NetApp ONTAP: Added support to Active/Active mode in NFS driver

This patch enables Active/Active support in NetApp NFS driver.

To support replication in A/A mode, the method failover_host was
splitted in two phases (failover and failover_completed) as
required by the spec [1].

[1] https://specs.openstack.org/openstack/cinder-specs/specs/ocata/ha-aa-replication.html

Change-Id: I401ca89440d44e04f460741981cf24a42e5264a0

cinder/tests/unit/volume/drivers/netapp/dataontap/test_nfs_cmode.py

@@ -1573,6 +1573,72 @@ class NetAppCmodeNfsDriverTestCase(test.TestCase):
         self.assertEqual('dev1', actual_active)
         self.assertEqual([], vol_updates)
 
+    @ddt.data({'secondary_id': 'dev0', 'configured_targets': ['dev1']},
+              {'secondary_id': 'dev3', 'configured_targets': ['dev1', 'dev2']},
+              {'secondary_id': 'dev1', 'configured_targets': []},
+              {'secondary_id': None, 'configured_targets': []})
+    @ddt.unpack
+    def test_failover_invalid_replication_target(self, secondary_id,
+                                                 configured_targets):
+        """This tests executes a method in the DataMotionMixin."""
+        self.driver.backend_name = 'dev0'
+        self.mock_object(data_motion.DataMotionMixin,
+                         'get_replication_backend_names',
+                         return_value=configured_targets)
+        complete_failover_call = self.mock_object(
+            data_motion.DataMotionMixin, '_complete_failover')
+
+        self.assertRaises(exception.InvalidReplicationTarget,
+                          self.driver.failover, 'fake_context', [],
+                          secondary_id=secondary_id)
+        self.assertFalse(complete_failover_call.called)
+
+    def test_failover_unable_to_failover(self):
+        """This tests executes a method in the DataMotionMixin."""
+        self.driver.backend_name = 'dev0'
+        self.mock_object(data_motion.DataMotionMixin, '_complete_failover',
+                         side_effect=na_utils.NetAppDriverException)
+        self.mock_object(data_motion.DataMotionMixin,
+                         'get_replication_backend_names',
+                         return_value=['dev1', 'dev2'])
+        self.mock_object(self.driver.ssc_library, 'get_ssc_flexvol_names',
+                         return_value=fake_ssc.SSC.keys())
+        self.mock_object(self.driver, '_update_zapi_client')
+
+        self.assertRaises(exception.UnableToFailOver,
+                          self.driver.failover, 'fake_context', [],
+                          secondary_id='dev1')
+        data_motion.DataMotionMixin._complete_failover.assert_called_once_with(
+            'dev0', ['dev1', 'dev2'], fake_ssc.SSC.keys(), [],
+            failover_target='dev1')
+        self.assertFalse(self.driver._update_zapi_client.called)
+
+    def test_failover(self):
+        """This tests executes a method in the DataMotionMixin."""
+        self.driver.backend_name = 'dev0'
+        self.mock_object(data_motion.DataMotionMixin, '_complete_failover',
+                         return_value=('dev1', []))
+        self.mock_object(data_motion.DataMotionMixin,
+                         'get_replication_backend_names',
+                         return_value=['dev1', 'dev2'])
+        self.mock_object(self.driver.ssc_library, 'get_ssc_flexvol_names',
+                         return_value=fake_ssc.SSC.keys())
+        self.mock_object(self.driver, '_update_zapi_client')
+
+        actual_active, vol_updates, __ = self.driver.failover(
+            'fake_context', [], secondary_id='dev1', groups=[])
+
+        data_motion.DataMotionMixin._complete_failover.assert_called_once_with(
+            'dev0', ['dev1', 'dev2'], fake_ssc.SSC.keys(), [],
+            failover_target='dev1')
+
+    def test_failover_completed(self):
+        self.mock_object(self.driver, '_update_zapi_client')
+        self.driver.failover_completed('fake_context', secondary_id='dev1')
+        self.driver._update_zapi_client.assert_called_once_with('dev1')
+        self.assertTrue(self.driver.failed_over)
+        self.assertEqual('dev1', self.driver.failed_over_backend_name)
+
     def test_delete_group_snapshot(self):
         mock_delete_backing_file = self.mock_object(
             self.driver, '_delete_backing_file_for_snapshot')

cinder/tests/unit/volume/drivers/netapp/dataontap/utils/test_data_motion.py

@@ -1027,7 +1027,7 @@ class NetAppCDOTDataMotionMixinTestCase(test.TestCase):
         self.assertEqual('fallback2', target)
         self.assertFalse(mock_debug_log.called)
 
-    def test__failover_host_no_suitable_target(self):
+    def test__complete_failover_no_suitable_target(self):
         flexvols = ['nvol1', 'nvol2']
         replication_backends = ['fallback1', 'fallback2']
         self.mock_object(self.dm_mixin, '_choose_failover_target',
@@ -1045,7 +1045,7 @@ class NetAppCDOTDataMotionMixinTestCase(test.TestCase):
         self.assertFalse(self.dm_mixin.break_snapmirrors.called)
 
     @ddt.data('fallback1', None)
-    def test__failover_host(self, failover_target):
+    def test__complete_failover(self, failover_target):
         flexvols = ['nvol1', 'nvol2', 'nvol3']
         replication_backends = ['fallback1', 'fallback2']
         volumes = [

cinder/volume/drivers/netapp/dataontap/nfs_cmode.py

@@ -66,12 +66,15 @@ class NetAppCmodeNfsDriver(nfs_base.NetAppNfsDriver,
                 Implement FlexGroup pool
         3.0.0 - Add support for Intra-cluster Storage assisted volume migration
                 Add support for revert to snapshot
+        4.0.0 - Add Cinder Active/Active support (High Availability)
+                Implement Active/Active replication support
 
     """
 
-    VERSION = "3.0.0"
+    VERSION = "4.0.0"
 
     REQUIRED_CMODE_FLAGS = ['netapp_vserver']
+    SUPPORTS_ACTIVE_ACTIVE = True
 
     def __init__(self, *args, **kwargs):
         super(NetAppCmodeNfsDriver, self).__init__(*args, **kwargs)
@@ -867,9 +870,23 @@ class NetAppCmodeNfsDriver(nfs_base.NetAppNfsDriver,
         super(NetAppCmodeNfsDriver, self).unmanage(volume)
 
     def failover_host(self, context, volumes, secondary_id=None, groups=None):
-        """Failover a backend to a secondary replication target."""
+        """Failover a backend to a secondary replication target.
 
-        return self._failover_host(volumes, secondary_id=secondary_id)
+           This function combines failover() and failover_completed()
+           to perform failover when Active/Active is not enabled.
+        """
+        active_backend_name, volume_updates, group_updates = (
+            self._failover(context, volumes, secondary_id, groups))
+        self._failover_completed(context, active_backend_name)
+        return active_backend_name, volume_updates, group_updates
+
+    def failover(self, context, volumes, secondary_id=None, groups=None):
+        """Failover to replication target."""
+        return self._failover(context, volumes, secondary_id, groups)
+
+    def failover_completed(self, context, secondary_id=None):
+        """Update volume node when `failover` is completed."""
+        return self._failover_completed(context, secondary_id)
 
     def _get_backing_flexvol_names(self):
         """Returns a list of backing flexvol names."""

cinder/volume/drivers/netapp/dataontap/utils/data_motion.py

@@ -787,6 +787,48 @@ class DataMotionMixin(object):
 
         return active_backend_name, volume_updates, []
 
+    def _failover(self, context, volumes, secondary_id=None, groups=None):
+        """Failover to replication target."""
+        if secondary_id == self.backend_name:
+            msg = _("Cannot failover to the same host as the primary.")
+            raise exception.InvalidReplicationTarget(reason=msg)
+
+        replication_targets = self.get_replication_backend_names(
+            self.configuration)
+
+        if not replication_targets:
+            msg = _("No replication targets configured for backend "
+                    "%s. Cannot failover.")
+            raise exception.InvalidReplicationTarget(reason=msg % self.host)
+        elif secondary_id and secondary_id not in replication_targets:
+            msg = _("%(target)s is not among replication targets configured "
+                    "for back end %(host)s. Cannot failover.")
+            payload = {
+                'target': secondary_id,
+                'host': self.host,
+            }
+            raise exception.InvalidReplicationTarget(reason=msg % payload)
+
+        flexvols = self.ssc_library.get_ssc_flexvol_names()
+
+        try:
+            active_backend_name, volume_updates = self._complete_failover(
+                self.backend_name, replication_targets, flexvols, volumes,
+                failover_target=secondary_id)
+        except na_utils.NetAppDriverException as e:
+            msg = _("Could not complete failover: %s") % e
+            raise exception.UnableToFailOver(reason=msg)
+
+        return active_backend_name, volume_updates, []
+
+    def _failover_completed(self, context, secondary_id=None):
+        """Update volume node when `failover` is completed."""
+        # Update the ZAPI client to the backend we failed over to
+        self._update_zapi_client(secondary_id)
+
+        self.failed_over = True
+        self.failed_over_backend_name = secondary_id
+
     def _get_replication_volume_online_timeout(self):
         return self.configuration.netapp_replication_volume_online_timeout
 

doc/source/reference/support-matrix.ini

@@ -157,7 +157,10 @@ title=NEC Storage M Series Driver (iSCSI, FC)
 title=NEC Storage V Series Driver (iSCSI, FC)
 
 [driver.netapp_ontap]
-title=NetApp Data ONTAP Driver (iSCSI, NFS, FC, NVMe/TCP)
+title=NetApp Data ONTAP Driver (iSCSI, FC, NVMe/TCP)
+
+[driver.netapp_ontap_nfs]
+title=NetApp Data ONTAP Driver (NFS)
 
 [driver.netapp_solidfire]
 title=NetApp Solidfire Driver (iSCSI)
@@ -286,6 +289,7 @@ driver.macrosan=complete
 driver.nec=complete
 driver.nec_v=complete
 driver.netapp_ontap=complete
+driver.netapp_ontap_nfs=complete
 driver.netapp_solidfire=complete
 driver.nexenta=complete
 driver.nfs=complete
@@ -363,6 +367,7 @@ driver.macrosan=complete
 driver.nec=complete
 driver.nec_v=complete
 driver.netapp_ontap=complete
+driver.netapp_ontap_nfs=complete
 driver.netapp_solidfire=complete
 driver.nexenta=complete
 driver.nfs=missing
@@ -443,6 +448,7 @@ driver.macrosan=complete
 driver.nec=complete
 driver.nec_v=missing
 driver.netapp_ontap=complete
+driver.netapp_ontap_nfs=complete
 driver.netapp_solidfire=complete
 driver.nexenta=missing
 driver.nfs=missing
@@ -522,6 +528,7 @@ driver.macrosan=complete
 driver.nec=missing
 driver.nec_v=missing
 driver.netapp_ontap=complete
+driver.netapp_ontap_nfs=complete
 driver.netapp_solidfire=complete
 driver.nexenta=missing
 driver.nfs=missing
@@ -602,6 +609,7 @@ driver.macrosan=missing
 driver.nec=missing
 driver.nec_v=complete
 driver.netapp_ontap=complete
+driver.netapp_ontap_nfs=complete
 driver.netapp_solidfire=complete
 driver.nexenta=missing
 driver.nfs=missing
@@ -681,6 +689,7 @@ driver.macrosan=complete
 driver.nec=complete
 driver.nec_v=complete
 driver.netapp_ontap=complete
+driver.netapp_ontap_nfs=complete
 driver.netapp_solidfire=complete
 driver.nexenta=missing
 driver.nfs=complete
@@ -761,6 +770,7 @@ driver.macrosan=complete
 driver.nec=complete
 driver.nec_v=missing
 driver.netapp_ontap=complete
+driver.netapp_ontap_nfs=complete
 driver.netapp_solidfire=complete
 driver.nexenta=missing
 driver.nfs=missing
@@ -841,6 +851,7 @@ driver.macrosan=missing
 driver.nec=complete
 driver.nec_v=complete
 driver.netapp_ontap=complete
+driver.netapp_ontap_nfs=complete
 driver.netapp_solidfire=complete
 driver.nexenta=missing
 driver.nfs=missing
@@ -918,6 +929,7 @@ driver.macrosan=missing
 driver.nec=complete
 driver.nec_v=complete
 driver.netapp_ontap=complete
+driver.netapp_ontap_nfs=complete
 driver.netapp_solidfire=complete
 driver.nexenta=missing
 driver.nfs=missing
@@ -999,6 +1011,7 @@ driver.macrosan=complete
 driver.nec=missing
 driver.nec_v=missing
 driver.netapp_ontap=missing
+driver.netapp_ontap_nfs=complete
 driver.netapp_solidfire=complete
 driver.nexenta=missing
 driver.nfs=missing

releasenotes/notes/netapp-nfs-aa-support-477ddf585c5aa578.yaml

@@ -0,0 +1,5 @@
+---
+features:
+  - |
+    NetApp ONTAP NFS driver: Enabled support for Active/Active environments
+    in the NetApp NFS driver (including replication).