def test_device_delete_with_bricks(self): """Validate device deletion with existing bricks on the device""" h_node, h_url = self.heketi_client_node, self.heketi_server_url # Create volume vol_size = 1 vol_info = heketi_volume_create(h_node, h_url, vol_size, json=True) self.addCleanup(heketi_volume_delete, h_node, h_url, vol_info['id']) device_delete_id = vol_info['bricks'][0]['device'] node_id = vol_info['bricks'][0]['node'] device_info = heketi_device_info(h_node, h_url, device_delete_id, json=True) device_name = device_info['name'] # Disable the device heketi_device_disable(h_node, h_url, device_delete_id) self.addCleanup(heketi_device_enable, h_node, h_url, device_delete_id) # Delete device with bricks with self.assertRaises(AssertionError): heketi_device_delete(h_node, h_url, device_delete_id) self.addCleanup(heketi_device_add, h_node, h_url, device_name, node_id)
def test_heketi_device_delete(self): """Test Heketi device delete operation""" # Get list of additional devices for one of the Gluster nodes ip_with_devices = {} for gluster_server in g.config["gluster_servers"].values(): if not gluster_server.get("additional_devices"): continue ip_with_devices = { gluster_server['storage']: gluster_server['additional_devices'] } break # Skip test if no additional device is available if not ip_with_devices: self.skipTest( "No additional devices attached to any of the gluster nodes") # Select any additional device and get the node id of the gluster node h_node, h_server = self.heketi_client_node, self.heketi_server_url node_id, device_name = None, list(ip_with_devices.values())[0][0] topology_info = heketi_topology_info(h_node, h_server, json=True) for node in topology_info["clusters"][0]["nodes"]: if list(ip_with_devices.keys())[0] == ( node['hostnames']["storage"][0]): node_id = node["id"] break self.assertTrue(node_id) # Add additional device to the cluster heketi_device_add(h_node, h_server, device_name, node_id) # Get the device id and number of bricks on the device node_info_after_addition = heketi_node_info(h_node, h_server, node_id, json=True) device_id, bricks = None, None for device in node_info_after_addition["devices"]: if device["name"] == device_name: device_id, bricks = device["id"], len(device['bricks']) break self.assertTrue(device_id, "Device not added in expected node") # Delete heketi device heketi_device_disable(h_node, h_server, device_id) heketi_device_remove(h_node, h_server, device_id) heketi_device_delete(h_node, h_server, device_id) # Verify that there were no bricks on the newly added device msg = ( "Number of bricks on the device %s of the node %s should be zero" % (device_name, list(ip_with_devices.keys())[0])) self.assertEqual(0, bricks, msg) # Verify device deletion node_info_after_deletion = heketi_node_info(h_node, h_server, node_id) msg = ("Device %s should not be shown in node info of the node %s" "after the device deletion" % (device_id, node_id)) self.assertNotIn(device_id, node_info_after_deletion, msg)
def delete_node_and_devices_on_it(self, node_id): heketi_ops.heketi_node_disable(self.h_node, self.h_url, node_id) heketi_ops.heketi_node_remove(self.h_node, self.h_url, node_id) node_info = heketi_ops.heketi_node_info( self.h_node, self.h_url, node_id, json=True) for device in node_info['devices']: heketi_ops.heketi_device_delete( self.h_node, self.h_url, device['id']) heketi_ops.heketi_node_delete(self.h_node, self.h_url, node_id)
def detach_devices_attached(self, device_id_list): """ All the devices attached are gracefully detached in this function """ if not isinstance(device_id_list, (tuple, set, list)): device_id_list = [device_id_list] for device_id in device_id_list: device_disable = heketi_ops.heketi_device_disable( self.heketi_client_node, self.heketi_server_url, device_id) self.assertNotEqual(device_disable, False, "Device %s could not be disabled" % device_id) device_remove = heketi_ops.heketi_device_remove( self.heketi_client_node, self.heketi_server_url, device_id) self.assertNotEqual(device_remove, False, "Device %s could not be removed" % device_id) device_delete = heketi_ops.heketi_device_delete( self.heketi_client_node, self.heketi_server_url, device_id) self.assertNotEqual(device_delete, False, "Device %s could not be deleted" % device_id)
def detach_devices_attached(self, device_id_list): """ All the devices attached are gracefully detached in this function """ if not isinstance(device_id_list, (tuple, set, list)): device_id_list = [device_id_list] for device_id in device_id_list: device_disable = heketi_ops.heketi_device_disable( self.heketi_client_node, self.heketi_server_url, device_id) self.assertNotEqual( device_disable, False, "Device %s could not be disabled" % device_id) device_remove = heketi_ops.heketi_device_remove( self.heketi_client_node, self.heketi_server_url, device_id) self.assertNotEqual( device_remove, False, "Device %s could not be removed" % device_id) device_delete = heketi_ops.heketi_device_delete( self.heketi_client_node, self.heketi_server_url, device_id) self.assertNotEqual( device_delete, False, "Device %s could not be deleted" % device_id)
def test_heketi_device_remove(self, delete_device): """Validate remove/delete device using heketi-cli""" gluster_server_0 = list(g.config["gluster_servers"].values())[0] try: device_name = gluster_server_0["additional_devices"][0] except (KeyError, IndexError): self.skipTest( "Additional disk is not specified for node with following " "hostnames and IP addresses: %s, %s." % (gluster_server_0.get( 'manage', '?'), gluster_server_0.get('storage', '?'))) manage_hostname = gluster_server_0["manage"] # Get node ID of the Gluster hostname topo_info = heketi_topology_info(self.heketi_client_node, self.heketi_server_url, json=True) self.assertTrue(topo_info["clusters"][0]["nodes"], "Cluster info command returned empty list of nodes.") node_id = None for node in topo_info["clusters"][0]["nodes"]: if manage_hostname == node['hostnames']["manage"][0]: node_id = node["id"] break self.assertNotEqual( node_id, None, "No information about node_id for %s" % manage_hostname) # Iterate chosen node devices and pick the smallest online one. lowest_device_size = lowest_device_id = None online_hosts = self.get_online_nodes_disable_redundant() for host in online_hosts[0:3]: if node_id != host["id"]: continue for device in host["devices"]: if device["state"].strip().lower() != "online": continue if (lowest_device_size is None or device["storage"]["total"] < lowest_device_size): lowest_device_size = device["storage"]["total"] lowest_device_id = device["id"] lowest_device_name = device["name"] if lowest_device_id is None: self.skipTest( "Didn't find suitable device for disablement on '%s' node." % (node_id)) # Create volume vol_size = 1 vol_info = heketi_volume_create(self.heketi_client_node, self.heketi_server_url, vol_size, json=True) self.addCleanup(heketi_volume_delete, self.heketi_client_node, self.heketi_server_url, vol_info['id']) # Add extra device, then remember it's ID and size heketi_device_add(self.heketi_client_node, self.heketi_server_url, device_name, node_id) node_info_after_addition = heketi_node_info(self.heketi_client_node, self.heketi_server_url, node_id, json=True) for device in node_info_after_addition["devices"]: if device["name"] != device_name: continue device_id_new = device["id"] device_size_new = device["storage"]["total"] self.addCleanup(heketi_device_delete, self.heketi_client_node, self.heketi_server_url, device_id_new) self.addCleanup(heketi_device_remove, self.heketi_client_node, self.heketi_server_url, device_id_new) self.addCleanup(heketi_device_disable, self.heketi_client_node, self.heketi_server_url, device_id_new) if lowest_device_size > device_size_new: skip_msg = ("Skip test case, because newly added disk %s is " "smaller than device which we want to remove %s." % (device_size_new, lowest_device_size)) self.skipTest(skip_msg) g.log.info("Removing device id %s" % lowest_device_id) with self.assertRaises(AssertionError): out = heketi_device_remove(self.heketi_client_node, self.heketi_server_url, lowest_device_id) self.addCleanup(heketi_device_enable, self.heketi_client_node, self.heketi_server_url, lowest_device_id) self.addCleanup(heketi_device_disable, self.heketi_client_node, self.heketi_server_url, lowest_device_id) self.assertFalse(True, "Device removal didn't fail: %s" % out) g.log.info("Device removal failed as expected") # Need to disable device before removing heketi_device_disable(self.heketi_client_node, self.heketi_server_url, lowest_device_id) if not delete_device: self.addCleanup(heketi_device_enable, self.heketi_client_node, self.heketi_server_url, lowest_device_id) # Remove device from Heketi try: heketi_device_remove(self.heketi_client_node, self.heketi_server_url, lowest_device_id) except Exception: if delete_device: self.addCleanup(heketi_device_enable, self.heketi_client_node, self.heketi_server_url, lowest_device_id) raise if not delete_device: self.addCleanup(heketi_device_disable, self.heketi_client_node, self.heketi_server_url, lowest_device_id) if delete_device: try: heketi_device_delete(self.heketi_client_node, self.heketi_server_url, lowest_device_id) except Exception: self.addCleanup(heketi_device_enable, self.heketi_client_node, self.heketi_server_url, lowest_device_id) self.addCleanup(heketi_device_disable, self.heketi_client_node, self.heketi_server_url, lowest_device_id) raise self.addCleanup(heketi_device_add, self.heketi_client_node, self.heketi_server_url, lowest_device_name, node_id) # Create volume vol_info = heketi_volume_create(self.heketi_client_node, self.heketi_server_url, vol_size, json=True) self.addCleanup(heketi_volume_delete, self.heketi_client_node, self.heketi_server_url, vol_info['id']) if delete_device: return # Check that none of volume's bricks is present on the device present = self.check_any_of_bricks_present_in_device( vol_info['bricks'], lowest_device_id) self.assertFalse( present, "Some of the '%s' volume bricks is present of the removed " "'%s' device." % (vol_info['id'], lowest_device_id))
def test_heketi_metrics_validation_after_node(self, condition): """Validate heketi metrics after adding and remove node""" # Get additional node additional_host_info = g.config.get("additional_gluster_servers") if not additional_host_info: self.skipTest( "Skipping this test case as additional gluster server is " "not provied in config file") additional_host_info = list(additional_host_info.values())[0] storage_hostname = additional_host_info.get("manage") storage_ip = additional_host_info.get("storage") if not (storage_hostname and storage_ip): self.skipTest( "Config options 'additional_gluster_servers.manage' " "and 'additional_gluster_servers.storage' must be set.") h_client, h_server = self.heketi_client_node, self.heketi_server_url initial_node_count, final_node_count = 0, 0 # Get initial node count from prometheus metrics metric_result = self._fetch_metric_from_promtheus_pod( metric='heketi_nodes_count') initial_node_count = reduce( lambda x, y: x + y, [result.get('value')[1] for result in metric_result]) # Switch to storage project openshift_ops.switch_oc_project( self._master, self.storage_project_name) # Configure node before adding node self.configure_node_to_run_gluster(storage_hostname) # Get cluster list cluster_info = heketi_ops.heketi_cluster_list( h_client, h_server, json=True) # Add node to the cluster heketi_node_info = heketi_ops.heketi_node_add( h_client, h_server, len(self.gluster_servers), cluster_info.get('clusters')[0], storage_hostname, storage_ip, json=True) heketi_node_id = heketi_node_info.get("id") self.addCleanup( heketi_ops.heketi_node_delete, h_client, h_server, heketi_node_id, raise_on_error=False) self.addCleanup( heketi_ops.heketi_node_remove, h_client, h_server, heketi_node_id, raise_on_error=False) self.addCleanup( heketi_ops.heketi_node_disable, h_client, h_server, heketi_node_id, raise_on_error=False) self.addCleanup( openshift_ops.switch_oc_project, self._master, self.storage_project_name) if condition == 'delete': # Switch to openshift-monitoring project openshift_ops.switch_oc_project( self.ocp_master_node[0], self._prometheus_project_name) # Get initial node count from prometheus metrics for w in waiter.Waiter(timeout=60, interval=10): metric_result = self._fetch_metric_from_promtheus_pod( metric='heketi_nodes_count') node_count = reduce( lambda x, y: x + y, [result.get('value')[1] for result in metric_result]) if node_count != initial_node_count: break if w.expired: raise exceptions.ExecutionError( "Failed to get updated node details from prometheus") # Remove node from cluster heketi_ops.heketi_node_disable(h_client, h_server, heketi_node_id) heketi_ops.heketi_node_remove(h_client, h_server, heketi_node_id) for device in heketi_node_info.get('devices'): heketi_ops.heketi_device_delete( h_client, h_server, device.get('id')) heketi_ops.heketi_node_delete(h_client, h_server, heketi_node_id) # Switch to openshift-monitoring project openshift_ops.switch_oc_project( self.ocp_master_node[0], self._prometheus_project_name) # Get final node count from prometheus metrics for w in waiter.Waiter(timeout=60, interval=10): metric_result = self._fetch_metric_from_promtheus_pod( metric='heketi_nodes_count') final_node_count = reduce( lambda x, y: x + y, [result.get('value')[1] for result in metric_result]) if condition == 'delete': if final_node_count < node_count: break else: if final_node_count > initial_node_count: break if w.expired: raise exceptions.ExecutionError( "Failed to update node details in prometheus")
def test_heketi_prometheus_usedbytes_brickcount_on_device_delete( self, operation): """Validate used bytes,device count on heketi and prometheus""" h_node, h_server = self.heketi_client_node, self.heketi_server_url # Get list of additional devices for one of the Gluster nodes gluster_server_0 = list(self.gluster_servers_info.values())[0] manage_hostname = gluster_server_0.get("manage") self.assertTrue( manage_hostname, "IP Address is not specified for " "node {}".format(gluster_server_0)) device_name = gluster_server_0.get("additional_devices")[0] self.assertTrue( device_name, "Additional devices are not specified for " "node {}".format(gluster_server_0)) # Get node ID of the Gluster hostname node_list = heketi_ops.heketi_topology_info( h_node, h_server, json=True).get("clusters")[0].get("nodes") self.assertTrue( node_list, "Cluster info command returned empty list of nodes") node_id = [ node.get("id") for node in node_list if manage_hostname == node.get("hostnames").get("manage")[0]] self.assertTrue( node_id, "Failed to get node_id for {}".format(manage_hostname)) node_id = node_id[0] # Adding heketi device heketi_ops.heketi_device_add(h_node, h_server, device_name, node_id) node_info_after_addition = heketi_ops.heketi_node_info( h_node, h_server, node_id, json=True) device_id, bricks = None, None for device in node_info_after_addition.get("devices"): if device.get("name") == device_name: device_id, bricks = ( device.get("id"), len(device.get("bricks"))) break # Verify zero bricks on the device msg = ( "Number of bricks on the device {} of the nodes should be" "zero".format(device_name)) self.assertFalse(bricks, msg) self.addCleanup( heketi_ops.heketi_device_delete, h_node, h_server, device_id, raise_on_error=False) self.addCleanup( heketi_ops.heketi_device_remove, h_node, h_server, device_id, raise_on_error=False) self.addCleanup( heketi_ops.heketi_device_disable, h_node, h_server, device_id, raise_on_error=False) # Disable,Remove and Delete heketi device heketi_ops.heketi_device_disable(h_node, h_server, device_id) heketi_ops.heketi_device_remove(h_node, h_server, device_id) heketi_ops.heketi_device_delete(h_node, h_server, device_id) # Verify device deletion node_info_after_deletion = ( heketi_ops.heketi_node_info(h_node, h_server, node_id)) msg = ("Device {} should not be shown in node info of the node {}" "after the device deletion".format(device_id, node_id)) self.assertNotIn(device_id, node_info_after_deletion, msg) if operation == "usedbytes": # Validate heketi and prometheus device used bytes for w in waiter.Waiter(timeout=60, interval=10): device_used_bytes_prometheus = 0 device_used_bytes_metrics = 0 openshift_ops.switch_oc_project( self.ocp_master_node[0], 'openshift-monitoring') metric_result = self._fetch_metric_from_promtheus_pod( metric='heketi_device_used_bytes') for result in metric_result: if (node_id == result.get('cluster') and device_name == result.get('device')): device_used_bytes_prometheus += ( int(result.get('value')[1])) openshift_ops.switch_oc_project( self.ocp_master_node[0], 'glusterfs') metrics = heketi_ops.get_heketi_metrics(h_node, h_server) heketi_device_count_metric = ( metrics.get('heketi_device_used_bytes')) for result in heketi_device_count_metric: if (node_id == result.get('cluster') and device_name == result.get('device')): device_used_bytes_metrics = int(result.get('value')) if device_used_bytes_prometheus == device_used_bytes_metrics: break if w.expired: raise exceptions.ExecutionError( "Failed to update device details in prometheus") elif operation == "brickcount": # Validate heketi and prometheus device brick count for w in waiter.Waiter(timeout=60, interval=10): device_brick_count_prometheus = 0 device_brick_count_metrics = 0 metrics = heketi_ops.get_heketi_metrics(h_node, h_server) heketi_device_count_metric = metrics.get( 'heketi_device_brick_count') for result in heketi_device_count_metric: device_brick_count_metrics += int(result.get('value')) openshift_ops.switch_oc_project( self.ocp_master_node[0], 'openshift-monitoring') metric_result = self._fetch_metric_from_promtheus_pod( metric='heketi_device_brick_count') for result in metric_result: device_brick_count_prometheus += ( int(result.get('value')[1])) if device_brick_count_prometheus == device_brick_count_metrics: break if w.expired: raise exceptions.ExecutionError( "Failed to update device details in prometheus")
def test_heketi_device_replacement_in_node(self): """Validate device replacement operation on single node""" h_client, h_server = self.heketi_client_node, self.heketi_server_url try: gluster_server_0 = list(g.config["gluster_servers"].values())[0] manage_hostname = gluster_server_0["manage"] add_device_name = gluster_server_0["additional_devices"][0] except (KeyError, IndexError): self.skipTest( "Additional disk is not specified for node with following " "hostnames and IP addresses: {}, {}".format( gluster_server_0.get('manage', '?'), gluster_server_0.get('storage', '?'))) # Get existing heketi volume list existing_volumes = heketi_volume_list(h_client, h_server, json=True) # Add cleanup function to clean stale volumes created during test self.addCleanup(self._cleanup_heketi_volumes, existing_volumes.get("volumes")) # Get nodes info node_id_list = heketi_node_list(h_client, h_server) # Disable 4th and other nodes if len(node_id_list) > 3: for node_id in node_id_list[3:]: heketi_node_disable(h_client, h_server, node_id) self.addCleanup(heketi_node_enable, h_client, h_server, node_id) # Create volume when 3 nodes are online vol_size, vol_count = 2, 4 for _ in range(vol_count): vol_info = heketi_blockvolume_create(h_client, h_server, vol_size, json=True) self.addCleanup(heketi_blockvolume_delete, h_client, h_server, vol_info['id']) # Get node ID of the Gluster hostname topology_info = heketi_topology_info(h_client, h_server, json=True) self.assertIsNotNone(topology_info, "Failed to get topology info") self.assertIn("clusters", topology_info.keys(), "Failed to get cluster " "details from topology info") node_list = topology_info["clusters"][0]["nodes"] self.assertTrue(node_list, "Cluster info command returned empty list of nodes") node_id = None for node in node_list: if manage_hostname == node['hostnames']["manage"][0]: node_id = node["id"] break self.assertTrue( node_id, "Failed to get node info for node id '{}'".format( manage_hostname)) # Add extra device, then remember it's ID and size device_id_new, device_size_new = self._add_heketi_device( add_device_name, node_id) # Remove one of the existing devices on node except new device device_name, device_id = None, None node_info_after_addition = heketi_node_info(h_client, h_server, node_id, json=True) for device in node_info_after_addition["devices"]: if (device["name"] != add_device_name and device["storage"]["total"] == device_size_new): device_name = device["name"] device_id = device["id"] break self.assertIsNotNone(device_name, "Failed to get device name") self.assertIsNotNone(device_id, "Failed to get device id") self.addCleanup(heketi_device_enable, h_client, h_server, device_id, raise_on_error=False) self.addCleanup(heketi_device_add, h_client, h_server, device_name, node_id, raise_on_error=False) heketi_device_disable(h_client, h_server, device_id) heketi_device_remove(h_client, h_server, device_id) heketi_device_delete(h_client, h_server, device_id)
def _add_new_device_and_remove_existing_device( self, is_delete_device, add_device_name, node_id, add_back_again=False, skip_cleanup_addition=False): """Delete or remove device and also add one device on the same node. """ h_client, h_url = self.heketi_client_node, self.heketi_server_url raise_on_error = False if add_back_again else True # Iterate chosen node devices and pick the smallest online one. lowest_device_size = lowest_device_id = None online_hosts = self.get_online_nodes_disable_redundant() for host in online_hosts[0:3]: if node_id != host["id"]: continue for device in host["devices"]: if device["state"].strip().lower() != "online": continue if (lowest_device_size is None or device["storage"]["total"] < lowest_device_size): lowest_device_size = device["storage"]["total"] lowest_device_id = device["id"] lowest_device_name = device["name"] if lowest_device_id is None: self.skipTest( "Didn't find suitable device for disablement on '%s' node." % (node_id)) # Create volume vol_size = 1 vol_info = heketi_volume_create(h_client, h_url, vol_size, json=True) self.addCleanup(heketi_volume_delete, self.heketi_client_node, self.heketi_server_url, vol_info['id']) # Add extra device, then remember it's ID and size device_id_new, device_size_new = self._add_heketi_device( add_device_name, node_id, raise_on_error) if lowest_device_size > device_size_new: skip_msg = ("Skip test case, because newly added disk %s is " "smaller than device which we want to remove %s." % (device_size_new, lowest_device_size)) self.skipTest(skip_msg) g.log.info("Removing device id %s" % lowest_device_id) with self.assertRaises(AssertionError): out = heketi_device_remove(h_client, h_url, lowest_device_id) self.addCleanup(heketi_device_enable, h_client, h_url, lowest_device_id) self.addCleanup(heketi_device_disable, h_client, h_url, lowest_device_id) self.assertFalse(True, "Device removal didn't fail: %s" % out) g.log.info("Device removal failed as expected") # Need to disable device before removing heketi_device_disable(h_client, h_url, lowest_device_id) if not is_delete_device: self.addCleanup(heketi_device_enable, h_client, h_url, lowest_device_id) # Remove device from Heketi try: heketi_device_remove(h_client, h_url, lowest_device_id) except Exception: if is_delete_device: self.addCleanup(heketi_device_enable, h_client, h_url, lowest_device_id, raise_on_error=raise_on_error) raise if not is_delete_device: self.addCleanup(heketi_device_disable, h_client, h_url, lowest_device_id) if is_delete_device: try: heketi_device_delete(h_client, h_url, lowest_device_id) except Exception: self.addCleanup(heketi_device_enable, h_client, h_url, lowest_device_id, raise_on_error=raise_on_error) self.addCleanup(heketi_device_disable, h_client, h_url, lowest_device_id, raise_on_error=raise_on_error) raise if not skip_cleanup_addition: # Do not add the additional device back, intially added self.addCleanup(heketi_device_add, h_client, h_url, lowest_device_name, node_id, raise_on_error=raise_on_error) # Create volume vol_info = heketi_volume_create(h_client, h_url, vol_size, json=True) self.addCleanup(heketi_volume_delete, h_client, h_url, vol_info['id']) if is_delete_device: return lowest_device_name # Check that none of volume's bricks is present on the device present = self.check_any_of_bricks_present_in_device( vol_info['bricks'], lowest_device_id) self.assertFalse( present, "Some of the '%s' volume bricks is present of the removed " "'%s' device." % (vol_info['id'], lowest_device_id))
def test_dev_path_mapping_heketi_node_delete(self): """Validate dev path mapping for heketi node deletion lifecycle""" h_client, h_url = self.heketi_client_node, self.heketi_server_url node_ids = heketi_ops.heketi_node_list(h_client, h_url) self.assertTrue(node_ids, "Failed to get heketi node list") # Fetch #4th node for the operations h_disable_node = node_ids[3] # Fetch bricks on the devices before volume create h_node_details_before, h_node = self._get_bricks_and_device_details() # Bricks count on the node before pvc creation brick_count_before = [count[1] for count in h_node_details_before] # Create file volume with app pod and verify IO's # and compare path, UUID, vg_name pod_name, dc_name, use_percent = self._create_app_pod_and_verify_pvs() # Check if IO's are running use_percent_after = self._get_space_use_percent_in_app_pod(pod_name) self.assertNotEqual( use_percent, use_percent_after, "Failed to execute IO's in the app pod {} after respin".format( pod_name)) # Fetch bricks on the devices after volume create h_node_details_after, h_node = self._get_bricks_and_device_details() # Bricks count on the node after pvc creation brick_count_after = [count[1] for count in h_node_details_after] self.assertGreater( sum(brick_count_after), sum(brick_count_before), "Failed to add bricks on the node {}".format(h_node)) self.addCleanup(heketi_ops.heketi_node_disable, h_client, h_url, h_disable_node) # Enable the #4th node heketi_ops.heketi_node_enable(h_client, h_url, h_disable_node) node_info = heketi_ops.heketi_node_info(h_client, h_url, h_disable_node, json=True) h_node_id = node_info['id'] self.assertEqual(node_info['state'], "online", "Failed to enable node {}".format(h_disable_node)) # Disable the node and check for brick migrations self.addCleanup(heketi_ops.heketi_node_enable, h_client, h_url, h_node, raise_on_error=False) heketi_ops.heketi_node_disable(h_client, h_url, h_node) node_info = heketi_ops.heketi_node_info(h_client, h_url, h_node, json=True) self.assertEqual(node_info['state'], "offline", "Failed to disable node {}".format(h_node)) # Before bricks migration h_node_info = heketi_ops.heketi_node_info(h_client, h_url, h_node, json=True) # Bricks before migration on the node i.e to be deleted bricks_counts_before = 0 for device in h_node_info['devices']: bricks_counts_before += (len(device['bricks'])) # Remove the node heketi_ops.heketi_node_remove(h_client, h_url, h_node) # After bricks migration h_node_info_after = heketi_ops.heketi_node_info(h_client, h_url, h_node, json=True) # Bricks after migration on the node i.e to be delete bricks_counts = 0 for device in h_node_info_after['devices']: bricks_counts += (len(device['bricks'])) self.assertFalse( bricks_counts, "Failed to remove all the bricks from node {}".format(h_node)) # Old node which is to deleted, new node were bricks resides old_node, new_node = h_node, h_node_id # Node info for the new node were brick reside after migration h_node_info_new = heketi_ops.heketi_node_info(h_client, h_url, new_node, json=True) bricks_counts_after = 0 for device in h_node_info_new['devices']: bricks_counts_after += (len(device['bricks'])) self.assertEqual( bricks_counts_before, bricks_counts_after, "Failed to migrated bricks from {} node to {}".format( old_node, new_node)) # Fetch device list i.e to be deleted h_node_info = heketi_ops.heketi_node_info(h_client, h_url, h_node, json=True) devices_list = [[device['id'], device['name']] for device in h_node_info['devices']] for device in devices_list: device_id = device[0] device_name = device[1] self.addCleanup(heketi_ops.heketi_device_add, h_client, h_url, device_name, h_node, raise_on_error=False) # Device deletion from heketi node device_delete = heketi_ops.heketi_device_delete( h_client, h_url, device_id) self.assertTrue(device_delete, "Failed to delete the device {}".format(device_id)) node_info = heketi_ops.heketi_node_info(h_client, h_url, h_node, json=True) cluster_id = node_info['cluster'] zone = node_info['zone'] storage_hostname = node_info['hostnames']['manage'][0] storage_ip = node_info['hostnames']['storage'][0] # Delete the node self.addCleanup(heketi_ops.heketi_node_add, h_client, h_url, zone, cluster_id, storage_hostname, storage_ip, raise_on_error=False) heketi_ops.heketi_node_delete(h_client, h_url, h_node) # Verify if the node is deleted node_ids = heketi_ops.heketi_node_list(h_client, h_url) self.assertNotIn(old_node, node_ids, "Failed to delete the node {}".format(old_node)) # Check if IO's are running use_percent_after = self._get_space_use_percent_in_app_pod(pod_name) self.assertNotEqual( use_percent, use_percent_after, "Failed to execute IO's in the app pod {} after respin".format( pod_name)) # Adding node back h_node_info = heketi_ops.heketi_node_add(h_client, h_url, zone, cluster_id, storage_hostname, storage_ip, json=True) self.assertTrue( h_node_info, "Failed to add the node in the cluster {}".format(cluster_id)) h_node_id = h_node_info["id"] # Adding devices to the new node for device in devices_list: storage_device = device[1] # Add device to the new heketi node heketi_ops.heketi_device_add(h_client, h_url, storage_device, h_node_id) heketi_node_info = heketi_ops.heketi_node_info(h_client, h_url, h_node_id, json=True) device_id = None for device in heketi_node_info["devices"]: if device["name"] == storage_device: device_id = device["id"] break self.assertTrue( device_id, "Failed to add device {} on node {}".format( storage_device, h_node_id)) # Create n pvc in order to verfiy if the bricks reside on the new node pvc_amount, pvc_size = 5, 1 # Fetch bricks on the devices before volume create h_node_details_before, h_node = self._get_bricks_and_device_details() # Bricks count on the node before pvc creation brick_count_before = [count[1] for count in h_node_details_before] # Create file volumes pvc_name = self.create_and_wait_for_pvcs(pvc_size=pvc_size, pvc_amount=pvc_amount) self.assertEqual(len(pvc_name), pvc_amount, "Failed to create {} pvc".format(pvc_amount)) # Fetch bricks on the devices before volume create h_node_details_after, h_node = self._get_bricks_and_device_details() # Bricks count on the node after pvc creation brick_count_after = [count[1] for count in h_node_details_after] self.assertGreater( sum(brick_count_after), sum(brick_count_before), "Failed to add bricks on the new node {}".format(new_node)) # Check if IO's are running after new node is added use_percent_after = self._get_space_use_percent_in_app_pod(pod_name) self.assertNotEqual( use_percent, use_percent_after, "Failed to execute IO's in the app pod {} after respin".format( pod_name))
def test_dev_path_mapping_heketi_device_delete(self): """Validate dev path mapping for heketi device delete lifecycle""" h_client, h_url = self.heketi_client_node, self.heketi_server_url node_ids = heketi_ops.heketi_node_list(h_client, h_url) self.assertTrue(node_ids, "Failed to get heketi node list") # Fetch #4th node for the operations h_disable_node = node_ids[3] # Fetch bricks on the devices before volume create h_node_details_before, h_node = self._get_bricks_and_device_details() # Bricks count on the node before pvc creation brick_count_before = [count[1] for count in h_node_details_before] # Create file volume with app pod and verify IO's # and compare path, UUID, vg_name pod_name, dc_name, use_percent = self._create_app_pod_and_verify_pvs() # Check if IO's are running use_percent_after = self._get_space_use_percent_in_app_pod(pod_name) self.assertNotEqual( use_percent, use_percent_after, "Failed to execute IO's in the app pod {} after respin".format( pod_name)) # Fetch bricks on the devices after volume create h_node_details_after, h_node = self._get_bricks_and_device_details() # Bricks count on the node after pvc creation brick_count_after = [count[1] for count in h_node_details_after] self.assertGreater( sum(brick_count_after), sum(brick_count_before), "Failed to add bricks on the node {}".format(h_node)) # Enable the #4th node heketi_ops.heketi_node_enable(h_client, h_url, h_disable_node) node_info = heketi_ops.heketi_node_info(h_client, h_url, h_disable_node, json=True) h_node_id = node_info['id'] self.assertEqual(node_info['state'], "online", "Failed to enable node {}".format(h_disable_node)) # Fetch device list i.e to be deleted h_node_info = heketi_ops.heketi_node_info(h_client, h_url, h_node, json=True) devices_list = [[device['id'], device['name']] for device in h_node_info['devices']] # Device deletion operation for device in devices_list: device_id, device_name = device[0], device[1] self.addCleanup(heketi_ops.heketi_device_enable, h_client, h_url, device_id, raise_on_error=False) # Disable device from heketi device_disable = heketi_ops.heketi_device_disable( h_client, h_url, device_id) self.assertTrue( device_disable, "Device {} could not be disabled".format(device_id)) device_info = heketi_ops.heketi_device_info(h_client, h_url, device_id, json=True) self.assertEqual(device_info['state'], "offline", "Failed to disable device {}".format(device_id)) # Remove device from heketi device_remove = heketi_ops.heketi_device_remove( h_client, h_url, device_id) self.assertTrue(device_remove, "Device {} could not be removed".format(device_id)) # Bricks after device removal device_info = heketi_ops.heketi_device_info(h_client, h_url, device_id, json=True) bricks_count_after = len(device_info['bricks']) self.assertFalse( bricks_count_after, "Failed to remove the bricks from the device {}".format( device_id)) # Delete device from heketi self.addCleanup(heketi_ops.heketi_device_add, h_client, h_url, device_name, h_node, raise_on_error=False) device_delete = heketi_ops.heketi_device_delete( h_client, h_url, device_id) self.assertTrue(device_delete, "Device {} could not be deleted".format(device_id)) # Check if IO's are running after device is deleted use_percent_after = self._get_space_use_percent_in_app_pod(pod_name) self.assertNotEqual( use_percent, use_percent_after, "Failed to execute IO's in the app pod {} after respin".format( pod_name)) # Add device operations for device in devices_list: device_name = device[1] # Add device back to the node heketi_ops.heketi_device_add(h_client, h_url, device_name, h_node) # Fetch device info after device add node_info = heketi_ops.heketi_node_info(h_client, h_url, h_node, json=True) device_id = None for device in node_info["devices"]: if device["name"] == device_name: device_id = device["id"] break self.assertTrue( device_id, "Failed to add device {} on node" " {}".format(device_name, h_node)) # Disable the #4th node heketi_ops.heketi_node_disable(h_client, h_url, h_node_id) node_info = heketi_ops.heketi_node_info(h_client, h_url, h_node_id, json=True) self.assertEqual(node_info['state'], "offline", "Failed to disable node {}".format(h_node_id)) pvc_amount, pvc_size = 5, 1 # Fetch bricks on the devices before volume create h_node_details_before, h_node = self._get_bricks_and_device_details() # Bricks count on the node before pvc creation brick_count_before = [count[1] for count in h_node_details_before] # Create file volumes pvc_name = self.create_and_wait_for_pvcs(pvc_size=pvc_size, pvc_amount=pvc_amount) self.assertEqual(len(pvc_name), pvc_amount, "Failed to create {} pvc".format(pvc_amount)) # Fetch bricks on the devices after volume create h_node_details_after, h_node = self._get_bricks_and_device_details() # Bricks count on the node after pvc creation brick_count_after = [count[1] for count in h_node_details_after] self.assertGreater( sum(brick_count_after), sum(brick_count_before), "Failed to add bricks on the node {}".format(h_node)) # Check if IO's are running after new device is added use_percent_after = self._get_space_use_percent_in_app_pod(pod_name) self.assertNotEqual( use_percent, use_percent_after, "Failed to execute IO's in the app pod {} after respin".format( pod_name))