53,8 +53,8 @@ def process(element, compiler, **kw):
 
 
 def _create_shadow_tables(connection):
-    meta = sa.MetaData(connection)
-    meta.reflect(connection)
+    meta = sa.MetaData()
+    meta.reflect(bind=connection)
     table_names = list(meta.tables.keys())
 
     for table_name in table_names:

nova/tests/fixtures/nova.py

@@ -847,18 +847,6 @@ def setUp(self):
             message=r'The current statement is being autocommitted .*',
             category=sqla_exc.SADeprecationWarning)
 
-        warnings.filterwarnings(
-            'ignore',
-            module='nova',
-            message=r'The MetaData.bind argument is deprecated .*',
-            category=sqla_exc.SADeprecationWarning)
-
-        warnings.filterwarnings(
-            'ignore',
-            module='nova',
-            message=r'The ``bind`` argument for schema methods .*',
-            category=sqla_exc.SADeprecationWarning)
-
         warnings.filterwarnings(
             'ignore',
             module='nova',

nova/tests/functional/db/test_archive.py

@@ -177,8 +177,8 @@ def test_archive_deleted_rows_incomplete(self):
     def _get_table_counts(self):
         engine = db.get_engine()
         conn = engine.connect()
-        meta = sa.MetaData(engine)
-        meta.reflect()
+        meta = sa.MetaData()
+        meta.reflect(bind=engine)
         shadow_tables = db._purgeable_tables(meta)
         results = {}
         for table in shadow_tables:

nova/tests/unit/db/main/test_api.py

@@ -5652,7 +5652,7 @@ class ArchiveTestCase(test.TestCase, ModelsObjectComparatorMixin):
     def setUp(self):
         super(ArchiveTestCase, self).setUp()
         self.engine = db.get_engine()
-        self.metadata = sa.MetaData(self.engine)
+        self.metadata = sa.MetaData()
         self.conn = self.engine.connect()
         self.instance_id_mappings = models.InstanceIdMapping.__table__
         self.shadow_instance_id_mappings = sqlalchemyutils.get_table(
@@ -5684,8 +5684,8 @@ def _assert_shadow_tables_empty_except(self, *exceptions):
         except for specificially named exceptions, are empty. This
         makes sure that archiving isn't moving unexpected content.
         """
-        metadata = sa.MetaData(bind=self.engine)
-        metadata.reflect()
+        metadata = sa.MetaData()
+        metadata.reflect(bind=self.engine)
         for table in metadata.tables:
             if table.startswith("shadow_") and table not in exceptions:
                 rows = self.conn.exec_driver_sql(
@@ -5698,8 +5698,8 @@ def test_shadow_tables(self):
 
         Shadow tables should have an identical schema to the main table.
         """
-        metadata = sa.MetaData(bind=self.engine)
-        metadata.reflect()
+        metadata = sa.MetaData()
+        metadata.reflect(bind=self.engine)
         for table_name in metadata.tables:
             # some tables don't have shadow tables so skip these
             if table_name in [
@@ -5942,7 +5942,9 @@ def _test_archive_deleted_rows_for_one_uuid_table(self, tablename):
         self.assertEqual(len(rows), 0)
         # Archive 2 rows
         db._archive_deleted_rows_for_table(
-            self.metadata, tablename, max_rows=2, before=None, task_log=False)
+            self.metadata, self.engine, tablename, max_rows=2, before=None,
+            task_log=False,
+        )
         # Verify we have 4 left in main
         rows = self.conn.execute(qmt).fetchall()
         self.assertEqual(len(rows), 4)
@@ -5951,7 +5953,9 @@ def _test_archive_deleted_rows_for_one_uuid_table(self, tablename):
         self.assertEqual(len(rows), 2)
         # Archive 2 more rows
         db._archive_deleted_rows_for_table(
-            self.metadata, tablename, max_rows=2, before=None, task_log=False)
+            self.metadata, self.engine, tablename, max_rows=2, before=None,
+            task_log=False,
+        )
         # Verify we have 2 left in main
         rows = self.conn.execute(qmt).fetchall()
         self.assertEqual(len(rows), 2)
@@ -5960,7 +5964,9 @@ def _test_archive_deleted_rows_for_one_uuid_table(self, tablename):
         self.assertEqual(len(rows), 4)
         # Try to archive more, but there are no deleted rows left.
         db._archive_deleted_rows_for_table(
-            self.metadata, tablename, max_rows=2, before=None, task_log=False)
+            self.metadata, self.engine, tablename, max_rows=2, before=None,
+            task_log=False,
+        )
         # Verify we still have 2 left in main
         rows = self.conn.execute(qmt).fetchall()
         self.assertEqual(len(rows), 2)
@@ -6019,8 +6025,8 @@ def test_archive_deleted_rows_for_migrations(self):
         # Archiving instances should result in migrations related to the
         # instances also being archived.
         num = db._archive_deleted_rows_for_table(
-            self.metadata, "instances", max_rows=None, before=None,
-            task_log=False)
+            self.metadata, self.engine, "instances", max_rows=None,
+            before=None, task_log=False)
         self.assertEqual(1, num[0])
         self._assert_shadow_tables_empty_except(
             'shadow_instances',
@@ -6386,7 +6392,8 @@ def call_api(*args, **kwargs):
 
 
 class TestSqlalchemyTypesRepr(
-        test_fixtures.OpportunisticDBTestMixin, test.NoDBTestCase):
+    test_fixtures.OpportunisticDBTestMixin, test.NoDBTestCase,
+):
 
     def setUp(self):
         # NOTE(sdague): the oslo_db base test case completely
@@ -6397,15 +6404,15 @@ def setUp(self):
 
         super(TestSqlalchemyTypesRepr, self).setUp()
         self.engine = enginefacade.writer.get_engine()
-        meta = sa.MetaData(bind=self.engine)
+        meta = sa.MetaData()
         self.table = sa.Table(
             'cidr_tbl',
             meta,
             sa.Column('id', sa.Integer, primary_key=True),
             sa.Column('addr', col_types.CIDR())
         )
-        self.table.create()
-        self.addCleanup(meta.drop_all)
+        meta.create_all(self.engine)
+        self.addCleanup(meta.drop_all, self.engine)
 
     def test_cidr_repr(self):
         addrs = [('192.168.3.0/24', '192.168.3.0/24'),