openstack
/
cinder
Code Issues Proposed changes
cinder/cinder/volume/drivers/dell_emc/powermax/common.py

6387 lines
284 KiB
Python
Raw Blame History

# Copyright (c) 2017-2018 Dell Inc. or its subsidiaries.
# All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License"); you may
# not use this file except in compliance with the License. You may obtain
# a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
# License for the specific language governing permissions and limitations
# under the License.
import ast
from copy import deepcopy
import math
import random
import sys
import time
from oslo_config import cfg
from oslo_config import types
from oslo_log import log as logging
from oslo_utils import strutils
import six
from cinder import coordination
from cinder import exception
from cinder.i18n import _
from cinder.objects import fields
from cinder.utils import retry
from cinder.volume import configuration
from cinder.volume.drivers.dell_emc.powermax import masking
from cinder.volume.drivers.dell_emc.powermax import metadata as volume_metadata
from cinder.volume.drivers.dell_emc.powermax import provision
from cinder.volume.drivers.dell_emc.powermax import rest
from cinder.volume.drivers.dell_emc.powermax import utils
from cinder.volume import volume_types
from cinder.volume import volume_utils
LOG = logging.getLogger(__name__)
CONF = cfg.CONF
BACKENDNAME = 'volume_backend_name'
PREFIXBACKENDNAME = 'capabilities:volume_backend_name'
# Replication
REPLICATION_DISABLED = fields.ReplicationStatus.DISABLED
REPLICATION_ENABLED = fields.ReplicationStatus.ENABLED
REPLICATION_FAILOVER = fields.ReplicationStatus.FAILED_OVER
FAILOVER_ERROR = fields.ReplicationStatus.FAILOVER_ERROR
REPLICATION_ERROR = fields.ReplicationStatus.ERROR
retry_exc_tuple = (exception.VolumeBackendAPIException,)
powermax_opts = [
cfg.IntOpt('interval',
default=3,
help='Use this value to specify '
'length of the interval in seconds.'),
cfg.IntOpt('retries',
default=200,
help='Use this value to specify '
'number of retries.'),
cfg.IntOpt(utils.VMAX_SNAPVX_UNLINK_LIMIT,
default=3,
help='DEPRECATED: vmax_snapvc_unlink_limit.',
deprecated_for_removal=True,
deprecated_reason='Replaced by powermax_snapvx_unlink_limit.'),
cfg.BoolOpt('initiator_check',
default=False,
help='Use this value to enable '
'the initiator_check.'),
cfg.StrOpt(utils.VMAX_ARRAY,
help='DEPRECATED: vmax_array.',
deprecated_for_removal=True,
deprecated_reason='Replaced by powermax_array.'),
cfg.StrOpt(utils.VMAX_SRP,
help='DEPRECATED: vmax_srp.',
deprecated_for_removal=True,
deprecated_reason='Replaced by powermax_srp.'),
cfg.StrOpt(utils.VMAX_SERVICE_LEVEL,
help='DEPRECATED: vmax_service_level.',
deprecated_for_removal=True,
deprecated_reason='Replaced by powermax_service_level.'),
cfg.StrOpt(utils.VMAX_WORKLOAD,
help='Workload, setting this as an extra spec in '
'pool_name is preferable.'),
cfg.ListOpt(utils.VMAX_PORT_GROUPS,
bounds=True,
help='DEPRECATED: vmax_port_groups.',
deprecated_for_removal=True,
deprecated_reason='Replaced by powermax_port_groups.'),
cfg.IntOpt(utils.U4P_FAILOVER_TIMEOUT,
default=20.0,
help='How long to wait for the server to send data before '
'giving up.'),
cfg.IntOpt(utils.U4P_FAILOVER_RETRIES,
default=3,
help='The maximum number of retries each connection should '
'attempt. Note, this applies only to failed DNS lookups, '
'socket connections and connection timeouts, never to '
'requests where data has made it to the server.'),
cfg.IntOpt(utils.U4P_FAILOVER_BACKOFF_FACTOR,
default=1,
help='A backoff factor to apply between attempts after the '
'second try (most errors are resolved immediately by a '
'second try without a delay). Retries will sleep for: '
'{backoff factor} * (2 ^ ({number of total retries} - 1)) '
'seconds.'),
cfg.BoolOpt(utils.U4P_FAILOVER_AUTOFAILBACK,
default=True,
help='If the driver should automatically failback to the '
'primary instance of Unisphere when a successful '
'connection is re-established.'),
cfg.MultiOpt(utils.U4P_FAILOVER_TARGETS,
item_type=types.Dict(),
help='Dictionary of Unisphere failover target info.'),
cfg.IntOpt(utils.POWERMAX_SNAPVX_UNLINK_LIMIT,
default=3,
help='Use this value to specify '
'the maximum number of unlinks '
'for the temporary snapshots '
'before a clone operation.'),
cfg.StrOpt(utils.POWERMAX_ARRAY,
help='Serial number of the array to connect to.'),
cfg.StrOpt(utils.POWERMAX_SRP,
help='Storage resource pool on array to use for '
'provisioning.'),
cfg.StrOpt(utils.POWERMAX_SERVICE_LEVEL,
help='Service level to use for provisioning storage. '
'Setting this as an extra spec in pool_name '
'is preferable.'),
cfg.ListOpt(utils.POWERMAX_PORT_GROUPS,
bounds=True,
help='List of port groups containing frontend ports '
'configured prior for server connection.'),
cfg.ListOpt(utils.POWERMAX_ARRAY_TAG_LIST,
bounds=True,
help='List of user assigned name for storage array.'),
cfg.StrOpt(utils.POWERMAX_SHORT_HOST_NAME_TEMPLATE,
default='shortHostName',
help='User defined override for short host name.'),
cfg.StrOpt(utils.POWERMAX_PORT_GROUP_NAME_TEMPLATE,
default='portGroupName',
help='User defined override for port group name.')]
CONF.register_opts(powermax_opts, group=configuration.SHARED_CONF_GROUP)
class PowerMaxCommon(object):
"""Common class for Rest based PowerMax volume drivers.
This common class is for Dell EMC PowerMax volume drivers
based on UniSphere Rest API.
It supports VMAX 3 and VMAX All Flash and PowerMax arrays.
"""
pool_info = {'backend_name': None,
'config_file': None,
'arrays_info': {},
'max_over_subscription_ratio': None,
'reserved_percentage': 0,
'replication_enabled': False}
def __init__(self, prtcl, version, configuration=None,
active_backend_id=None):
self.rest = rest.PowerMaxRest()
self.utils = utils.PowerMaxUtils()
self.masking = masking.PowerMaxMasking(prtcl, self.rest)
self.provision = provision.PowerMaxProvision(self.rest)
self.volume_metadata = volume_metadata.PowerMaxVolumeMetadata(
self.rest, version, LOG.isEnabledFor(logging.DEBUG))
# Configuration/Attributes
self.protocol = prtcl
self.configuration = configuration
self.configuration.append_config_values(powermax_opts)
self.active_backend_id = active_backend_id
self.version = version
self.version_dict = {}
self.ucode_level = None
self.next_gen = False
self.replication_enabled = False
self.rep_devices = []
self.failover = True if active_backend_id else False
self.powermax_array_tag_list = None
self.powermax_short_host_name_template = None
self.powermax_port_group_name_template = None
# Gather environment info
self._get_replication_info()
self._get_u4p_failover_info()
self._gather_info()
self.rest.validate_unisphere_version()
def _gather_info(self):
"""Gather the relevant information for update_volume_stats."""
self._get_attributes_from_config()
array_info = self.get_attributes_from_cinder_config()
if array_info is None:
LOG.error("Unable to get attributes from cinder.conf. Please "
"refer to the current online documentation for correct "
"configuration and note that the xml file is no "
"longer supported.")
self.rest.set_rest_credentials(array_info)
if array_info:
serial_number = array_info['SerialNumber']
self.array_model, self.next_gen = (
self.rest.get_array_model_info(serial_number))
self.ucode_level = self.rest.get_array_ucode_version(serial_number)
if self.replication_enabled:
if serial_number in self.replication_targets:
msg = (_("The same array serial number (%s) is defined "
"for powermax_array and replication_device in "
"cinder.conf. Please ensure your "
"target_device_id points to a different "
"array." % serial_number))
LOG.error(msg)
raise exception.InvalidConfigurationValue(msg)
finalarrayinfolist = self._get_slo_workload_combinations(
array_info)
self.pool_info['arrays_info'] = finalarrayinfolist
def _get_attributes_from_config(self):
"""Get relevent details from configuration file."""
self.interval = self.configuration.safe_get('interval')
self.retries = self.configuration.safe_get('retries')
self.snapvx_unlink_limit = self._get_unlink_configuration_value(
utils.VMAX_SNAPVX_UNLINK_LIMIT,
utils.POWERMAX_SNAPVX_UNLINK_LIMIT)
self.powermax_array_tag_list = self.configuration.safe_get(
utils.POWERMAX_ARRAY_TAG_LIST)
self.powermax_short_host_name_template = self.configuration.safe_get(
utils.POWERMAX_SHORT_HOST_NAME_TEMPLATE)
self.powermax_port_group_name_template = self.configuration.safe_get(
utils.POWERMAX_PORT_GROUP_NAME_TEMPLATE)
self.pool_info['backend_name'] = (
self.configuration.safe_get('volume_backend_name'))
mosr = volume_utils.get_max_over_subscription_ratio(
self.configuration.safe_get('max_over_subscription_ratio'), True)
self.pool_info['max_over_subscription_ratio'] = mosr
self.pool_info['reserved_percentage'] = (
self.configuration.safe_get('reserved_percentage'))
LOG.debug(
"Updating volume stats on Cinder backend %(backendName)s.",
{'backendName': self.pool_info['backend_name']})
def _get_u4p_failover_info(self):
"""Gather Unisphere failover target information, if provided."""
key_dict = {'san_ip': 'RestServerIp',
'san_api_port': 'RestServerPort',
'san_login': 'RestUserName',
'san_password': 'RestPassword',
'driver_ssl_cert_verify': 'SSLVerify',
'driver_ssl_cert_path': 'SSLPath'}
if self.configuration.safe_get('u4p_failover_target'):
u4p_targets = self.configuration.safe_get('u4p_failover_target')
formatted_target_list = list()
for target in u4p_targets:
formatted_target = {key_dict[key]: value for key, value in
target.items()}
try:
formatted_target['SSLVerify'] = formatted_target['SSLPath']
del formatted_target['SSLPath']
except KeyError:
if formatted_target['SSLVerify'] == 'False':
formatted_target['SSLVerify'] = False
else:
formatted_target['SSLVerify'] = True
formatted_target_list.append(formatted_target)
u4p_failover_config = dict()
u4p_failover_config['u4p_failover_targets'] = formatted_target_list
u4p_failover_config['u4p_failover_backoff_factor'] = (
self.configuration.safe_get('u4p_failover_backoff_factor'))
u4p_failover_config['u4p_failover_retries'] = (
self.configuration.safe_get('u4p_failover_retries'))
u4p_failover_config['u4p_failover_timeout'] = (
self.configuration.safe_get('u4p_failover_timeout'))
u4p_failover_config['u4p_failover_autofailback'] = (
self.configuration.safe_get('u4p_failover_autofailback'))
u4p_failover_config['u4p_primary'] = (
self.get_attributes_from_cinder_config())
self.rest.set_u4p_failover_config(u4p_failover_config)
else:
LOG.warning("There has been no failover instances of Unisphere "
"configured for this instance of Cinder. If your "
"primary instance of Unisphere goes down then your "
"PowerMax/VMAX will be inaccessible until the "
"Unisphere REST API is responsive again.")
def retest_primary_u4p(self):
"""Retest connection to the primary instance of Unisphere."""
primary_array_info = self.get_attributes_from_cinder_config()
temp_conn = rest.PowerMaxRest()
temp_conn.set_rest_credentials(primary_array_info)
LOG.debug(
"Running connection check to primary instance of Unisphere "
"at %(primary)s", {
'primary': primary_array_info['RestServerIp']})
sc, response = temp_conn.request(target_uri='/system/version',
method='GET', u4p_check=True,
request_object=None)
if sc and int(sc) == 200:
self._get_u4p_failover_info()
self.rest.set_rest_credentials(primary_array_info)
self.rest.u4p_in_failover = False
LOG.info("Connection to primary instance of Unisphere at "
"%(primary)s restored, available failover instances of "
"Unisphere reset to default.", {
'primary': primary_array_info['RestServerIp']})
else:
LOG.debug(
"Connection check to primary instance of Unisphere at "
"%(primary)s failed, maintaining session with backup "
"instance of Unisphere at %(bu_in_use)s", {
'primary': primary_array_info['RestServerIp'],
'bu_in_use': self.rest.base_uri})
temp_conn.session.close()
def _get_initiator_check_flag(self):
"""Reads the configuration for initator_check flag.
:returns: flag
"""
return self.configuration.safe_get('initiator_check')
def _get_replication_info(self):
"""Gather replication information, if provided."""
self.rep_configs = None
self.replication_targets = []
if hasattr(self.configuration, 'replication_device'):
self.rep_devices = self.configuration.safe_get(
'replication_device')
if self.rep_devices:
if len(self.rep_devices) > 1:
self.utils.validate_multiple_rep_device(self.rep_devices)
self.rep_configs = self.utils.get_replication_config(
self.rep_devices)
# use self.replication_enabled for update_volume_stats
self.replication_enabled = True
self.replication_targets = self.utils.get_replication_targets(
self.rep_configs)
LOG.debug("The replication configuration is %(rep_configs)s.",
{'rep_configs': self.rep_configs})
if self.next_gen:
for rc in self.rep_configs:
rc[utils.RDF_CONS_EXEMPT] = True
else:
for rc in self.rep_configs:
rc[utils.RDF_CONS_EXEMPT] = False
def _get_slo_workload_combinations(self, array_info):
"""Method to query the array for SLO and Workloads.
Takes the arrayinfolist object and generates a set which has
all available SLO & Workload combinations
:param array_info: the array information
:returns: finalarrayinfolist
:raises: VolumeBackendAPIException:
"""
try:
upgraded_afa = False
if self.array_model in utils.VMAX_HYBRID_MODELS:
sls = deepcopy(utils.HYBRID_SLS)
wls = deepcopy(utils.HYBRID_WLS)
elif self.array_model in utils.VMAX_AFA_MODELS:
wls = deepcopy(utils.AFA_WLS)
if not self.next_gen:
sls = deepcopy(utils.AFA_H_SLS)
else:
sls = deepcopy(utils.AFA_P_SLS)
upgraded_afa = True
elif self.array_model in utils.PMAX_MODELS:
sls, wls = deepcopy(utils.PMAX_SLS), deepcopy(utils.PMAX_WLS)
else:
raise exception.VolumeBackendAPIException(
message="Unable to determine array model.")
if self.next_gen:
LOG.warning(
"Workloads have been deprecated for arrays running "
"PowerMax OS uCode level 5978 or higher. Any supplied "
"workloads will be treated as None values. It is "
"recommended to create a new volume type without a "
"workload specified.")
# Add service levels:
pools = sls
# Array Specific SL/WL Combos
pools += (
['{}:{}'.format(x, y) for x in sls for y in wls
if x.lower() not in ['optimized', 'none']])
# Add Optimized & None combinations
pools += (
['{}:{}'.format(x, y) for x in ['Optimized', 'NONE', 'None']
for y in ['NONE', 'None']])
if upgraded_afa:
# Cleanup is required here for service levels that were not
# present in AFA HyperMax but added for AFA PowerMax, we
# do not need these SL/WL combinations for backwards
# compatibility but we do for Diamond SL
afa_pool = list()
for p in pools:
try:
pl = p.split(':')
if (pl[0] not in [
'Platinum', 'Gold', 'Silver', 'Bronze']) or (
pl[1] not in [
'OLTP', 'OLTP_REP', 'DSS', 'DSS_REP']):
afa_pool.append(p)
except IndexError:
# Pool has no workload present
afa_pool.append(p)
pools = afa_pool
# Build array pool of SL/WL combinations
array_pool = list()
for pool in pools:
_array_info = array_info.copy()
try:
slo, workload = pool.split(':')
_array_info['SLO'] = slo
_array_info['Workload'] = workload
except ValueError:
_array_info['SLO'] = pool
array_pool.append(_array_info)
except Exception as e:
exception_message = (_(
"Unable to get the SLO/Workload combinations from the array. "
"Exception received was %(e)s") % {'e': six.text_type(e)})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
return array_pool
def create_volume(self, volume):
"""Creates a EMC(PowerMax/VMAX) volume from a storage group.
:param volume: volume object
:returns: model_update - dict
"""
model_update, rep_driver_data = dict(), dict()
volume_id = volume.id
extra_specs = self._initial_setup(volume)
if 'qos' in extra_specs:
del extra_specs['qos']
# Volume_name naming convention is 'OS-UUID'.
volume_name = self.utils.get_volume_element_name(volume_id)
volume_size = volume.size
volume_dict, rep_update, rep_info_dict = self._create_volume(
volume, volume_name, volume_size, extra_specs)
if rep_update:
rep_driver_data = rep_update['replication_driver_data']
model_update.update(rep_update)
# Add volume to group
group_name = self._add_to_group(
volume, volume_dict['device_id'], volume_name, volume.group_id,
volume.group, extra_specs, rep_driver_data)
# Gather Metadata
model_update.update(
{'provider_location': six.text_type(volume_dict)})
model_update = self.update_metadata(
model_update, volume.metadata, self.get_volume_metadata(
volume_dict['array'], volume_dict['device_id']))
if rep_update:
model_update['metadata']['BackendID'] = extra_specs[
utils.REP_CONFIG].get(utils.BACKEND_ID, 'None')
array_tag_list = self.get_tags_of_storage_array(
extra_specs[utils.ARRAY])
self.volume_metadata.capture_create_volume(
volume_dict['device_id'], volume, group_name, volume.group_id,
extra_specs, rep_info_dict, 'create',
array_tag_list=array_tag_list)
LOG.info("Leaving create_volume: %(name)s. Volume dict: %(dict)s.",
{'name': volume_name, 'dict': volume_dict})
return model_update
def _add_to_group(
self, volume, device_id, volume_name, group_id, group,
extra_specs, rep_driver_data=None):
"""Add a volume to a volume group
:param volume: volume object
:param device_id: the device id
:param volume_name: volume name
:param group_id: the group id
:param group: group object
:param extra_specs: extra specifications
:param rep_driver_data: replication data (optional)
:returns: group_id - string
"""
group_name = None
if group_id is not None:
if (volume_utils.is_group_a_cg_snapshot_type(group)
or group.is_replicated):
group_name = self._add_new_volume_to_volume_group(
volume, device_id, volume_name,
extra_specs, rep_driver_data)
return group_name
def _add_new_volume_to_volume_group(self, volume, device_id, volume_name,
extra_specs, rep_driver_data=None):
"""Add a new volume to a volume group.
This may also be called after extending a replicated volume.
:param volume: the volume object
:param device_id: the device id
:param volume_name: the volume name
:param extra_specs: the extra specifications
:param rep_driver_data: the replication driver data, optional
:returns: group_name string
"""
self.utils.check_replication_matched(volume, extra_specs)
group_name = self.provision.get_or_create_volume_group(
extra_specs[utils.ARRAY], volume.group, extra_specs)
self.masking.add_volume_to_storage_group(
extra_specs[utils.ARRAY], device_id,
group_name, volume_name, extra_specs)
# Add remote volume to remote group, if required
if volume.group.is_replicated:
self.masking.add_remote_vols_to_volume_group(
volume, volume.group, extra_specs, rep_driver_data)
return group_name
def create_volume_from_snapshot(self, volume, snapshot):
"""Creates a volume from a snapshot.
:param volume: volume object
:param snapshot: snapshot object
:returns: model_update
:raises: VolumeBackendAPIException:
"""
LOG.debug("Entering create_volume_from_snapshot.")
model_update, rep_info_dict = {}, {}
extra_specs = self._initial_setup(volume)
# Check if legacy snapshot
sourcedevice_id = self._find_device_on_array(
snapshot, extra_specs)
from_snapvx = False if sourcedevice_id else True
clone_dict, rep_update, rep_info_dict = self._create_cloned_volume(
volume, snapshot, extra_specs, from_snapvx=from_snapvx)
# Update model with replication session info if applicable
if rep_update:
model_update.update(rep_update)
model_update.update(
{'provider_location': six.text_type(clone_dict)})
model_update = self.update_metadata(
model_update, volume.metadata, self.get_volume_metadata(
clone_dict['array'], clone_dict['device_id']))
if rep_update:
model_update['metadata']['BackendID'] = extra_specs[
utils.REP_CONFIG].get(utils.BACKEND_ID, 'None')
array_tag_list = self.get_tags_of_storage_array(
extra_specs[utils.ARRAY])
self.volume_metadata.capture_create_volume(
clone_dict['device_id'], volume, None, None,
extra_specs, rep_info_dict, 'createFromSnapshot',
source_snapshot_id=snapshot.id, array_tag_list=array_tag_list)
return model_update
def create_cloned_volume(self, clone_volume, source_volume):
"""Creates a clone of the specified volume.
:param clone_volume: clone volume Object
:param source_volume: volume object
:returns: model_update, dict
"""
model_update, rep_info_dict = {}, {}
rep_driver_data = None
extra_specs = self._initial_setup(clone_volume)
array = extra_specs[utils.ARRAY]
source_device_id = self._find_device_on_array(
source_volume, extra_specs)
if not self.next_gen and (
extra_specs.get('rep_mode', None) == utils.REP_METRO):
self._sync_check(array, source_device_id, extra_specs)
else:
self._clone_check(array, source_device_id, extra_specs)
clone_dict, rep_update, rep_info_dict = self._create_cloned_volume(
clone_volume, source_volume, extra_specs)
# Update model with replication session info if applicable
if rep_update:
rep_driver_data = rep_update['replication_driver_data']
model_update.update(rep_update)
# Add volume to group
group_name = self._add_to_group(
clone_volume, clone_dict['device_id'], clone_volume.name,
source_volume.group_id, source_volume.group, extra_specs,
rep_driver_data)
model_update.update(
{'provider_location': six.text_type(clone_dict)})
model_update = self.update_metadata(
model_update, clone_volume.metadata, self.get_volume_metadata(
clone_dict['array'], clone_dict['device_id']))
if rep_update:
model_update['metadata']['BackendID'] = extra_specs[
utils.REP_CONFIG].get(utils.BACKEND_ID, 'None')
array_tag_list = self.get_tags_of_storage_array(
extra_specs[utils.ARRAY])
self.volume_metadata.capture_create_volume(
clone_dict['device_id'], clone_volume, group_name,
source_volume.group_id, extra_specs, rep_info_dict,
'createFromVolume',
temporary_snapvx=clone_dict.get('snap_name'),
source_device_id=clone_dict.get('source_device_id'),
array_tag_list=array_tag_list)
return model_update
def delete_volume(self, volume):
"""Deletes a EMC(PowerMax/VMAX) volume.
:param volume: volume object
"""
LOG.info("Deleting Volume: %(volume)s",
{'volume': volume.name})
volume_name = self._delete_volume(volume)
self.volume_metadata.capture_delete_info(volume)
LOG.info("Leaving delete_volume: %(volume_name)s.",
{'volume_name': volume_name})
def create_snapshot(self, snapshot, volume):
"""Creates a snapshot.
:param snapshot: snapshot object
:param volume: volume Object to create snapshot from
:returns: dict -- the cloned volume dictionary
"""
extra_specs = self._initial_setup(volume)
snapshot_dict, __, __ = self._create_cloned_volume(
snapshot, volume, extra_specs, is_snapshot=True)
model_update = {
'provider_location': six.text_type(snapshot_dict)}
model_update = self.update_metadata(
model_update, snapshot.metadata, self.get_snapshot_metadata(
extra_specs['array'], snapshot_dict['source_id'],
snapshot_dict['snap_name']))
if snapshot.metadata:
model_update['metadata'].update(snapshot.metadata)
self.volume_metadata.capture_snapshot_info(
volume, extra_specs, 'createSnapshot', snapshot_dict['snap_name'])
return model_update
def delete_snapshot(self, snapshot, volume):
"""Deletes a snapshot.
:param snapshot: snapshot object
:param volume: source volume
"""
LOG.info("Delete Snapshot: %(snapshotName)s.",
{'snapshotName': snapshot.name})
extra_specs = self._initial_setup(volume)
sourcedevice_id, snap_name = self._parse_snap_info(
extra_specs[utils.ARRAY], snapshot)
if not sourcedevice_id and not snap_name:
# Check if legacy snapshot
sourcedevice_id = self._find_device_on_array(
snapshot, extra_specs)
if sourcedevice_id:
self._delete_volume(snapshot)
else:
LOG.info("No snapshot found on the array")
elif not sourcedevice_id or not snap_name:
LOG.info("No snapshot found on the array")
else:
# Ensure snap has not been recently deleted
self.provision.delete_volume_snap_check_for_links(
extra_specs[utils.ARRAY], snap_name,
sourcedevice_id, extra_specs)
LOG.info("Leaving delete_snapshot: %(ssname)s.",
{'ssname': snap_name})
self.volume_metadata.capture_snapshot_info(
volume, extra_specs, 'deleteSnapshot', None)
def _remove_members(self, array, volume, device_id,
extra_specs, connector, is_multiattach,
async_grp=None, host_template=None):
"""This method unmaps a volume from a host.
Removes volume from the storage group that belongs to a masking view.
:param array: the array serial number
:param volume: volume object
:param device_id: the PowerMax/VMAX volume device id
:param extra_specs: extra specifications
:param connector: the connector object
:param is_multiattach: flag to indicate if this is a multiattach case
:param async_grp: the name if the async group, if applicable
"""
volume_name = volume.name
LOG.debug("Detaching volume %s.", volume_name)
reset = False if is_multiattach else True
self.masking.remove_and_reset_members(
array, volume, device_id, volume_name,
extra_specs, reset, connector, async_grp=async_grp,
host_template=host_template)
if is_multiattach:
self.masking.return_volume_to_fast_managed_group(
array, device_id, extra_specs)
def _unmap_lun(self, volume, connector):
"""Unmaps a volume from the host.
:param volume: the volume Object
:param connector: the connector Object
"""
mv_list, sg_list = None, None
extra_specs = self._initial_setup(volume)
rep_config = None
rep_extra_specs = None
current_host_occurances = 0
if 'qos' in extra_specs:
del extra_specs['qos']
if self.utils.is_replication_enabled(extra_specs):
backend_id = self._get_replicated_volume_backend_id(volume)
rep_config = self.utils.get_rep_config(
backend_id, self.rep_configs)
extra_specs[utils.FORCE_VOL_REMOVE] = True
rep_extra_specs = self._get_replication_extra_specs(
extra_specs, rep_config)
if self.utils.is_volume_failed_over(volume):
extra_specs = rep_extra_specs
volume_name = volume.name
mgmt_sg_name = None
LOG.info("Unmap volume: %(volume)s.", {'volume': volume})
if connector is not None:
host_name = connector.get('host')
attachment_list = volume.volume_attachment
LOG.debug("Volume attachment list: %(atl)s. "
"Attachment type: %(at)s",
{'atl': attachment_list, 'at': type(attachment_list)})
try:
att_list = attachment_list.objects
except AttributeError:
att_list = attachment_list
if att_list is not None:
host_list = [att.connector['host'] for att in att_list if
att is not None and att.connector is not None]
current_host_occurances = host_list.count(host_name)
else:
LOG.warning("Cannot get host name from connector object - "
"assuming force-detach.")
host_name = None
device_info, is_multiattach = (
self.find_host_lun_id(volume, host_name, extra_specs))
if 'hostlunid' not in device_info:
LOG.info("Volume %s is not mapped. No volume to unmap.",
volume_name)
return
if current_host_occurances > 1:
LOG.info("Volume is attached to multiple instances on "
"this host. Not removing the volume from the "
"masking view.")
else:
array = extra_specs[utils.ARRAY]
if self.utils.does_vol_need_rdf_management_group(extra_specs):
mgmt_sg_name = self.utils.get_rdf_management_group_name(
rep_config)
self._remove_members(
array, volume, device_info['device_id'], extra_specs,
connector, is_multiattach, async_grp=mgmt_sg_name,
host_template=self.powermax_short_host_name_template)
if self.utils.is_metro_device(rep_config, extra_specs):
# Need to remove from remote masking view
device_info, __ = (self.find_host_lun_id(
volume, host_name, extra_specs, rep_extra_specs))
if 'hostlunid' in device_info:
self._remove_members(
rep_extra_specs[utils.ARRAY], volume,
device_info['device_id'], rep_extra_specs, connector,
is_multiattach, async_grp=mgmt_sg_name,
host_template=self.powermax_short_host_name_template)
else:
# Make an attempt to clean up initiator group
self.masking.attempt_ig_cleanup(
connector, self.protocol,
rep_extra_specs[utils.ARRAY], True,
host_template=self.powermax_short_host_name_template)
if is_multiattach and LOG.isEnabledFor(logging.DEBUG):
mv_list, sg_list = (
self._get_mvs_and_sgs_from_volume(
extra_specs[utils.ARRAY],
device_info['device_id']))
self.volume_metadata.capture_detach_info(
volume, extra_specs, device_info['device_id'], mv_list,
sg_list)
def initialize_connection(self, volume, connector):
"""Initializes the connection and returns device and connection info.
The volume may be already mapped, if this is so the deviceInfo tuple
is returned. If the volume is not already mapped then we need to
gather information to either 1. Create an new masking view or 2. Add
the volume to an existing storage group within an already existing
maskingview.
The naming convention is the following:
.. code-block:: none
initiator_group_name = OS-<shortHostName>-<shortProtocol>-IG
e.g OS-myShortHost-I-IG
storage_group_name = OS-<shortHostName>-<srpName>-<shortProtocol>-SG
e.g OS-myShortHost-SRP_1-I-SG
port_group_name = OS-<target>-PG The port_group_name will come from
the cinder.conf or as an extra spec on the volume
type. These are precreated. If the portGroup does not
exist then an error will be returned to the user
maskingview_name = OS-<shortHostName>-<srpName>-<shortProtocol>-MV
e.g OS-myShortHost-SRP_1-I-MV
:param volume: volume Object
:param connector: the connector Object
:returns: dict -- device_info_dict - device information dict
"""
extra_specs = self._initial_setup(volume)
is_multipath = connector.get('multipath', False)
rep_config = extra_specs.get(utils.REP_CONFIG)
rep_extra_specs = self._get_replication_extra_specs(
extra_specs, rep_config)
remote_port_group = None
volume_name = volume.name
LOG.info("Initialize connection: %(volume)s.",
{'volume': volume_name})
if (self.utils.is_metro_device(rep_config, extra_specs)
and not is_multipath and self.protocol.lower() == 'iscsi'):
LOG.warning("Either multipathing is not correctly/currently "
"enabled on your system or the volume was created "
"prior to multipathing being enabled. Please refer "
"to the online PowerMax Cinder driver documentation "
"for this release for further details.")
return
if self.utils.is_volume_failed_over(volume):
extra_specs = rep_extra_specs
device_info_dict, is_multiattach = (
self.find_host_lun_id(volume, connector.get('host'), extra_specs))
masking_view_dict = self._populate_masking_dict(
volume, connector, extra_specs)
masking_view_dict[utils.IS_MULTIATTACH] = is_multiattach
if self.rest.is_next_gen_array(extra_specs['array']):
masking_view_dict['workload'] = 'NONE'
temp_pool = masking_view_dict['storagegroup_name']
splitPool = temp_pool.split('+')
if len(splitPool) == 4:
splitPool[1] = 'NONE'
masking_view_dict['storagegroup_name'] = '+'.join(splitPool)
if ('hostlunid' in device_info_dict and
device_info_dict['hostlunid'] is not None):
hostlunid = device_info_dict['hostlunid']
LOG.info("Volume %(volume)s is already mapped to host %(host)s. "
"The hostlunid is %(hostlunid)s.",
{'volume': volume_name, 'host': connector['host'],
'hostlunid': hostlunid})
port_group_name = (
self.get_port_group_from_masking_view(
extra_specs[utils.ARRAY],
device_info_dict['maskingview']))
if self.utils.is_metro_device(rep_config, extra_specs):
remote_info_dict, is_multiattach = (
self.find_host_lun_id(volume, connector.get('host'),
extra_specs, rep_extra_specs))
if remote_info_dict.get('hostlunid') is None:
# Need to attach on remote side
metro_host_lun, remote_port_group = (
self._attach_metro_volume(
volume, connector, is_multiattach, extra_specs,
rep_extra_specs))
else:
metro_host_lun = remote_info_dict['hostlunid']
remote_port_group = self.get_port_group_from_masking_view(
rep_extra_specs[utils.ARRAY],
remote_info_dict['maskingview'])
device_info_dict['metro_hostlunid'] = metro_host_lun
else:
if is_multiattach and extra_specs[utils.SLO]:
# Need to move volume to a non-fast managed storagegroup
# before attach on subsequent host(s)
masking_view_dict = self.masking.pre_multiattach(
extra_specs[utils.ARRAY],
masking_view_dict[utils.DEVICE_ID], masking_view_dict,
extra_specs)
device_info_dict, port_group_name = (
self._attach_volume(
volume, connector, extra_specs, masking_view_dict))
if self.utils.is_metro_device(rep_config, extra_specs):
# Need to attach on remote side
metro_host_lun, remote_port_group = self._attach_metro_volume(
volume, connector, is_multiattach, extra_specs,
rep_extra_specs)
device_info_dict['metro_hostlunid'] = metro_host_lun
if self.protocol.lower() == 'iscsi':
device_info_dict['ip_and_iqn'] = (
self._find_ip_and_iqns(
extra_specs[utils.ARRAY], port_group_name))
if self.utils.is_metro_device(rep_config, extra_specs):
device_info_dict['metro_ip_and_iqn'] = (
self._find_ip_and_iqns(
rep_extra_specs[utils.ARRAY], remote_port_group))
device_info_dict['is_multipath'] = is_multipath
array_tag_list = self.get_tags_of_storage_array(
extra_specs[utils.ARRAY])
if array_tag_list:
masking_view_dict['array_tag_list'] = array_tag_list
if is_multiattach and LOG.isEnabledFor(logging.DEBUG):
masking_view_dict['mv_list'], masking_view_dict['sg_list'] = (
self._get_mvs_and_sgs_from_volume(
extra_specs[utils.ARRAY],
masking_view_dict[utils.DEVICE_ID]))
elif not is_multiattach and LOG.isEnabledFor(logging.DEBUG):
masking_view_dict['tag_list'] = self.get_tags_of_storage_group(
extra_specs[utils.ARRAY], masking_view_dict[utils.SG_NAME])
self.volume_metadata.capture_attach_info(
volume, extra_specs, masking_view_dict, connector['host'],
is_multipath, is_multiattach)
return device_info_dict
def get_tags_of_storage_group(self, array, storage_group_name):
"""Get the tag information from a storage group
:param array: serial number of array
:param storage_group_name: storage group name
:returns: tag list
"""
try:
storage_group = self.rest.get_storage_group(
array, storage_group_name)
except Exception:
return None
return storage_group.get('tags')
def get_tags_of_storage_array(self, array):
"""Get the tag information from an array
:param array: serial number of array
:returns: tag list
"""
tag_name_list = None
try:
tag_name_list = self.rest.get_array_tags(array)
except Exception:
pass
return tag_name_list
def _attach_metro_volume(self, volume, connector, is_multiattach,
extra_specs, rep_extra_specs):
"""Helper method to attach a metro volume.
Metro protected volumes point to two PowerMax/VMAX devices on
different arrays, which are presented as a single device to the host.
This method masks the remote device to the host.
:param volume: the volume object
:param connector: the connector dict
:param is_multiattach: flag to indicate if this a multiattach case
:param extra_specs: the extra specifications
:param rep_extra_specs: replication extra specifications
:returns: hostlunid, remote_port_group
"""
remote_mv_dict = self._populate_masking_dict(
volume, connector, extra_specs, rep_extra_specs)
remote_mv_dict[utils.IS_MULTIATTACH] = (
True if is_multiattach else False)
if is_multiattach and rep_extra_specs[utils.SLO]:
# Need to move volume to a non-fast managed sg
# before attach on subsequent host(s)
remote_mv_dict = self.masking.pre_multiattach(
rep_extra_specs[utils.ARRAY], remote_mv_dict[utils.DEVICE_ID],
remote_mv_dict, rep_extra_specs)
remote_info_dict, remote_port_group = (
self._attach_volume(
volume, connector, extra_specs, remote_mv_dict,
rep_extra_specs=rep_extra_specs))
remote_port_group = self.get_port_group_from_masking_view(
rep_extra_specs[utils.ARRAY], remote_info_dict['maskingview'])
return remote_info_dict['hostlunid'], remote_port_group
def _attach_volume(self, volume, connector, extra_specs,
masking_view_dict, rep_extra_specs=None):
"""Attach a volume to a host.
:param volume: the volume object
:param connector: the connector object
:param extra_specs: extra specifications
:param masking_view_dict: masking view information
:param rep_extra_specs: rep extra specs are passed if metro device
:returns: dict -- device_info_dict
String -- port group name
:raises: VolumeBackendAPIException
"""
m_specs = extra_specs if rep_extra_specs is None else rep_extra_specs
rollback_dict = self.masking.setup_masking_view(
masking_view_dict[utils.ARRAY], volume,
masking_view_dict, m_specs)
# Find host lun id again after the volume is exported to the host.
device_info_dict, __ = self.find_host_lun_id(
volume, connector.get('host'), extra_specs, rep_extra_specs)
if 'hostlunid' not in device_info_dict:
# Did not successfully attach to host, so a rollback is required.
error_message = (_("Error Attaching volume %(vol)s. Cannot "
"retrieve hostlunid.") % {'vol': volume.id})
LOG.error(error_message)
self.masking.check_if_rollback_action_for_masking_required(
masking_view_dict[utils.ARRAY], volume,
masking_view_dict[utils.DEVICE_ID], rollback_dict)
raise exception.VolumeBackendAPIException(
message=error_message)
return device_info_dict, rollback_dict[utils.PORTGROUPNAME]
def terminate_connection(self, volume, connector):
"""Disallow connection from connector.
:param volume: the volume Object
:param connector: the connector Object
"""
volume_name = volume.name
LOG.info("Terminate connection: %(volume)s.",
{'volume': volume_name})
self._unmap_lun(volume, connector)
def extend_volume(self, volume, new_size):
"""Extends an existing volume.
:param volume: the volume Object
:param new_size: the new size to increase the volume to
:raises: VolumeBackendAPIException:
"""
# Set specific attributes for extend operation
ex_specs = self._initial_setup(volume)
array = ex_specs[utils.ARRAY]
device_id = self._find_device_on_array(volume, ex_specs)
vol_name = volume.name
orig_vol_size = volume.size
rep_enabled = self.utils.is_replication_enabled(ex_specs)
rdf_grp_no = None
legacy_extend = False
# Run validation and capabilities checks
self._extend_vol_validation_checks(
array, device_id, vol_name, ex_specs, orig_vol_size, new_size)
# Get extend workflow dependent on array gen and replication status
if rep_enabled:
rep_config = ex_specs[utils.REP_CONFIG]
rdf_grp_no, __ = self.get_rdf_details(array, rep_config)
r1_ode, r1_ode_metro, r2_ode, r2_ode_metro = (
self._array_ode_capabilities_check(array, True))
if self.next_gen:
if self.utils.is_metro_device(rep_config, ex_specs):
if not r1_ode_metro or not r2_ode or not r2_ode_metro:
legacy_extend = True
else:
legacy_extend = True
# Handle the extend process using workflow info from previous steps
if legacy_extend:
rep_config = ex_specs[utils.REP_CONFIG]
if rep_config.get('allow_extend', False):
LOG.info("Legacy extend volume %(volume)s to %(new_size)d GBs",
{'volume': vol_name, 'new_size': int(new_size)})
self._extend_legacy_replicated_vol(
array, volume, device_id, vol_name, new_size, ex_specs,
rdf_grp_no)
else:
exception_message = (
"Extending a replicated volume on this backend is not "
"permitted. Please set 'allow_extend:True' in your "
"PowerMax replication target_backend configuration.")
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
else:
LOG.info("ODE extend volume %(volume)s to %(new_size)d GBs",
{'volume': vol_name,
'new_size': int(new_size)})
self.provision.extend_volume(
array, device_id, new_size, ex_specs, rdf_grp_no)
self.volume_metadata.capture_extend_info(
volume, new_size, device_id, ex_specs, array)
LOG.debug("Leaving extend_volume: %(volume_name)s. ",
{'volume_name': vol_name})
def _extend_vol_validation_checks(self, array, device_id, vol_name,
ex_specs, orig_vol_size, new_size):
"""Run validation checks on settings for extend volume operation.
:param array: the array serial number
:param device_id: the device id
:param vol_name: the volume name
:param ex_specs: extra specifications
:param orig_vol_size: the original volume size
:param new_size: the new size the volume should be
:raises: VolumeBackendAPIException:
"""
# 1 - Check device exists
if device_id is None:
exception_message = (_(
"Cannot find Volume: %(volume_name)s. Extend operation. "
"Exiting....") % {'volume_name': vol_name})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
# 2 - Check if volume is part of an on-going clone operation or if vol
# has source snapshots but not next-gen array
self._sync_check(array, device_id, ex_specs)
__, snapvx_src, __ = self.rest.is_vol_in_rep_session(array, device_id)
if snapvx_src:
if not self.next_gen:
exception_message = (
_("The volume: %(volume)s is a snapshot source. "
"Extending a volume with snapVx snapshots is only "
"supported on PowerMax/VMAX from OS version 5978 "
"onwards. Exiting...") % {'volume': vol_name})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
# 3 - Check new size is larger than old size
if int(orig_vol_size) >= int(new_size):
exception_message = (_(
"Your original size: %(orig_vol_size)s GB is greater "
"than or the same as: %(new_size)s GB. Only extend ops are "
"supported. Exiting...") % {'orig_vol_size': orig_vol_size,
'new_size': new_size})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
def _array_ode_capabilities_check(self, array, rep_config,
rep_enabled=False):
"""Given an array, check Online Device Expansion (ODE) support.
:param array: the array serial number
:param rep_config: the replication configuration
:param rep_enabled: if replication is enabled for backend
:returns: r1_ode: (bool) If R1 array supports ODE
:returns: r1_ode_metro: (bool) If R1 array supports ODE with Metro vols
:returns: r2_ode: (bool) If R2 array supports ODE
:returns: r2_ode_metro: (bool) If R2 array supports ODE with Metro vols
"""
r1_ucode = self.ucode_level.split('.')
r1_ode, r1_ode_metro = False, False
r2_ode, r2_ode_metro = False, False
if self.next_gen:
r1_ode = True
if rep_enabled:
__, r2_array = self.get_rdf_details(array, rep_config)
r2_ucode = self.rest.get_array_ucode_version(r2_array)
if int(r1_ucode[2]) > utils.UCODE_5978_ELMSR:
r1_ode_metro = True
r2_ucode = r2_ucode.split('.')
if self.rest.is_next_gen_array(r2_array):
r2_ode = True
if int(r2_ucode[2]) > utils.UCODE_5978_ELMSR:
r2_ode_metro = True
return r1_ode, r1_ode_metro, r2_ode, r2_ode_metro
@coordination.synchronized('emc-{rdf_group_no}-rdf')
def _extend_legacy_replicated_vol(
self, array, volume, device_id, volume_name, new_size, extra_specs,
rdf_group_no):
"""Extend a legacy OS volume without Online Device Expansion
:param array: the array serial number
:param volume: the volume objcet
:param device_id: the volume device id
:param volume_name: the volume name
:param new_size: the new size the volume should be
:param extra_specs: extra specifications
:param rdf_group_no: the RDF group number
"""
try:
# Break the RDF device pair relationship and cleanup R2
LOG.info("Breaking replication relationship...")
self.break_rdf_device_pair_session(
array, device_id, volume_name, extra_specs)
# Extend the R1 volume
LOG.info("Extending source volume...")
self.provision.extend_volume(
array, device_id, new_size, extra_specs)
# Setup volume replication again for source volume
LOG.info("Recreating replication relationship...")
rep_status, __, __, rep_extra_specs, resume_rdf = (
self.configure_volume_replication(
array, volume, device_id, extra_specs))
# If first/only volume in SG then RDF protect SG
if rep_status == 'first_vol_in_rdf_group':
self._protect_storage_group(
array, device_id, volume, volume_name, rep_extra_specs)
# If more than one volume in SG then resume replication
if resume_rdf:
self.rest.srdf_resume_replication(
array, rep_extra_specs['mgmt_sg_name'],
rep_extra_specs['rdf_group_no'], extra_specs)
except Exception as e:
exception_message = (_("Error extending volume. Error received "
"was %(e)s") % {'e': e})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
def update_volume_stats(self):
"""Retrieve stats info."""
if self.rest.u4p_in_failover and self.rest.u4p_failover_autofailback:
self.retest_primary_u4p()
pools = []
# Dictionary to hold the arrays for which the SRP details
# have already been queried.
arrays = {}
total_capacity_gb = 0
free_capacity_gb = 0
provisioned_capacity_gb = 0
location_info = None
backend_name = self.pool_info['backend_name']
max_oversubscription_ratio = (
self.pool_info['max_over_subscription_ratio'])
reserved_percentage = self.pool_info['reserved_percentage']
array_reserve_percent = None
array_info_list = self.pool_info['arrays_info']
already_queried = False
for array_info in array_info_list:
if self.failover:
rep_config = self.utils.get_rep_config(
self.active_backend_id, self.rep_configs)
array_info = self.get_secondary_stats_info(
rep_config, array_info)
# Add both SLO & Workload name in the pool name
# Only insert the array details in the dict once
if array_info['SerialNumber'] not in arrays:
(location_info, total_capacity_gb, free_capacity_gb,
provisioned_capacity_gb,
array_reserve_percent) = self._update_srp_stats(array_info)
arrays[array_info['SerialNumber']] = (
[total_capacity_gb, free_capacity_gb,
provisioned_capacity_gb, array_reserve_percent])
else:
already_queried = True
try:
pool_name = ("%(slo)s+%(workload)s+%(srpName)s+%(array)s"
% {'slo': array_info['SLO'],
'workload': array_info['Workload'],
'srpName': array_info['srpName'],
'array': array_info['SerialNumber']})
except KeyError:
pool_name = ("%(slo)s+%(srpName)s+%(array)s"
% {'slo': array_info['SLO'],
'srpName': array_info['srpName'],
'array': array_info['SerialNumber']})
if already_queried:
# The dictionary will only have one key per PowerMax/VMAX
# Construct the location info
pool = self._construct_location_info_and_pool(
array_info, pool_name, arrays, max_oversubscription_ratio,
reserved_percentage)
else:
pool = {'pool_name': pool_name,
'total_capacity_gb': total_capacity_gb,
'free_capacity_gb': free_capacity_gb,
'provisioned_capacity_gb': provisioned_capacity_gb,
'QoS_support': False,
'location_info': location_info,
'consistencygroup_support': False,
'thin_provisioning_support': True,
'thick_provisioning_support': False,
'consistent_group_snapshot_enabled': True,
'max_over_subscription_ratio':
max_oversubscription_ratio,
'reserved_percentage': reserved_percentage,
'replication_enabled': self.replication_enabled,
'group_replication_enabled': self.replication_enabled,
'consistent_group_replication_enabled':
self.replication_enabled
}
if array_reserve_percent:
if isinstance(reserved_percentage, int):
if array_reserve_percent > reserved_percentage:
pool['reserved_percentage'] = array_reserve_percent
else:
pool['reserved_percentage'] = array_reserve_percent
pools.append(pool)
pools = self.utils.add_legacy_pools(pools)
data = {'vendor_name': "Dell EMC",
'driver_version': self.version,
'storage_protocol': 'unknown',
'volume_backend_name': backend_name or
self.__class__.__name__,
# Use zero capacities here so we always use a pool.
'total_capacity_gb': 0,
'free_capacity_gb': 0,
'provisioned_capacity_gb': 0,
'reserved_percentage': 0,
'replication_enabled': self.replication_enabled,
'replication_targets': self.replication_targets,
'pools': pools}
return data
def _construct_location_info_and_pool(
self, array_info, pool_name, arrays, max_oversubscription_ratio,
reserved_percentage):
"""Construct the location info string and the pool dict
:param array_info: array information dict
:param pool_name: pool name
:param arrays: arrays dict
:param max_oversubscription_ratio: max oversubscription ratio
:param reserved_percentage: reserved percentage
:returns: pool - dict
"""
try:
temp_location_info = (
("%(arrayName)s#%(srpName)s#%(slo)s#%(workload)s"
% {'arrayName': array_info['SerialNumber'],
'srpName': array_info['srpName'],
'slo': array_info['SLO'],
'workload': array_info['Workload']}))
except KeyError:
temp_location_info = (
("%(arrayName)s#%(srpName)s#%(slo)s"
% {'arrayName': array_info['SerialNumber'],
'srpName': array_info['srpName'],
'slo': array_info['SLO']}))
pool = {'pool_name': pool_name,
'total_capacity_gb':
arrays[array_info['SerialNumber']][0],
'free_capacity_gb':
arrays[array_info['SerialNumber']][1],
'provisioned_capacity_gb':
arrays[array_info['SerialNumber']][2],
'QoS_support': False,
'location_info': temp_location_info,
'thin_provisioning_support': True,
'thick_provisioning_support': False,
'consistent_group_snapshot_enabled': True,
'max_over_subscription_ratio':
max_oversubscription_ratio,
'reserved_percentage': reserved_percentage,
'replication_enabled': self.replication_enabled,
'multiattach': True}
if arrays[array_info['SerialNumber']][3]:
if reserved_percentage:
if (arrays[array_info['SerialNumber']][3] >
reserved_percentage):
pool['reserved_percentage'] = (
arrays[array_info['SerialNumber']][3])
else:
pool['reserved_percentage'] = (
arrays[array_info['SerialNumber']][3])
return pool
def _update_srp_stats(self, array_info):
"""Update SRP stats.
:param array_info: array information
:returns: location_info
:returns: totalManagedSpaceGbs
:returns: remainingManagedSpaceGbs
:returns: provisionedManagedSpaceGbs
:returns: array_reserve_percent
:returns: wlpEnabled
"""
(totalManagedSpaceGbs, remainingManagedSpaceGbs,
provisionedManagedSpaceGbs, array_reserve_percent) = (
self.provision.get_srp_pool_stats(
array_info['SerialNumber'], array_info))
LOG.info("Capacity stats for SRP pool %(srpName)s on array "
"%(arrayName)s total_capacity_gb=%(total_capacity_gb)lu, "
"free_capacity_gb=%(free_capacity_gb)lu, "
"provisioned_capacity_gb=%(provisioned_capacity_gb)lu",
{'srpName': array_info['srpName'],
'arrayName': array_info['SerialNumber'],
'total_capacity_gb': totalManagedSpaceGbs,
'free_capacity_gb': remainingManagedSpaceGbs,
'provisioned_capacity_gb': provisionedManagedSpaceGbs})
try:
location_info = ("%(arrayName)s#%(srpName)s#%(slo)s#%(workload)s"
% {'arrayName': array_info['SerialNumber'],
'srpName': array_info['srpName'],
'slo': array_info['SLO'],
'workload': array_info['Workload']})
except KeyError:
location_info = ("%(arrayName)s#%(srpName)s#%(slo)s"
% {'arrayName': array_info['SerialNumber'],
'srpName': array_info['srpName'],
'slo': array_info['SLO']})
return (location_info, totalManagedSpaceGbs,
remainingManagedSpaceGbs, provisionedManagedSpaceGbs,
array_reserve_percent)
def _set_config_file_and_get_extra_specs(self, volume,
volume_type_id=None):
"""Given the volume object get the associated volumetype.
Given the volume object get the associated volumetype and the
extra specs associated with it.
Based on the name of the config group, register the config file
:param volume: the volume object including the volume_type_id
:param volume_type_id: Optional override of volume.volume_type_id
:returns: dict -- the extra specs dict
:returns: dict -- QoS specs
"""
qos_specs = {}
extra_specs = self.utils.get_volumetype_extra_specs(
volume, volume_type_id)
type_id = volume.volume_type_id
if type_id:
res = volume_types.get_volume_type_qos_specs(type_id)
qos_specs = res['qos_specs']
# If there are no extra specs then the default case is assumed.
if extra_specs:
if extra_specs.get('replication_enabled') == '<is> True':
extra_specs[utils.IS_RE] = True
backend_id = self._get_replicated_volume_backend_id(volume)
rep_config = self.utils.get_rep_config(
backend_id, self.rep_configs)
if rep_config is None:
msg = _('Could not determine which rep_device to use '
'from cinder.conf')
raise exception.VolumeBackendAPIException(msg)
extra_specs[utils.REP_CONFIG] = rep_config
if rep_config.get('mode'):
extra_specs[utils.REP_MODE] = rep_config['mode']
if rep_config.get(utils.METROBIAS):
extra_specs[utils.METROBIAS] = (
rep_config[utils.METROBIAS])
return extra_specs, qos_specs
def _get_replicated_volume_backend_id(self, volume):
"""Given a volume, return its rep device backend id.
:param volume: volume used to retrieve backend id -- volume
:returns: backend id -- str
"""
backend_id = utils.BACKEND_ID_LEGACY_REP
volume_extra_specs = self.utils.get_volumetype_extra_specs(volume)
if volume_extra_specs:
volume_backend_id = volume_extra_specs.get(
utils.REPLICATION_DEVICE_BACKEND_ID)
if volume_backend_id:
backend_id = volume_backend_id
return backend_id
def _find_device_on_array(self, volume, extra_specs):
"""Given the volume get the PowerMax/VMAX device Id.
:param volume: volume object
:param extra_specs: the extra Specs
:returns: array, device_id
"""
founddevice_id = None
volume_name = volume.id
try:
name_id = volume._name_id
except AttributeError:
name_id = None
loc = volume.provider_location
if isinstance(loc, six.string_types):
name = ast.literal_eval(loc)
array = extra_specs[utils.ARRAY]
if name.get('device_id'):
device_id = name['device_id']
elif name.get('keybindings'):
device_id = name['keybindings']['DeviceID']
else:
device_id = None
try:
founddevice_id = self.rest.check_volume_device_id(
array, device_id, volume_name, name_id)
except exception.VolumeBackendAPIException:
pass
if founddevice_id is None:
LOG.debug("Volume %(volume_name)s not found on the array.",
{'volume_name': volume_name})
else:
LOG.debug("Volume name: %(volume_name)s Volume device id: "
"%(founddevice_id)s.",
{'volume_name': volume_name,
'founddevice_id': founddevice_id})
return founddevice_id
def find_host_lun_id(self, volume, host, extra_specs,
rep_extra_specs=None):
"""Given the volume dict find the host lun id for a volume.
:param volume: the volume dict
:param host: host from connector (can be None on a force-detach)
:param extra_specs: the extra specs
:param rep_extra_specs: rep extra specs, passed in if metro device
:returns: dict -- the data dict
"""
maskedvols = {}
is_multiattach = False
volume_name = volume.name
device_id = self._find_device_on_array(volume, extra_specs)
if rep_extra_specs:
rdf_pair_info = self.rest.get_rdf_pair_volume(
extra_specs[utils.ARRAY], rep_extra_specs['rdf_group_no'],
device_id)
device_id = rdf_pair_info.get('remoteVolumeName', None)
extra_specs = rep_extra_specs
host_name = self.utils.get_host_name_label(
host, self.powermax_short_host_name_template) if host else None
if device_id:
array = extra_specs[utils.ARRAY]
# Return only masking views for this host
host_maskingviews, all_masking_view_list = (
self._get_masking_views_from_volume(
array, device_id, host_name))
if not host_maskingviews:
# Backward compatibility if a new template was added to
# an existing backend.
host_name = self.utils.get_host_short_name(
host) if host else None
host_maskingviews, all_masking_view_list = (
self._get_masking_views_from_volume_for_host(
all_masking_view_list, host_name))
for maskingview in host_maskingviews:
host_lun_id = self.rest.find_mv_connections_for_vol(
array, maskingview, device_id)
if host_lun_id is not None:
devicedict = {'hostlunid': host_lun_id,
'maskingview': maskingview,
'array': array,
'device_id': device_id}
maskedvols = devicedict
if not maskedvols:
LOG.debug(
"Host lun id not found for volume: %(volume_name)s "
"with the device id: %(device_id)s on host: %(host)s.",
{'volume_name': volume_name,
'device_id': device_id, 'host': host_name})
if len(all_masking_view_list) > len(host_maskingviews):
other_maskedvols = []
for maskingview in all_masking_view_list:
host_lun_id = self.rest.find_mv_connections_for_vol(
array, maskingview, device_id)
if host_lun_id is not None:
devicedict = {'hostlunid': host_lun_id,
'maskingview': maskingview,
'array': array,
'device_id': device_id}
other_maskedvols.append(devicedict)
if len(other_maskedvols) > 0:
LOG.debug("Volume is masked to a different host "
"than %(host)s - Live Migration or Multi-Attach "
"use case.", {'host': host})
is_multiattach = True
else:
exception_message = (_("Cannot retrieve volume %(vol)s "
"from the array.") % {'vol': volume_name})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(exception_message)
return maskedvols, is_multiattach
def get_masking_views_from_volume(self, array, volume, device_id, host):
"""Get all masking views from a volume.
:param array: array serial number
:param volume: the volume object
:param device_id: the volume device id
:param host: the host
:returns: masking view list, is metro
"""
is_metro = False
extra_specs = self._initial_setup(volume)
mv_list, __ = self._get_masking_views_from_volume(array, device_id,
host)
if self.utils.is_metro_device(
extra_specs.get(utils.REP_CONFIG), extra_specs):
is_metro = True
return mv_list, is_metro
def _get_masking_views_from_volume(self, array, device_id, host):
"""Helper function to retrieve masking view list for a volume.
:param array: array serial number
:param device_id: the volume device id
:param host: the host
:returns: masking view list, all masking view list
"""
LOG.debug("Getting masking views from volume")
mvs, __ = self._get_mvs_and_sgs_from_volume(array, device_id)
return self._get_masking_views_from_volume_for_host(mvs, host)
def _get_masking_views_from_volume_for_host(
self, masking_views, host_name):
"""Check all masking views for host_name
:param masking_views: list of masking view
:param host_name: the host name for comparision
:returns: masking view list, all masking view list
"""
LOG.debug("Getting masking views from volume for host %(host)s ",
{'host': host_name})
host_masking_view_list, all_masking_view_list = [], []
for masking_view in masking_views:
all_masking_view_list.append(masking_view)
if host_name:
if host_name.lower() in masking_view.lower():
host_masking_view_list.append(masking_view)
host_masking_view_list = (host_masking_view_list if host_name else
all_masking_view_list)
return host_masking_view_list, all_masking_view_list
def _get_mvs_and_sgs_from_volume(self, array, device_id):
"""Helper function to retrieve masking views and storage groups.
:param array: array serial number
:param device_id: the volume device id
:returns: masking view list, storage group list
"""
final_masking_view_list = []
storage_group_list = self.rest.get_storage_groups_from_volume(
array, device_id)
for sg in storage_group_list:
masking_view_list = self.rest.get_masking_views_from_storage_group(
array, sg)
final_masking_view_list.extend(masking_view_list)
return final_masking_view_list, storage_group_list
def _initial_setup(self, volume, volume_type_id=None):
"""Necessary setup to accumulate the relevant information.
The volume object has a host in which we can parse the
config group name. The config group name is the key to our EMC
configuration file. The emc configuration file contains srp name
and array name which are mandatory fields.
:param volume: the volume object
:param volume_type_id: optional override of volume.volume_type_id
:returns: dict -- extra spec dict
:raises: VolumeBackendAPIException:
"""
try:
array_info = self.get_attributes_from_cinder_config()
if array_info:
extra_specs, qos_specs = (
self._set_config_file_and_get_extra_specs(
volume, volume_type_id))
else:
exception_message = (_(
"Unable to get corresponding record for srp. Please "
"refer to the current online documentation for correct "
"configuration and note that the xml file is no longer "
"supported."))
raise exception.VolumeBackendAPIException(
message=exception_message)
extra_specs = self._set_vmax_extra_specs(extra_specs, array_info)
if qos_specs and qos_specs.get('consumer') != "front-end":
extra_specs['qos'] = qos_specs.get('specs')
except Exception:
exception_message = (_(
"Unable to get configuration information necessary to "
"create a volume: %(errorMessage)s.")
% {'errorMessage': sys.exc_info()[1]})
raise exception.VolumeBackendAPIException(
message=exception_message)
return extra_specs
def _populate_masking_dict(self, volume, connector,
extra_specs, rep_extra_specs=None):
"""Get all the names of the maskingview and sub-components.
:param volume: the volume object
:param connector: the connector object
:param extra_specs: extra specifications
:param rep_extra_specs: replication extra specs, if metro volume
:returns: dict -- a dictionary with masking view information
"""
masking_view_dict = {}
volume_name = volume.name
device_id = self._find_device_on_array(volume, extra_specs)
if rep_extra_specs is not None:
rdf_pair_info = self.rest.get_rdf_pair_volume(
extra_specs[utils.ARRAY], rep_extra_specs['rdf_group_no'],
device_id)
device_id = rdf_pair_info.get('remoteVolumeName', None)
extra_specs = rep_extra_specs
if not device_id:
exception_message = (_("Cannot retrieve volume %(vol)s "
"from the array. ") % {'vol': volume_name})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(exception_message)
protocol = self.utils.get_short_protocol_type(self.protocol)
short_host_name = self.utils.get_host_name_label(
connector['host'], self.powermax_short_host_name_template)
masking_view_dict[utils.USED_HOST_NAME] = short_host_name
masking_view_dict[utils.SLO] = extra_specs[utils.SLO]
masking_view_dict[utils.WORKLOAD] = 'NONE' if self.next_gen else (
extra_specs[utils.WORKLOAD])
masking_view_dict[utils.ARRAY] = extra_specs[utils.ARRAY]
masking_view_dict[utils.SRP] = extra_specs[utils.SRP]
if not extra_specs[utils.PORTGROUPNAME]:
LOG.warning("You must supply a valid pre-created port group "
"in cinder.conf or as an extra spec. Port group "
"cannot be left empty as creating a new masking "
"view will fail.")
masking_view_dict[utils.PORT_GROUP_LABEL] = (
self.utils.get_port_name_label(
extra_specs[utils.PORTGROUPNAME],
self.powermax_port_group_name_template))
masking_view_dict[utils.PORTGROUPNAME] = (
extra_specs[utils.PORTGROUPNAME])
masking_view_dict[utils.INITIATOR_CHECK] = (
self._get_initiator_check_flag())
child_sg_name, do_disable_compression, rep_enabled = (
self.utils.get_child_sg_name(
short_host_name, extra_specs,
masking_view_dict[utils.PORT_GROUP_LABEL]))
masking_view_dict[utils.DISABLECOMPRESSION] = do_disable_compression
masking_view_dict[utils.IS_RE] = rep_enabled
mv_prefix = (
"OS-%(shortHostName)s-%(protocol)s-%(pg)s"
% {'shortHostName': short_host_name,
'protocol': protocol,
'pg': masking_view_dict[utils.PORT_GROUP_LABEL]})
masking_view_dict[utils.SG_NAME] = child_sg_name
masking_view_dict[utils.MV_NAME] = ("%(prefix)s-MV"
% {'prefix': mv_prefix})
masking_view_dict[utils.PARENT_SG_NAME] = ("%(prefix)s-SG"
% {'prefix': mv_prefix})
masking_view_dict[utils.IG_NAME] = (
("OS-%(shortHostName)s-%(protocol)s-IG"
% {'shortHostName': short_host_name,
'protocol': protocol}))
masking_view_dict[utils.CONNECTOR] = connector
masking_view_dict[utils.DEVICE_ID] = device_id
masking_view_dict[utils.VOL_NAME] = volume_name
return masking_view_dict
def _create_cloned_volume(
self, volume, source_volume, extra_specs, is_snapshot=False,
from_snapvx=False):
"""Create a clone volume from the source volume.
:param volume: clone volume
:param source_volume: source of the clone volume
:param extra_specs: extra specs
:param is_snapshot: boolean -- Defaults to False
:param from_snapvx: bool -- Defaults to False
:returns: dict -- cloneDict the cloned volume dictionary
:raises: VolumeBackendAPIException:
"""
clone_name = volume.name
snap_name = None
rep_update, rep_info_dict = dict(), dict()
LOG.info("Create a replica from Volume: Clone Volume: %(clone_name)s "
"from Source Volume: %(source_name)s.",
{'clone_name': clone_name,
'source_name': source_volume.name})
array = extra_specs[utils.ARRAY]
is_clone_license = self.rest.is_snapvx_licensed(array)
if not is_clone_license:
exception_message = (_(
"SnapVx feature is not licensed on %(array)s.")
% {'array': array})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
if from_snapvx:
source_device_id, snap_name = self._parse_snap_info(
array, source_volume)
else:
source_device_id = self._find_device_on_array(
source_volume, extra_specs)
if not source_device_id:
exception_message = (_(
"Cannot find source device on %(array)s.")
% {'array': array})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
# Perform any snapvx cleanup if required before creating the clone
if is_snapshot or from_snapvx:
self._clone_check(array, source_device_id, extra_specs)
if not is_snapshot:
clone_dict, rep_update, rep_info_dict = self._create_replica(
array, volume, source_device_id, extra_specs,
snap_name=snap_name)
else:
clone_dict = self._create_snapshot(
array, volume, source_device_id, extra_specs)
LOG.debug("Leaving _create_cloned_volume: Volume: "
"%(clone_name)s Source Device Id: %(source_name)s ",
{'clone_name': clone_name,
'source_name': source_device_id})
return clone_dict, rep_update, rep_info_dict
def _parse_snap_info(self, array, snapshot):
"""Given a snapshot object, parse the provider_location.
:param array: the array serial number
:param snapshot: the snapshot object
:returns: sourcedevice_id, foundsnap_name
"""
foundsnap_name = None
sourcedevice_id = None
volume_name = snapshot.id
loc = snapshot.provider_location
if isinstance(loc, six.string_types):
name = ast.literal_eval(loc)
try:
sourcedevice_id = name['source_id']
snap_name = name['snap_name']
except KeyError:
LOG.info("Error retrieving snapshot details. Assuming "
"legacy structure of snapshot...")
return None, None
# Ensure snapvx is on the array.
try:
snap_details = self.rest.get_volume_snap(
array, sourcedevice_id, snap_name)
if snap_details:
foundsnap_name = snap_name
except Exception as e:
LOG.info("Exception in retrieving snapshot: %(e)s.",
{'e': e})
foundsnap_name = None
if foundsnap_name is None or sourcedevice_id is None:
LOG.debug("Error retrieving snapshot details. "
"Snapshot name: %(snap)s",
{'snap': volume_name})
else:
LOG.debug("Source volume: %(volume_name)s Snap name: "
"%(foundsnap_name)s.",
{'volume_name': sourcedevice_id,
'foundsnap_name': foundsnap_name})
return sourcedevice_id, foundsnap_name
def _create_snapshot(self, array, snapshot,
source_device_id, extra_specs):
"""Create a snap Vx of a volume.
:param array: the array serial number
:param snapshot: the snapshot object
:param source_device_id: the source device id
:param extra_specs: the extra specifications
:returns: snap_dict
"""
clone_name = self.utils.get_volume_element_name(snapshot.id)
snap_name = self.utils.truncate_string(clone_name, 19)
try:
self.provision.create_volume_snapvx(array, source_device_id,
snap_name, extra_specs)
except Exception as e:
exception_message = (_("Error creating snap Vx of %(vol)s. "
"Exception received: %(e)s.")
% {'vol': source_device_id,
'e': six.text_type(e)})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
snap_dict = {'snap_name': snap_name, 'source_id': source_device_id}
return snap_dict
def _delete_volume(self, volume):
"""Helper function to delete the specified volume.
Pass in host if is snapshot
:param volume: volume object to be deleted
:returns: volume_name (string vol name)
"""
source_device_id = None
volume_name = volume.name
extra_specs = self._initial_setup(volume)
prov_loc = volume.provider_location
if isinstance(prov_loc, six.string_types):
name = ast.literal_eval(prov_loc)
source_device_id = name.get('source_device_id')
device_id = self._find_device_on_array(volume, extra_specs)
if device_id is None:
LOG.error("Volume %(name)s not found on the array. "
"No volume to delete.",
{'name': volume_name})
return volume_name
array = extra_specs[utils.ARRAY]
if self.utils.is_replication_enabled(extra_specs):
self._validate_rdfg_status(array, extra_specs)
# Check if the volume being deleted is a
# source or target for copy session
self._sync_check(array, device_id, extra_specs,
source_device_id=source_device_id)
# Remove from any storage groups and cleanup replication
self._remove_vol_and_cleanup_replication(
array, device_id, volume_name, extra_specs, volume)
# Check if volume is in any storage group
sg_list = self.rest.get_storage_groups_from_volume(array, device_id)
if sg_list:
LOG.error("Device %(device_id)s is in storage group(s) "
"%(sg_list)s prior to delete. Delete will fail.",
{'device_id': device_id, 'sg_list': sg_list})
self._delete_from_srp(
array, device_id, volume_name, extra_specs)
return volume_name
def _create_volume(self, volume, volume_name, volume_size, extra_specs):
"""Create a volume.
:param volume: the volume
:param volume_name: the volume name
:param volume_size: the volume size
:param extra_specs: extra specifications
:returns: volume_dict, rep_update, rep_info_dict --dict
"""
# Set Create Volume options
is_re, rep_mode, storagegroup_name = False, None, None
rep_info_dict, rep_update = dict(), dict()
# Get Array details
array = extra_specs[utils.ARRAY]
array_model, next_gen = self.rest.get_array_model_info(array)
if next_gen:
extra_specs[utils.WORKLOAD] = 'NONE'
# Verify valid SL/WL combination
is_valid_slo, is_valid_workload = self.provision.verify_slo_workload(
array, extra_specs[utils.SLO],
extra_specs[utils.WORKLOAD], next_gen, array_model)
if not is_valid_slo or not is_valid_workload:
exception_message = (_(
"Either SLO: %(slo)s or workload %(workload)s is invalid. "
"Examine previous error statement for valid values.")
% {'slo': extra_specs[utils.SLO],
'workload': extra_specs[utils.WORKLOAD]})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
LOG.debug("Create Volume: %(volume)s Srp: %(srp)s "
"Array: %(array)s "
"Size: %(size)lu.",
{'volume': volume_name,
'srp': extra_specs[utils.SRP],
'array': array,
'size': volume_size})
do_disable_compression = self.utils.is_compression_disabled(
extra_specs)
if self.utils.is_replication_enabled(extra_specs):
is_re, rep_mode = True, extra_specs['rep_mode']
storagegroup_name = self.masking.get_or_create_default_storage_group(
array, extra_specs[utils.SRP], extra_specs[utils.SLO],
extra_specs[utils.WORKLOAD], extra_specs,
do_disable_compression, is_re, rep_mode)
if not is_re:
volume_dict = self._create_non_replicated_volume(
array, volume, volume_name, storagegroup_name,
volume_size, extra_specs)
else:
volume_dict, rep_update, rep_info_dict = (
self._create_replication_enabled_volume(
array, volume, volume_name, volume_size, extra_specs,
storagegroup_name, rep_mode))
# Compare volume ID against identifier on array. Update if needed.
# This can occur in cases where multiple edits are occurring at once.
found_device_id = self.rest.find_volume_device_id(array, volume_name)
returning_device_id = volume_dict['device_id']
if found_device_id != returning_device_id:
volume_dict['device_id'] = found_device_id
return volume_dict, rep_update, rep_info_dict
@coordination.synchronized("emc-nonrdf-vol-{storagegroup_name}-{array}")
def _create_non_replicated_volume(
self, array, volume, volume_name, storagegroup_name, volume_size,
extra_specs):
"""Create a volume without replication enabled
:param array: the primary array -- string
:param volume: the volume -- dict
:param volume_name: the volume name -- string
:param storagegroup_name: the storage group name -- string
:param volume_size: the volume size -- string
:param extra_specs: extra specifications -- dict
:return: volume_dict -- dict
:raises: VolumeBackendAPIException:
"""
existing_devices = self.rest.get_volumes_in_storage_group(
array, storagegroup_name)
try:
volume_dict = self.provision.create_volume_from_sg(
array, volume_name, storagegroup_name,
volume_size, extra_specs, rep_info=None)
return volume_dict
except Exception as e:
try:
# Attempt cleanup of storage group post exception.
updated_devices = set(self.rest.get_volumes_in_storage_group(
array, storagegroup_name))
devices_to_delete = [device for device in updated_devices
if device not in existing_devices]
if devices_to_delete:
self._cleanup_non_rdf_volume_create_post_failure(
volume, volume_name, extra_specs, devices_to_delete)
elif not existing_devices:
self.rest.delete_storage_group(array, storagegroup_name)
finally:
# Pass actual exception that was raised now that cleanup
# attempt is finished. Mainly VolumeBackendAPIException raised
# from error status codes returned from the various REST jobs.
raise e
@coordination.synchronized('emc-rdf-vol-{storagegroup_name}-{array}')
def _create_replication_enabled_volume(
self, array, volume, volume_name, volume_size, extra_specs,
storagegroup_name, rep_mode):
"""Create a volume with replication enabled
:param array: the primary array
:param volume: the volume
:param volume_name: the volume name
:param volume_size: the volume size
:param extra_specs: extra specifications
:param storagegroup_name: the storage group name
:param rep_mode: the replication mode
:returns: volume_dict, rep_update, rep_info_dict --dict
:raises: VolumeBackendAPIException:
"""
def _is_first_vol_in_replicated_sg():
vol_dict = dict()
first_vol, rep_ex_specs, rep_info, rdfg_empty = (
self.prepare_replication_details(extra_specs))
if first_vol:
vol_dict = self.provision.create_volume_from_sg(
array, volume_name, storagegroup_name,
volume_size, extra_specs, rep_info)
rep_vol = deepcopy(vol_dict)
rep_vol.update({'device_uuid': volume_name,
'storage_group': storagegroup_name,
'size': volume_size})
if first_vol and rdfg_empty:
# First volume in SG, first volume in RDFG
self.srdf_protect_storage_group(
extra_specs, rep_ex_specs, rep_vol)
elif not rdfg_empty and not rep_info:
# First volume in SG, not first in RDFG
__, __, __, rep_ex_specs, resume_rdf = (
self.configure_volume_replication(
array, volume, vol_dict['device_id'], extra_specs))
if resume_rdf:
self.rest.srdf_resume_replication(
array, rep_ex_specs['mgmt_sg_name'],
rep_ex_specs['rdf_group_no'], extra_specs)
return first_vol, rep_ex_specs, vol_dict
existing_devices = self.rest.get_volumes_in_storage_group(
array, storagegroup_name)
try:
is_first_volume, rep_extra_specs, volume_info_dict = (
_is_first_vol_in_replicated_sg())
if not is_first_volume:
self._validate_rdfg_status(array, extra_specs)
__, rep_extra_specs, rep_info_dict, __ = (
self.prepare_replication_details(extra_specs))
volume_info_dict = self.provision.create_volume_from_sg(
array, volume_name, storagegroup_name,
volume_size, extra_specs, rep_info_dict)
rep_vol_dict = deepcopy(volume_info_dict)
rep_vol_dict.update({'device_uuid': volume_name,
'storage_group': storagegroup_name,
'size': volume_size})
remote_device_id = self.get_and_set_remote_device_uuid(
extra_specs, rep_extra_specs, rep_vol_dict)
rep_vol_dict.update({'remote_device_id': remote_device_id})
rep_update, rep_info_dict = self.gather_replication_updates(
extra_specs, rep_extra_specs, rep_vol_dict)
if rep_mode in [utils.REP_ASYNC, utils.REP_METRO]:
self._add_volume_to_rdf_management_group(
array, volume_info_dict['device_id'], volume_name,
rep_extra_specs['array'], remote_device_id,
extra_specs)
return volume_info_dict, rep_update, rep_info_dict
except Exception as e:
try:
# Attempt cleanup of rdfg & storage group post exception
updated_devices = set(self.rest.get_volumes_in_storage_group(
array, storagegroup_name))
devices_to_delete = [device for device in updated_devices
if device not in existing_devices]
if devices_to_delete:
self._cleanup_rdf_volume_create_post_failure(
volume, volume_name, extra_specs, devices_to_delete)
elif not existing_devices:
self.rest.delete_storage_group(array, storagegroup_name)
finally:
# Pass actual exception that was raised now that cleanup
# attempt is finished. Mainly VolumeBackendAPIException raised
# from error status codes returned from the various REST jobs.
raise e
def _set_vmax_extra_specs(self, extra_specs, pool_record):
"""Set the PowerMax/VMAX extra specs.
The pool_name extra spec must be set, otherwise a default slo/workload
will be chosen. The portgroup can either be passed as an extra spec
on the volume type (e.g. 'storagetype:portgroupname = os-pg1-pg'), or
can be chosen from a list provided in the cinder.conf
:param extra_specs: extra specifications
:param pool_record: pool record
:returns: dict -- the extra specifications dictionary
"""
# set extra_specs from pool_record
extra_specs[utils.SRP] = pool_record['srpName']
extra_specs[utils.ARRAY] = pool_record['SerialNumber']
try:
if not extra_specs.get(utils.PORTGROUPNAME):
extra_specs[utils.PORTGROUPNAME] = pool_record['PortGroup']
except Exception:
error_message = (_("Port group name has not been provided - "
"please configure the "
"'storagetype:portgroupname' extra spec on "
"the volume type, or enter a list of "
"portgroups in the cinder.conf associated with "
"this backend."))
LOG.error(error_message)
raise exception.VolumeBackendAPIException(message=error_message)
self._validate_storage_group_tag_list(extra_specs)
extra_specs[utils.INTERVAL] = self.interval
LOG.debug("The interval is set at: %(intervalInSecs)s.",
{'intervalInSecs': self.interval})
extra_specs[utils.RETRIES] = self.retries
LOG.debug("Retries are set at: %(retries)s.",
{'retries': self.retries})
# Set pool_name slo and workload
if 'pool_name' in extra_specs:
pool_name = extra_specs['pool_name']
pool_details = pool_name.split('+')
slo_from_extra_spec = pool_details[0]
workload_from_extra_spec = pool_details[1]
# Check if legacy pool chosen
if (workload_from_extra_spec == pool_record['srpName'] or
self.next_gen):
workload_from_extra_spec = 'NONE'
elif pool_record.get('ServiceLevel'):
slo_from_extra_spec = pool_record['ServiceLevel']
workload_from_extra_spec = pool_record.get('Workload', 'None')
# If workload is None in cinder.conf, convert to string
if not workload_from_extra_spec or self.next_gen:
workload_from_extra_spec = 'NONE'
LOG.info("Pool_name is not present in the extra_specs "
"- using slo/ workload from cinder.conf: %(slo)s/%(wl)s.",
{'slo': slo_from_extra_spec,
'wl': workload_from_extra_spec})
else:
slo_list = self.rest.get_slo_list(
pool_record['SerialNumber'], self.next_gen, self.array_model)
if 'Optimized' in slo_list:
slo_from_extra_spec = 'Optimized'
elif 'Diamond' in slo_list:
slo_from_extra_spec = 'Diamond'
else:
slo_from_extra_spec = 'None'
workload_from_extra_spec = 'NONE'
LOG.warning("Pool_name is not present in the extra_specs "
"so no slo/ workload information is present "
"using default slo/ workload combination: "
"%(slo)s/%(wl)s.",
{'slo': slo_from_extra_spec,
'wl': workload_from_extra_spec})
# Standardize slo and workload 'NONE' naming conventions
if workload_from_extra_spec.lower() == 'none':
workload_from_extra_spec = 'NONE'
if slo_from_extra_spec.lower() == 'none':
slo_from_extra_spec = None
extra_specs[utils.SLO] = slo_from_extra_spec
extra_specs[utils.WORKLOAD] = workload_from_extra_spec
if self.rest.is_compression_capable(extra_specs[utils.ARRAY]):
if extra_specs.get(utils.DISABLECOMPRESSION):
# If not True remove it.
if not strutils.bool_from_string(
extra_specs[utils.DISABLECOMPRESSION]):
extra_specs.pop(utils.DISABLECOMPRESSION, None)
else:
extra_specs.pop(utils.DISABLECOMPRESSION, None)
self._check_and_add_tags_to_storage_array(
extra_specs[utils.ARRAY], self.powermax_array_tag_list,
extra_specs)
LOG.debug("SRP is: %(srp)s, Array is: %(array)s "
"SLO is: %(slo)s, Workload is: %(workload)s.",
{'srp': extra_specs[utils.SRP],
'array': extra_specs[utils.ARRAY],
'slo': extra_specs[utils.SLO],
'workload': extra_specs[utils.WORKLOAD]})
if self.version_dict:
self.volume_metadata.print_pretty_table(self.version_dict)
else:
self.version_dict = (
self.volume_metadata.gather_version_info(
extra_specs[utils.ARRAY]))
return extra_specs
def _validate_storage_group_tag_list(self, extra_specs):
"""Validate the storagetype:storagegrouptags list
:param extra_specs: the extra specifications
:raises: VolumeBackendAPIException:
"""
tag_list = extra_specs.get(utils.STORAGE_GROUP_TAGS)
if tag_list:
if not self.utils.verify_tag_list(tag_list.split(',')):
exception_message = (_(
"Unable to get verify "
"storagetype:storagegrouptags in the Volume Type. "
"Only alpha-numeric, dashes and underscores "
"allowed. List values must be separated by commas. "
"The number of values must not exceed 8"))
raise exception.VolumeBackendAPIException(
message=exception_message)
else:
LOG.info("The tag list %(tag_list)s has been verified.",
{'tag_list': tag_list})
def _validate_array_tag_list(self, array_tag_list):
"""Validate the array tag list
:param array_tag_list: the array tag list
:raises: VolumeBackendAPIException:
"""
if array_tag_list:
if not self.utils.verify_tag_list(array_tag_list):
exception_message = (_(
"Unable to get verify "
"config option powermax_array_tag_list. "
"Only alpha-numeric, dashes and underscores "
"allowed. List values must be separated by commas. "
"The number of values must not exceed 8"))
raise exception.VolumeBackendAPIException(
message=exception_message)
else:
LOG.info("The tag list %(tag_list)s has been verified.",
{'tag_list': array_tag_list})
def _delete_from_srp(self, array, device_id, volume_name,
extra_specs):
"""Delete from srp.
:param array: the array serial number
:param device_id: the device id
:param volume_name: the volume name
:param extra_specs: the extra specifications
:raises: VolumeBackendAPIException:
"""
try:
LOG.debug("Delete Volume: %(name)s. device_id: %(device_id)s.",
{'name': volume_name, 'device_id': device_id})
self.provision.delete_volume_from_srp(
array, device_id, volume_name)
except Exception as e:
error_message = (_("Failed to delete volume %(volume_name)s. "
"Exception received: %(e)s") %
{'volume_name': volume_name,
'e': six.text_type(e)})
LOG.error(error_message)
raise exception.VolumeBackendAPIException(message=error_message)
def _remove_vol_and_cleanup_replication(
self, array, device_id, volume_name, extra_specs, volume):
"""Remove a volume from its storage groups and cleanup replication.
:param array: the array serial number
:param device_id: the device id
:param volume_name: the volume name
:param extra_specs: the extra specifications
:param volume: the volume object
"""
if volume and volume.migration_status == 'deleting':
extra_specs = self.utils.get_migration_delete_extra_specs(
volume, extra_specs, self.rep_configs)
# Cleanup remote replication
if self.utils.is_replication_enabled(extra_specs):
rdf_group_no, __ = self.get_rdf_details(
array, extra_specs[utils.REP_CONFIG])
self.cleanup_rdf_device_pair(array, rdf_group_no, device_id,
extra_specs)
else:
self.masking.remove_and_reset_members(
array, volume, device_id, volume_name, extra_specs, False)
@coordination.synchronized('emc-{rdf_group_no}-rdf')
def cleanup_rdf_device_pair(self, array, rdf_group_no, device_id,
extra_specs):
"""Cleanup replication on a RDF device pair, leave only source volume.
:param array: the array serial number
:param rdf_group_no: the rdf group number
:param device_id: the device id
:param extra_specs: the extra specifications
:raises: exception.VolumeBackendAPIException
"""
resume_replication, rdf_mgmt_cleanup = False, False
rdf_mgmt_sg, vols_in_mgmt_sg = None, None
rep_config = extra_specs[utils.REP_CONFIG]
rep_mode = extra_specs['rep_mode']
if rep_mode in [utils.REP_METRO, utils.REP_ASYNC]:
extra_specs['force_vol_remove'] = True
rdf_group_no, remote_array = self.get_rdf_details(array, rep_config)
rep_extra_specs = self._get_replication_extra_specs(
extra_specs, rep_config)
# 1. Get the remote device ID so it can be deleted later
remote_device = self.rest.get_rdf_pair_volume(
array, rdf_group_no, device_id)
remote_device_id = remote_device['remoteVolumeName']
vol_sg_list = self.rest.get_storage_groups_from_volume(
array, device_id)
# 2. If replication mode is async or metro, get RDF mgmt group info and
# suspend RDFG before proceeding to delete operation
if rep_mode in [utils.REP_METRO, utils.REP_ASYNC]:
# Make sure devices are in a valid state before continuing
self.rest.wait_for_rdf_pair_sync(
array, rdf_group_no, device_id, rep_extra_specs)
rdf_mgmt_sg = self.utils.get_rdf_management_group_name(rep_config)
vols_in_mgmt_sg = self.rest.get_num_vols_in_sg(array, rdf_mgmt_sg)
if vols_in_mgmt_sg > 1:
resume_replication = True
else:
rdf_mgmt_cleanup = True
self.rest.srdf_suspend_replication(
array, rdf_mgmt_sg, rdf_group_no, rep_extra_specs)
try:
# 3. Check vol doesnt live in any SGs outside OpenStack managed SGs
if rdf_mgmt_sg and rdf_mgmt_sg in vol_sg_list:
vol_sg_list.remove(rdf_mgmt_sg)
if len(vol_sg_list) > 1:
exception_message = (_(
"There is more than one storage group associated with "
"device %(dev)s not including RDF management groups. "
"Please check device is not member of non-OpenStack "
"managed storage groups") % {'dev': device_id})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(exception_message)
else:
vol_src_sg = vol_sg_list[0]
# 4. Remove device from SG and delete RDFG device pair
self.rest.srdf_remove_device_pair_from_storage_group(
array, vol_src_sg, rep_extra_specs['array'], device_id,
rep_extra_specs)
# 5. Remove the volume from any additional SGs
if rdf_mgmt_sg:
self.rest.remove_vol_from_sg(
array, rdf_mgmt_sg, device_id, extra_specs)
self.rest.remove_vol_from_sg(
remote_array, rdf_mgmt_sg, remote_device_id,
rep_extra_specs)
# 6. Delete the r2 volume
self.rest.delete_volume(remote_array, remote_device_id)
# 7. Delete the SGs if there are no volumes remaining
self._cleanup_rdf_storage_groups_post_r2_delete(
array, remote_array, vol_src_sg, rdf_mgmt_sg, rdf_mgmt_cleanup)
# 8. Resume replication if RDFG still contains volumes
if resume_replication:
self.rest.srdf_resume_replication(
array, rdf_mgmt_sg, rep_extra_specs['rdf_group_no'],
rep_extra_specs)
LOG.info('Remote device %(dev)s deleted from RDF Group %(grp)s',
{'dev': remote_device_id,
'grp': rep_extra_specs['rdf_group_label']})
except Exception as e:
# Attempt to resume SRDF groups after exception to avoid leaving
# them in a suspended state.
try:
if rdf_mgmt_sg:
self.rest.srdf_resume_replication(
array, rdf_mgmt_sg, rdf_group_no, rep_extra_specs,
False)
elif len(vol_sg_list) == 1:
self.rest.srdf_resume_replication(
array, vol_sg_list[0], rdf_group_no, rep_extra_specs,
False)
except Exception:
LOG.debug('Could not resume SRDF group after exception '
'during cleanup_rdf_device_pair.')
raise e
def _cleanup_rdf_storage_groups_post_r2_delete(
self, array, remote_array, sg_name, rdf_mgmt_sg, rdf_mgmt_cleanup):
"""Cleanup storage groups after a RDF device pair has been deleted.
:param array: the array serial number
:param remote_array: the remote array serial number
:param sg_name: the storage group name
:param rdf_mgmt_sg: the RDF managment group name
:param rdf_mgmt_cleanup: is RDF management group cleanup required
"""
vols_in_sg = self.rest.get_num_vols_in_sg(array, sg_name)
vols_in_remote_sg = self.rest.get_num_vols_in_sg(remote_array, sg_name)
if not vols_in_sg:
parent_sg = self.masking.get_parent_sg_from_child(
array, sg_name)
self.rest.delete_storage_group(array, sg_name)
if not vols_in_remote_sg:
self.rest.delete_storage_group(remote_array, sg_name)
if parent_sg:
vols_in_parent = self.rest.get_num_vols_in_sg(
array, parent_sg)
if not vols_in_parent:
mv_name = self.rest.get_masking_views_from_storage_group(
array, parent_sg)
if mv_name:
self.rest.delete_masking_view(array, mv_name)
if sg_name != parent_sg:
self.rest.delete_storage_group(array, parent_sg)
self.rest.delete_storage_group(remote_array,
parent_sg)
if rdf_mgmt_cleanup:
self.rest.delete_storage_group(array, rdf_mgmt_sg)
self.rest.delete_storage_group(remote_array, rdf_mgmt_sg)
def get_target_wwns_from_masking_view(
self, volume, connector):
"""Find target WWNs via the masking view.
:param volume: volume to be attached
:param connector: the connector dict
:returns: list -- the target WWN list
"""
metro_wwns = []
host = connector['host']
short_host_name = self.utils.get_host_name_label(
host, self.powermax_short_host_name_template) if host else None
extra_specs = self._initial_setup(volume)
if self.utils.is_volume_failed_over(volume):
rep_extra_specs = self._get_replication_extra_specs(
extra_specs, extra_specs[utils.REP_CONFIG])
extra_specs = rep_extra_specs
device_id = self._find_device_on_array(volume, extra_specs)
target_wwns = self._get_target_wwns_from_masking_view(
device_id, short_host_name, extra_specs)
if extra_specs.get(utils.REP_CONFIG) and self.utils.is_metro_device(
extra_specs[utils.REP_CONFIG], extra_specs):
rdf_group_no, __ = self.get_rdf_details(
extra_specs[utils.ARRAY], extra_specs[utils.REP_CONFIG])
rdf_pair_info = self.rest.get_rdf_pair_volume(
extra_specs[utils.ARRAY], rdf_group_no, device_id)
remote_device_id = rdf_pair_info.get('remoteVolumeName', None)
rep_extra_specs = self._get_replication_extra_specs(
extra_specs, extra_specs[utils.REP_CONFIG])
metro_wwns = self._get_target_wwns_from_masking_view(
remote_device_id, short_host_name, rep_extra_specs)
return target_wwns, metro_wwns
def _get_target_wwns_from_masking_view(
self, device_id, short_host_name, extra_specs):
"""Helper function to get wwns from a masking view.
:param device_id: the device id
:param short_host_name: the short host name
:param extra_specs: the extra specs
:returns: target wwns -- list
"""
target_wwns = []
array = extra_specs[utils.ARRAY]
masking_view_list, __ = self._get_masking_views_from_volume(
array, device_id, short_host_name)
if masking_view_list:
portgroup = self.get_port_group_from_masking_view(
array, masking_view_list[0])
target_wwns = self.rest.get_target_wwns(array, portgroup)
LOG.info("Target wwns in masking view %(maskingView)s: "
"%(targetWwns)s.",
{'maskingView': masking_view_list[0],
'targetWwns': target_wwns})
return target_wwns
def get_port_group_from_masking_view(self, array, maskingview_name):
"""Get the port groups in a masking view.
:param array: the array serial number
:param maskingview_name: masking view name
:returns: port group name
"""
return self.rest.get_element_from_masking_view(
array, maskingview_name, portgroup=True)
def get_initiator_group_from_masking_view(self, array, maskingview_name):
"""Get the initiator group in a masking view.
:param array: the array serial number
:param maskingview_name: masking view name
:returns: initiator group name
"""
return self.rest.get_element_from_masking_view(
array, maskingview_name, host=True)
def get_common_masking_views(self, array, portgroup_name,
initiator_group_name):
"""Get common masking views, if any.
:param array: the array serial number
:param portgroup_name: port group name
:param initiator_group_name: ig name
:returns: list of masking views
"""
LOG.debug("Finding Masking Views for port group %(pg)s and %(ig)s.",
{'pg': portgroup_name, 'ig': initiator_group_name})
masking_view_list = self.rest.get_common_masking_views(
array, portgroup_name, initiator_group_name)
return masking_view_list
def _get_ip_and_iqn(self, array, port):
"""Get ip and iqn from the director port.
:param array: the array serial number
:param port: the director port on the array
:returns: ip_and_iqn - dict
"""
ip_iqn_list = []
ip_addresses, iqn = self.rest.get_iscsi_ip_address_and_iqn(
array, port)
for ip in ip_addresses:
ip_iqn_list.append({'iqn': iqn, 'ip': ip})
return ip_iqn_list
def _find_ip_and_iqns(self, array, port_group_name):
"""Find the list of ips and iqns for the ports in a portgroup.
:param array: the array serial number
:param port_group_name: the portgroup name
:returns: ip_and_iqn - list of dicts
"""
ips_and_iqns = []
LOG.debug("The portgroup name for iscsiadm is %(pg)s",
{'pg': port_group_name})
ports = self.rest.get_port_ids(array, port_group_name)
for port in ports:
ip_and_iqn = self._get_ip_and_iqn(array, port)
ips_and_iqns.extend(ip_and_iqn)
return ips_and_iqns
def _create_replica(
self, array, clone_volume, source_device_id,
extra_specs, snap_name=None):
"""Create a replica.
Create replica for source volume, source can be volume or snapshot.
:param array: the array serial number
:param clone_volume: the clone volume object
:param source_device_id: the device ID of the volume
:param extra_specs: extra specifications
:param snap_name: the snapshot name - optional
:returns: int -- return code
:returns: dict -- cloneDict
"""
clone_id, target_device_id = clone_volume.id, None
clone_name = self.utils.get_volume_element_name(clone_id)
create_snap, copy_mode, rep_extra_specs = False, False, dict()
volume_dict = self.rest.get_volume(array, source_device_id)
replication_enabled = self.utils.is_replication_enabled(extra_specs)
if replication_enabled:
copy_mode = True
__, rep_extra_specs, __, __ = (
self.prepare_replication_details(extra_specs))
# PowerMax/VMAX supports using a target volume that is bigger than
# the source volume, so we create the target volume the desired
# size at this point to avoid having to extend later
try:
clone_dict, rep_update, rep_info_dict = self._create_volume(
clone_volume, clone_name, clone_volume.size, extra_specs)
target_device_id = clone_dict['device_id']
if target_device_id:
clone_volume_dict = self.rest.get_volume(
array, target_device_id)
self.utils.compare_cylinders(
volume_dict['cap_cyl'], clone_volume_dict['cap_cyl'])
LOG.info("The target device id is: %(device_id)s.",
{'device_id': target_device_id})
if not snap_name:
snap_name = self.utils.get_temp_snap_name(source_device_id)
create_snap = True
if replication_enabled:
if rep_extra_specs[utils.REP_CONFIG]['mode'] in (
[utils.REP_ASYNC, utils.REP_METRO]):
rep_extra_specs['sg_name'] = (
self.utils.get_rdf_management_group_name(
rep_extra_specs[utils.REP_CONFIG]))
self.rest.srdf_suspend_replication(
array, rep_extra_specs['sg_name'],
rep_extra_specs['rdf_group_no'], rep_extra_specs)
self.provision.create_volume_replica(
array, source_device_id, target_device_id,
snap_name, extra_specs, create_snap, copy_mode)
if replication_enabled:
self.rest.rdf_resume_with_retries(array, rep_extra_specs)
except Exception as e:
if target_device_id:
LOG.warning("Create replica failed. Cleaning up the target "
"volume. Clone name: %(cloneName)s, Error "
"received is %(e)s.",
{'cloneName': clone_name, 'e': e})
self._cleanup_target(
array, target_device_id, source_device_id,
clone_name, snap_name, extra_specs,
target_volume=clone_volume)
# Re-throw the exception.
raise
# add source id and snap_name to the clone dict
clone_dict['source_device_id'] = source_device_id
clone_dict['snap_name'] = snap_name
return clone_dict, rep_update, rep_info_dict
def _cleanup_target(
self, array, target_device_id, source_device_id,
clone_name, snap_name, extra_specs, generation=0,
target_volume=None):
"""Cleanup target volume on failed clone/ snapshot creation.
:param array: the array serial number
:param target_device_id: the target device ID
:param source_device_id: the source device ID
:param clone_name: the name of the clone volume
:param extra_specs: the extra specifications
:param generation: the generation number of the snapshot
:param target_volume: the target volume object
"""
snap_session = self.rest.get_sync_session(
array, source_device_id, snap_name, target_device_id, generation)
if snap_session:
self.provision.unlink_snapvx_tgt_volume(
array, target_device_id, source_device_id,
snap_name, extra_specs, generation)
self._remove_vol_and_cleanup_replication(
array, target_device_id, clone_name, extra_specs, target_volume)
self._delete_from_srp(
array, target_device_id, clone_name, extra_specs)
@retry(retry_exc_tuple, interval=1, retries=3)
def _sync_check(self, array, device_id, extra_specs,
tgt_only=False, source_device_id=None):
"""Check if volume is part of a SnapVx sync process.
:param array: the array serial number
:param device_id: volume instance
:param tgt_only: Flag - return only sessions where device is target
:param extra_specs: extra specifications
:param tgt_only: Flag to specify if it is a target
:param source_device_id: source_device_id if it has one
"""
if not source_device_id and tgt_only:
source_device_id = self._get_target_source_device(
array, device_id)
if source_device_id:
@coordination.synchronized("emc-source-{src_device_id}")
def do_unlink_and_delete_snap(src_device_id):
# Check if source device exists on the array
try:
self.rest.get_volume(array, src_device_id)
except exception.VolumeBackendAPIException:
LOG.debug("Device %(device_id)s not found on array, no "
"sync check required.",
{'device_id': src_device_id})
return
self._do_sync_check(
array, src_device_id, extra_specs, tgt_only)
do_unlink_and_delete_snap(source_device_id)
else:
self._do_sync_check(
array, device_id, extra_specs, tgt_only)
def _do_sync_check(
self, array, device_id, extra_specs, tgt_only=False):
"""Check if volume is part of a SnapVx sync process.
:param array: the array serial number
:param device_id: volume instance
:param tgt_only: Flag - return only sessions where device is target
:param extra_specs: extra specifications
:param tgt_only: Flag to specify if it is a target
"""
snapvx_tgt, snapvx_src, __ = self.rest.is_vol_in_rep_session(
array, device_id)
get_sessions = True if snapvx_tgt or snapvx_src else False
if get_sessions:
src_sessions, tgt_session = self.rest.find_snap_vx_sessions(
array, device_id, tgt_only)
if tgt_session:
self._unlink_targets_and_delete_temp_snapvx(
tgt_session, array, extra_specs)
if src_sessions and not tgt_only:
src_sessions.sort(key=lambda k: k['generation'], reverse=True)
for session in src_sessions:
self._unlink_targets_and_delete_temp_snapvx(
session, array, extra_specs)
def _unlink_targets_and_delete_temp_snapvx(
self, session, array, extra_specs):
"""Unlink target and delete the temporary snapvx if valid candidate.
:param session: the snapvx session
:param array: the array serial number
:param extra_specs: extra specifications
"""
snap_name = session['snap_name']
source = session['source_vol_id']
generation = session['generation']
expired = session['expired']
target, cm_enabled = None, False
if session.get('target_vol_id'):
target = session['target_vol_id']
cm_enabled = session['copy_mode']
if target:
loop = True if cm_enabled else False
LOG.debug(
"Unlinking source from target. Source: %(vol)s, Target: "
"%(tgt)s, Generation: %(gen)s.", {'vol': source, 'tgt': target,
'gen': generation})
self.provision.unlink_snapvx_tgt_volume(
array, target, source, snap_name, extra_specs, generation,
loop)
# Candidates for deletion:
# 1. If legacy snapshot with 'EMC_SMI' in snapshot name
# 2. If snapVX snapshot with copy mode enabled
# 3. If snapVX snapshot with copy mode disabled and not expired
if ('EMC_SMI' in snap_name or cm_enabled or (
not cm_enabled and not expired)):
LOG.debug(
"Deleting temporary snapshot. Source: %(vol)s, snap name: "
"%(name)s, generation: %(gen)s.", {
'vol': source, 'name': snap_name, 'gen': generation})
self.provision.delete_temp_volume_snap(
array, snap_name, source, generation)
def _get_target_source_device(self, array, device_id):
"""Get the source device id of the target.
:param array: the array serial number
:param device_id: volume instance
return source_device_id
"""
LOG.debug("Getting the source device ID for target device %(tgt)s",
{'tgt': device_id})
source_device_id = None
snapvx_tgt, __, __ = self.rest.is_vol_in_rep_session(
array, device_id)
if snapvx_tgt:
__, tgt_session = self.rest.find_snap_vx_sessions(
array, device_id, tgt_only=True)
source_device_id = tgt_session['source_vol_id']
LOG.debug("Target %(tgt)s source device %(src)s",
{'tgt': device_id, 'src': source_device_id})
return source_device_id
def _clone_check(
self, array, device_id, extra_specs, force_unlink=False):
"""Perform any snapvx cleanup before creating clones or snapshots
:param array: the array serial
:param device_id: the device ID of the volume
:param extra_specs: extra specifications
:param force_unlink: force unlink even if not expired
"""
snapvx_tgt, snapvx_src, __ = self.rest.is_vol_in_rep_session(
array, device_id)
if snapvx_src or snapvx_tgt:
@coordination.synchronized("emc-source-{src_device_id}")
def do_unlink_and_delete_snap(src_device_id):
src_sessions, tgt_session = self.rest.find_snap_vx_sessions(
array, src_device_id)
count = 0
if tgt_session and count < self.snapvx_unlink_limit:
self._delete_valid_snapshot(
array, tgt_session, extra_specs, force_unlink)
count += 1
if src_sessions:
src_sessions.sort(
key=lambda k: k['generation'], reverse=True)
for session in src_sessions:
if count < self.snapvx_unlink_limit:
self._delete_valid_snapshot(
array, session, extra_specs, force_unlink)
count += 1
else:
break
do_unlink_and_delete_snap(device_id)
def _delete_valid_snapshot(
self, array, session, extra_specs, force_unlink=False):
"""Delete a snapshot if valid candidate for deletion.
:param array: the array serial
:param session: the snapvx session
:param extra_specs: extra specifications
:param force_unlink: force unlink even if not expired
"""
is_legacy = 'EMC_SMI' in session['snap_name']
is_temp = utils.CLONE_SNAPSHOT_NAME in session['snap_name']
is_expired = session['expired']
if force_unlink:
is_expired = True
is_valid = True if is_legacy or (is_temp and is_expired) else False
if is_valid:
try:
self._unlink_targets_and_delete_temp_snapvx(
session, array, extra_specs)
except exception.VolumeBackendAPIException as e:
# Ignore and continue as snapshot has been unlinked
# successfully with incorrect status code returned
if ('404' and session['snap_name'] and
'does not exist' in six.text_type(e)):
pass
def manage_existing(self, volume, external_ref):
"""Manages an existing PowerMax/VMAX Volume (import to Cinder).
Renames the existing volume to match the expected name for the volume.
Also need to consider things like QoS, Emulation, account/tenant.
:param volume: the volume object including the volume_type_id
:param external_ref: reference to the existing volume
:returns: dict -- model_update
"""
LOG.info("Beginning manage existing volume process")
rep_info_dict, resume_rdf, rep_status = dict(), False, None
rep_model_update, rep_driver_data = dict(), dict()
rep_extra_specs = dict()
extra_specs = self._initial_setup(volume)
array, device_id = self.utils.get_array_and_device_id(
volume, external_ref)
volume_id = volume.id
# Check if the existing volume is valid for cinder management
orig_vol_name, src_sg = self._check_lun_valid_for_cinder_management(
array, device_id, volume_id, external_ref)
# If volume name is not present, then assign the device id as the name
if not orig_vol_name:
orig_vol_name = device_id
LOG.debug("Original volume name %(vol)s and source sg: %(sg_name)s.",
{'vol': orig_vol_name, 'sg_name': src_sg})
# Rename the volume
volume_name = self.utils.get_volume_element_name(volume_id)
LOG.debug("Rename volume %(vol)s to %(element_name)s.",
{'vol': orig_vol_name,
'element_name': volume_name})
self.rest.rename_volume(array, device_id, volume_name)
provider_location = {'device_id': device_id, 'array': array}
model_update = {'provider_location': six.text_type(provider_location)}
# Set-up volume replication, if enabled
if self.utils.is_replication_enabled(extra_specs):
(rep_status, rep_driver_data, rep_info_dict,
rep_extra_specs, resume_rdf) = (
self.configure_volume_replication(
array, volume, device_id, extra_specs))
if rep_driver_data:
rep_model_update = {
'replication_status': rep_status,
'replication_driver_data': six.text_type(
{'device_id': rep_info_dict['target_device_id'],
'array': rep_info_dict['remote_array']})}
try:
# Add/move volume to default storage group
self.masking.add_volume_to_default_storage_group(
array, device_id, volume_name, extra_specs, src_sg=src_sg)
if rep_status == 'first_vol_in_rdf_group':
rep_status, rep_driver_data, rep_info_dict = (
self._protect_storage_group(
array, device_id, volume, volume_name,
rep_extra_specs))
except Exception as e:
exception_message = (_(
"Unable to move the volume to the default SG. "
"Exception received was %(e)s") % {'e': six.text_type(e)})
LOG.error(exception_message)
LOG.debug("Rename volume %(vol)s back to %(element_name)s.",
{'vol': volume_id, 'element_name': orig_vol_name})
self.rest.rename_volume(array, device_id, orig_vol_name)
raise exception.VolumeBackendAPIException(
message=exception_message)
if resume_rdf:
self.rest.srdf_resume_replication(
array, rep_extra_specs['mgmt_sg_name'],
rep_extra_specs['rdf_group_no'], extra_specs)
if rep_driver_data:
rep_model_update = {
'replication_status': rep_status,
'replication_driver_data': six.text_type(
{'device_id': rep_info_dict['target_device_id'],
'array': rep_info_dict['remote_array']})}
model_update.update(rep_model_update)
model_update = self.update_metadata(
model_update, volume.metadata, self.get_volume_metadata(
array, device_id))
if rep_model_update:
target_backend_id = extra_specs.get(
utils.REPLICATION_DEVICE_BACKEND_ID, 'None')
model_update['metadata']['BackendID'] = target_backend_id
self.volume_metadata.capture_manage_existing(
volume, rep_info_dict, device_id, extra_specs)
return model_update
def _protect_storage_group(
self, array, device_id, volume, volume_name, rep_extra_specs):
"""Enable RDF on a volume after it has been managed into OpenStack.
:param array: the array serial number
:param device_id: the device id
:param volume: the volume object
:param volume_name: the volume name
:param rep_extra_specs: replication information dictionary
:returns: replication status, device pair info, replication info --
str, dict, dict
"""
rdf_group_no = rep_extra_specs['rdf_group_no']
remote_array = rep_extra_specs['array']
rep_mode = rep_extra_specs['rep_mode']
rep_config = rep_extra_specs[utils.REP_CONFIG]
if rep_mode in [utils.REP_ASYNC, utils.REP_METRO]:
rep_extra_specs['mgmt_sg_name'] = (
self.utils.get_rdf_management_group_name(rep_config))
else:
rep_extra_specs['mgmt_sg_name'] = None
sg_list = self.rest.get_storage_groups_from_volume(array, device_id)
if len(sg_list) == 1:
sg_name = sg_list[0]
elif len(sg_list) > 1:
exception_message = (_(
"Unable to RDF protect device %(dev)s in OpenStack managed "
"storage group because it currently exists in one or more "
"user managed storage groups.") % {'dev': device_id})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(exception_message)
rep_status, pair_info, r2_device_id = (
self._post_retype_srdf_protect_storage_group(
array, sg_name, device_id, volume_name, rep_extra_specs))
target_name = self.utils.get_volume_element_name(volume.id)
rep_info_dict = self.volume_metadata.gather_replication_info(
volume.id, 'replication', False,
rdf_group_no=rdf_group_no, target_name=target_name,
remote_array=remote_array, target_device_id=r2_device_id,
replication_status=rep_status, rep_mode=rep_mode,
rdf_group_label=rep_config['rdf_group_label'],
target_array_model=rep_extra_specs['target_array_model'],
mgmt_sg_name=rep_extra_specs['mgmt_sg_name'])
return rep_status, pair_info, rep_info_dict
def _check_lun_valid_for_cinder_management(
self, array, device_id, volume_id, external_ref):
"""Check if a volume is valid for cinder management.
:param array: the array serial number
:param device_id: the device id
:param volume_id: the cinder volume id
:param external_ref: the external reference
:returns volume_identifier - name of the volume on PowerMax/VMAX
:returns sg - the storage group which the LUN belongs to
:raises: ManageExistingInvalidReference, ManageExistingAlreadyManaged:
"""
# Ensure the volume exists on the array
volume_details = self.rest.get_volume(array, device_id)
if not volume_details:
msg = (_('Unable to retrieve volume details from array for '
'device %(device_id)s') % {'device_id': device_id})
raise exception.ManageExistingInvalidReference(
existing_ref=external_ref, reason=msg)
# Check if volume is FBA emulation
fba_devices = self.rest.get_volume_list(array, "emulation=FBA")
if device_id not in fba_devices:
msg = (_("Unable to import volume %(device_id)s to cinder as it "
"is not an FBA volume. Only volumes with an emulation "
"type of FBA are supported.")
% {'device_id': device_id})
raise exception.ManageExistingVolumeTypeMismatch(reason=msg)
volume_identifier = None
# Check if volume is already cinder managed
if volume_details.get('volume_identifier'):
volume_identifier = volume_details['volume_identifier']
if volume_identifier.startswith(utils.VOLUME_ELEMENT_NAME_PREFIX):
raise exception.ManageExistingAlreadyManaged(
volume_ref=volume_id)
# Check if the volume is part of multiple SGs and
# check if the volume is attached by checking if in any masking view.
storagegrouplist = self.rest.get_storage_groups_from_volume(
array, device_id)
if storagegrouplist and len(storagegrouplist) > 1:
msg = (_("Unable to import volume %(device_id)s to cinder. "
"Volume is in multiple SGs.")
% {'device_id': device_id})
raise exception.ManageExistingInvalidReference(
existing_ref=external_ref, reason=msg)
sg = None
if storagegrouplist:
sg = storagegrouplist[0]
mvs = self.rest.get_masking_views_from_storage_group(
array, sg)
if mvs:
msg = (_("Unable to import volume %(device_id)s to cinder. "
"Volume is in masking view(s): %(mv)s.")
% {'device_id': device_id, 'mv': mvs})
raise exception.ManageExistingInvalidReference(
existing_ref=external_ref, reason=msg)
# Check if there are any replication sessions associated
# with the volume.
snapvx_tgt, __, rdf = self.rest.is_vol_in_rep_session(
array, device_id)
if snapvx_tgt or rdf:
msg = (_("Unable to import volume %(device_id)s to cinder. "
"It is part of a replication session.")
% {'device_id': device_id})
raise exception.ManageExistingInvalidReference(
existing_ref=external_ref, reason=msg)
return volume_identifier, sg
def manage_existing_get_size(self, volume, external_ref):
"""Return size of an existing PowerMax/VMAX volume to manage_existing.
:param self: reference to class
:param volume: the volume object including the volume_type_id
:param external_ref: reference to the existing volume
:returns: size of the volume in GB
"""
LOG.debug("Volume in manage_existing_get_size: %(volume)s.",
{'volume': volume})
array, device_id = self.utils.get_array_and_device_id(
volume, external_ref)
# Ensure the volume exists on the array
volume_details = self.rest.get_volume(array, device_id)
if not volume_details:
msg = (_('Unable to retrieve volume details from array for '
'device %(device_id)s') % {'device_id': device_id})
raise exception.ManageExistingInvalidReference(
existing_ref=external_ref, reason=msg)
size = float(self.rest.get_size_of_device_on_array(array, device_id))
if not size.is_integer():
exception_message = (
_("Cannot manage existing PowerMax/VMAX volume %(device_id)s "
"- it has a size of %(vol_size)s but only whole GB "
"sizes are supported. Please extend the "
"volume to the nearest GB value before importing.")
% {'device_id': device_id, 'vol_size': size, })
LOG.error(exception_message)
raise exception.ManageExistingInvalidReference(
existing_ref=external_ref, reason=exception_message)
LOG.debug("Size of volume %(device_id)s is %(vol_size)s GB.",
{'device_id': device_id, 'vol_size': int(size)})
return int(size)
def unmanage(self, volume):
"""Export PowerMax/VMAX volume from Cinder.
Leave the volume intact on the backend array.
:param volume: the volume object
"""
volume_name = volume.name
volume_id = volume.id
LOG.info("Unmanage volume %(name)s, id=%(id)s",
{'name': volume_name, 'id': volume_id})
extra_specs = self._initial_setup(volume)
device_id = self._find_device_on_array(volume, extra_specs)
if device_id is None:
LOG.error("Cannot find Volume: %(id)s for "
"unmanage operation. Exiting...",
{'id': volume_id})
else:
# Check if volume is snap source
self._clone_check(extra_specs['array'], device_id, extra_specs)
# Remove volume from any openstack storage groups
# and remove any replication
self._remove_vol_and_cleanup_replication(
extra_specs['array'], device_id,
volume_name, extra_specs, volume)
# Rename the volume to volumeId, thus remove the 'OS-' prefix.
self.rest.rename_volume(
extra_specs[utils.ARRAY], device_id, volume_id)
# First check/create the unmanaged sg
# Don't fail if we fail to create the SG
try:
self.provision.create_storage_group(
extra_specs[utils.ARRAY], utils.UNMANAGED_SG,
extra_specs[utils.SRP], None,
None, extra_specs=extra_specs)
except Exception as e:
msg = ("Exception creating %(sg)s. "
"Exception received was %(e)s."
% {'sg': utils.UNMANAGED_SG,
'e': six.text_type(e)})
LOG.warning(msg)
return
# Try to add the volume
self.masking._check_adding_volume_to_storage_group(
extra_specs[utils.ARRAY], device_id, utils.UNMANAGED_SG,
volume_id, extra_specs)
def manage_existing_snapshot(self, snapshot, existing_ref):
"""Manage an existing PowerMax/VMAX Snapshot (import to Cinder).
Renames the Snapshot to prefix it with OS- to indicate
it is managed by Cinder
:param snapshot: the snapshot object
:param existing_ref: the snapshot name on the backend VMAX
:raises: VolumeBackendAPIException
:returns: model update
"""
volume = snapshot.volume
extra_specs = self._initial_setup(volume)
array = extra_specs[utils.ARRAY]
device_id = self._find_device_on_array(volume, extra_specs)
try:
snap_name = existing_ref['source-name']
except KeyError:
snap_name = existing_ref['source-id']
if snapshot.display_name:
snap_display_name = snapshot.display_name
else:
snap_display_name = snapshot.id
if snap_name.startswith(utils.VOLUME_ELEMENT_NAME_PREFIX):
exception_message = (
_("Unable to manage existing Snapshot. Snapshot "
"%(snapshot)s is already managed by Cinder.") %
{'snapshot': snap_name})
raise exception.VolumeBackendAPIException(
message=exception_message)
if self.utils.is_volume_failed_over(volume):
exception_message = (
(_("Volume %(name)s is failed over from the source volume, "
"it is not possible to manage a snapshot of a failed over "
"volume.") % {'name': volume.id}))
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
if not self.rest.get_volume_snap(array, device_id, snap_name):
exception_message = (
_("Snapshot %(snap_name)s is not associated with specified "
"volume %(device_id)s, it is not possible to manage a "
"snapshot that is not associated with the specified "
"volume.")
% {'device_id': device_id, 'snap_name': snap_name})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
snap_backend_name = self.utils.modify_snapshot_prefix(
snap_name, manage=True)
try:
self.rest.modify_volume_snap(
array, device_id, device_id, snap_name,
extra_specs, rename=True, new_snap_name=snap_backend_name)
except Exception as e:
exception_message = (
_("There was an issue managing %(snap_name)s, it was not "
"possible to add the OS- prefix. Error Message: %(e)s.")
% {'snap_name': snap_name, 'e': six.text_type(e)})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
prov_loc = {'source_id': device_id, 'snap_name': snap_backend_name}
model_update = {
'display_name': snap_display_name,
'provider_location': six.text_type(prov_loc)}
model_update = self.update_metadata(
model_update, snapshot.metadata, self.get_snapshot_metadata(
array, device_id, snap_backend_name))
LOG.info("Managing SnapVX Snapshot %(snap_name)s of source "
"volume %(device_id)s, OpenStack Snapshot display name: "
"%(snap_display_name)s", {
'snap_name': snap_name, 'device_id': device_id,
'snap_display_name': snap_display_name})
return model_update
def manage_existing_snapshot_get_size(self, snapshot):
"""Return the size of the source volume for manage-existing-snapshot.
:param snapshot: the snapshot object
:returns: size of the source volume in GB
"""
volume = snapshot.volume
extra_specs = self._initial_setup(volume)
device_id = self._find_device_on_array(volume, extra_specs)
return self.rest.get_size_of_device_on_array(
extra_specs[utils.ARRAY], device_id)
def unmanage_snapshot(self, snapshot):
"""Export PowerMax/VMAX Snapshot from Cinder.
Leaves the snapshot intact on the backend VMAX
:param snapshot: the snapshot object
:raises: VolumeBackendAPIException
"""
volume = snapshot.volume
extra_specs = self._initial_setup(volume)
array = extra_specs[utils.ARRAY]
device_id, snap_name = self._parse_snap_info(array, snapshot)
if self.utils.is_volume_failed_over(volume):
exception_message = (
_("It is not possible to unmanage a snapshot where the "
"source volume is failed-over, revert back to source "
"PowerMax/VMAX to unmanage snapshot %(snap_name)s")
% {'snap_name': snap_name})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
new_snap_backend_name = self.utils.modify_snapshot_prefix(
snap_name, unmanage=True)
try:
self.rest.modify_volume_snap(
array, device_id, device_id, snap_name, extra_specs,
rename=True, new_snap_name=new_snap_backend_name)
except Exception as e:
exception_message = (
_("There was an issue unmanaging Snapshot, it "
"was not possible to remove the OS- prefix. Error "
"message is: %(e)s.")
% {'snap_name': snap_name, 'e': six.text_type(e)})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
LOG.info("Snapshot %(snap_name)s is no longer managed in "
"OpenStack but still remains on PowerMax/VMAX source "
"%(array_id)s", {'snap_name': snap_name, 'array_id': array})
LOG.warning("In order to remove the snapshot source volume from "
"OpenStack you will need to either delete the linked "
"SnapVX snapshot on the array or un-manage the volume "
"from Cinder.")
def get_manageable_volumes(self, marker, limit, offset, sort_keys,
sort_dirs):
"""Lists all manageable volumes.
:param marker: Begin returning volumes that appear later in the volume
list than that represented by this reference. This
reference should be json like. Default=None.
:param limit: Maximum number of volumes to return. Default=None.
:param offset: Number of volumes to skip after marker. Default=None.
:param sort_keys: Key to sort by, sort by size or reference. Valid
keys: size, reference. Default=None.
:param sort_dirs: Direction to sort by. Valid dirs: asc, desc.
Default=None.
:returns: List of dicts containing all volumes valid for management
"""
valid_vols = []
manageable_vols = []
array = self.pool_info['arrays_info'][0]["SerialNumber"]
LOG.info("Listing manageable volumes for array %(array_id)s", {
'array_id': array})
volumes = self.rest.get_private_volume_list(array)
# No volumes returned from PowerMax/VMAX
if not volumes:
LOG.info("There were no volumes found on the backend "
"PowerMax/VMAX. You need to create some volumes before "
"they can be managed into Cinder.")
return manageable_vols
for device in volumes:
# Determine if volume is valid for management
if self.utils.is_volume_manageable(device):
valid_vols.append(device['volumeHeader'])
# For all valid vols, extract relevant data for Cinder response
for vol in valid_vols:
volume_dict = {'reference': {'source-id': vol['volumeId']},
'safe_to_manage': True,
'size': int(math.ceil(vol['capGB'])),
'reason_not_safe': None, 'cinder_id': None,
'extra_info': {
'config': vol['configuration'],
'emulation': vol['emulationType']}}
manageable_vols.append(volume_dict)
# If volume list is populated, perform filtering on user params
if manageable_vols:
# If sort keys selected, determine if by size or reference, and
# direction of sort
manageable_vols = self._sort_manageable_volumes(
manageable_vols, marker, limit, offset, sort_keys, sort_dirs)
return manageable_vols
def _sort_manageable_volumes(
self, manageable_vols, marker, limit, offset, sort_keys,
sort_dirs):
"""Sort manageable volumes.
:param manageable_vols: Unsort list of dicts
:param marker: Begin returning volumes that appear later in the volume
list than that represented by this reference. This
reference should be json like. Default=None.
:param limit: Maximum number of volumes to return. Default=None.
:param offset: Number of volumes to skip after marker. Default=None.
:param sort_keys: Key to sort by, sort by size or reference. Valid
keys: size, reference. Default=None.
:param sort_dirs: Direction to sort by. Valid dirs: asc, desc.
Default=None.
:returns: manageable_vols -Sorted list of dicts
"""
# If sort keys selected, determine if by size or reference, and
# direction of sort
if sort_keys:
reverse = False
if sort_dirs:
if 'desc' in sort_dirs[0]:
reverse = True
if sort_keys[0] == 'size':
manageable_vols = sorted(manageable_vols,
key=lambda k: k['size'],
reverse=reverse)
if sort_keys[0] == 'reference':
manageable_vols = sorted(manageable_vols,
key=lambda k: k['reference'][
'source-id'],
reverse=reverse)
# If marker provided, return only manageable volumes after marker
if marker:
vol_index = None
for vol in manageable_vols:
if vol['reference']['source-id'] == marker:
vol_index = manageable_vols.index(vol)
if vol_index:
manageable_vols = manageable_vols[vol_index:]
else:
msg = _("Volume marker not found, please check supplied "
"device ID and try again.")
raise exception.VolumeBackendAPIException(msg)
# If offset or limit provided, offset or limit result list
if offset:
manageable_vols = manageable_vols[offset:]
if limit:
manageable_vols = manageable_vols[:limit]
return manageable_vols
def get_manageable_snapshots(self, marker, limit, offset, sort_keys,
sort_dirs):
"""Lists all manageable snapshots.
:param marker: Begin returning volumes that appear later in the volume
list than that represented by this reference. This
reference should be json like. Default=None.
:param limit: Maximum number of volumes to return. Default=None.
:param offset: Number of volumes to skip after marker. Default=None.
:param sort_keys: Key to sort by, sort by size or reference.
Valid keys: size, reference. Default=None.
:param sort_dirs: Direction to sort by. Valid dirs: asc, desc.
Default=None.
:returns: List of dicts containing all snapshots valid for management
"""
manageable_snaps = []
array = self.pool_info['arrays_info'][0]["SerialNumber"]
LOG.info("Listing manageable snapshots for array %(array_id)s", {
'array_id': array})
volumes = self.rest.get_private_volume_list(array)
# No volumes returned from PowerMax/VMAX
if not volumes:
LOG.info("There were no volumes found on the backend "
"PowerMax/VMAX. You need to create some volumes "
"before snashot can be created and managed into Cinder.")
return manageable_snaps
for device in volumes:
# Determine if volume is valid for management
manageable_snaps = self._is_snapshot_valid_for_management(
manageable_snaps, device)
# If snapshot list is populated, perform filtering on user params
if len(manageable_snaps) > 0:
# Order snapshots by source deviceID and not snapshot name
manageable_snaps = self._sort_manageable_snapshots(
manageable_snaps, marker, limit, offset, sort_keys, sort_dirs)
return manageable_snaps
def _sort_manageable_snapshots(
self, manageable_snaps, marker, limit, offset, sort_keys,
sort_dirs):
"""Sorts manageable snapshots list.
:param manageable_snaps: unsorted manageable snapshot list
:param marker: Begin returning volumes that appear later in the volume
list than that represented by this reference. This
reference should be json like. Default=None.
:param limit: Maximum number of volumes to return. Default=None.
:param offset: Number of volumes to skip after marker. Default=None.
:param sort_keys: Key to sort by, sort by size or reference.
Valid keys: size, reference. Default=None.
:param sort_dirs: Direction to sort by. Valid dirs: asc, desc.
Default=None.
:returns: List of dicts containing all snapshots valid for management
"""
manageable_snaps = sorted(
manageable_snaps,
key=lambda k: k['source_reference']['source-id'])
# If sort keys selected, determine if by size or reference, and
# direction of sort
if sort_keys:
reverse = False
if sort_dirs:
if 'desc' in sort_dirs[0]:
reverse = True
if sort_keys[0] == 'size':
manageable_snaps = sorted(manageable_snaps,
key=lambda k: k['size'],
reverse=reverse)
if sort_keys[0] == 'reference':
manageable_snaps = sorted(manageable_snaps,
key=lambda k: k['reference'][
'source-name'],
reverse=reverse)
# If marker provided, return only manageable volumes after marker
if marker:
snap_index = None
for snap in manageable_snaps:
if snap['reference']['source-name'] == marker:
snap_index = manageable_snaps.index(snap)
if snap_index:
manageable_snaps = manageable_snaps[snap_index:]
else:
msg = (_("Snapshot marker %(marker)s not found, marker "
"provided must be a valid PowerMax/VMAX "
"snapshot ID") %
{'marker': marker})
raise exception.VolumeBackendAPIException(msg)
# If offset or limit provided, offset or limit result list
if offset:
manageable_snaps = manageable_snaps[offset:]
if limit:
manageable_snaps = manageable_snaps[:limit]
return manageable_snaps
def _is_snapshot_valid_for_management(self, manageable_snaps, device):
"""check if snapshot is valid for management
:param manageable_snaps: list of manageable snapshots
:param device: the source device
:returns: List of dicts containing all snapshots valid for management
"""
if self.utils.is_snapshot_manageable(device):
# Snapshot valid, extract relevant snap info
snap_info = device['timeFinderInfo']['snapVXSession'][0][
'srcSnapshotGenInfo'][0]['snapshotHeader']
# Convert timestamp to human readable format
human_timestamp = time.strftime(
"%Y/%m/%d, %H:%M:%S", time.localtime(
float(six.text_type(
snap_info['timestamp'])[:-3])))
# If TTL is set, convert value to human readable format
if int(snap_info['timeToLive']) > 0:
human_ttl_timestamp = time.strftime(
"%Y/%m/%d, %H:%M:%S", time.localtime(
float(six.text_type(
snap_info['timeToLive']))))
else:
human_ttl_timestamp = 'N/A'
# For all valid snaps, extract relevant data for Cinder
# response
snap_dict = {
'reference': {
'source-name': snap_info['snapshotName']},
'safe_to_manage': True,
'size': int(
math.ceil(device['volumeHeader']['capGB'])),
'reason_not_safe': None, 'cinder_id': None,
'extra_info': {
'generation': snap_info['generation'],
'secured': snap_info['secured'],
'timeToLive': human_ttl_timestamp,
'timestamp': human_timestamp},
'source_reference': {'source-id': snap_info['device']}}
manageable_snaps.append(snap_dict)
return manageable_snaps
def retype(self, volume, new_type, host):
"""Migrate volume to another host using retype.
:param volume: the volume object including the volume_type_id
:param new_type: the new volume type.
:param host: The host dict holding the relevant target(destination)
information
:returns: boolean -- True if retype succeeded, False if error
"""
volume_name = volume.name
LOG.info("Migrating Volume %(volume)s via retype.",
{'volume': volume_name})
extra_specs = self._initial_setup(volume)
if not self.utils.is_retype_supported(volume, extra_specs,
new_type['extra_specs'],
self.rep_configs):
src_mode = extra_specs.get('rep_mode', 'non-replicated')
LOG.error("It is not possible to perform host-assisted retype "
"from %(src_mode)s to Metro replication type whilst the "
"volume is attached to a host. To perform this "
"operation please first detach the volume.",
{'src_mode': src_mode})
return False
device_id = self._find_device_on_array(volume, extra_specs)
if device_id is None:
LOG.error("Volume %(name)s not found on the array. "
"No volume to migrate using retype.",
{'name': volume_name})
return False
return self._slo_workload_migration(device_id, volume, host,
volume_name, new_type, extra_specs)
def _slo_workload_migration(self, device_id, volume, host,
volume_name, new_type, extra_specs):
"""Migrate from SLO/Workload combination to another.
:param device_id: the volume device id
:param volume: the volume object
:param host: the host dict
:param volume_name: the name of the volume
:param new_type: the type to migrate to
:param extra_specs: extra specifications
:returns: boolean -- True if migration succeeded, False if error.
"""
# Check if old type and new type have different replication types
do_change_replication = self.utils.change_replication(
extra_specs, new_type[utils.EXTRA_SPECS])
is_compression_disabled = self.utils.is_compression_disabled(
extra_specs)
# Check if old type and new type have different compression types
do_change_compression = (self.utils.change_compression_type(
is_compression_disabled, new_type))
is_valid, target_slo, target_workload = (
self._is_valid_for_storage_assisted_migration(
device_id, host, extra_specs[utils.ARRAY],
extra_specs[utils.SRP], volume_name,
do_change_compression, do_change_replication))
if not is_valid:
# Check if this is multiattach retype case
do_change_multiattach = self.utils.change_multiattach(
extra_specs, new_type['extra_specs'])
if do_change_multiattach:
return True
else:
LOG.error(
"Volume %(name)s is not suitable for storage "
"assisted migration using retype.",
{'name': volume_name})
return False
if (volume.host != host['host'] or do_change_compression
or do_change_replication):
LOG.debug(
"Retype Volume %(name)s from source host %(sourceHost)s "
"to target host %(targetHost)s. Compression change is %(cc)r. "
"Replication change is %(rc)s.",
{'name': volume_name, 'sourceHost': volume.host,
'targetHost': host['host'],
'cc': do_change_compression, 'rc': do_change_replication})
return self._migrate_volume(
extra_specs[utils.ARRAY], volume, device_id,
extra_specs[utils.SRP], target_slo,
target_workload, volume_name, new_type, extra_specs)
return False
def _migrate_volume(
self, array, volume, device_id, srp, target_slo,
target_workload, volume_name, new_type, extra_specs):
"""Migrate from one slo/workload combination to another.
This requires moving the volume from its current SG to a
new or existing SG that has the target attributes.
:param array: the array serial number
:param volume: the volume object
:param device_id: the device number
:param srp: the storage resource pool
:param target_slo: the target service level
:param target_workload: the target workload
:param volume_name: the volume name
:param new_type: the volume type to migrate to
:param extra_specs: the extra specifications
:returns: bool
"""
model_update, success = None, False
rep_info_dict, rep_extra_specs, rep_mode, rep_status = (
None, None, None, False)
resume_target_sg, resume_original_sg = False, False
resume_original_sg_dict = dict()
target_extra_specs = new_type['extra_specs']
target_extra_specs.update({
utils.SRP: srp, utils.ARRAY: array, utils.SLO: target_slo,
utils.WORKLOAD: target_workload,
utils.INTERVAL: extra_specs[utils.INTERVAL],
utils.RETRIES: extra_specs[utils.RETRIES]})
compression_disabled = self.utils.is_compression_disabled(
target_extra_specs)
target_extra_specs.update(
{utils.DISABLECOMPRESSION: compression_disabled})
was_rep_enabled = self.utils.is_replication_enabled(extra_specs)
if self.utils.is_replication_enabled(target_extra_specs):
target_backend_id = target_extra_specs.get(
utils.REPLICATION_DEVICE_BACKEND_ID,
utils.BACKEND_ID_LEGACY_REP)
target_rep_config = self.utils.get_rep_config(
target_backend_id, self.rep_configs)
rep_mode = target_rep_config['mode']
target_extra_specs[utils.REP_MODE] = rep_mode
target_extra_specs[utils.REP_CONFIG] = target_rep_config
is_rep_enabled = True
else:
is_rep_enabled = False
backend_ids_differ = False
if was_rep_enabled and is_rep_enabled:
curr_backend_id = extra_specs.get(
utils.REPLICATION_DEVICE_BACKEND_ID,
utils.BACKEND_ID_LEGACY_REP)
tgt_backend_id = target_extra_specs.get(
utils.REPLICATION_DEVICE_BACKEND_ID,
utils.BACKEND_ID_LEGACY_REP)
backend_ids_differ = curr_backend_id != tgt_backend_id
if was_rep_enabled:
self._validate_rdfg_status(array, extra_specs)
if is_rep_enabled:
self._validate_rdfg_status(array, target_extra_specs)
# Scenario 1: Rep -> Non-Rep
# Scenario 2: Cleanup for Rep -> Diff Rep type
if (was_rep_enabled and not is_rep_enabled) or backend_ids_differ:
rep_extra_specs, resume_original_sg = (
self.break_rdf_device_pair_session(
array, device_id, volume_name, extra_specs))
model_update = {
'replication_status': REPLICATION_DISABLED,
'replication_driver_data': None}
if resume_original_sg:
resume_original_sg_dict = {
utils.ARRAY: array,
utils.SG_NAME: rep_extra_specs['mgmt_sg_name'],
utils.RDF_GROUP_NO: rep_extra_specs['rdf_group_no'],
utils.EXTRA_SPECS: rep_extra_specs}
# Scenario 1: Non-Rep -> Rep
# Scenario 2: Rep -> Diff Rep type
if (not was_rep_enabled and is_rep_enabled) or backend_ids_differ:
self._sync_check(array, device_id, extra_specs)
(rep_status, rep_driver_data, rep_info_dict,
rep_extra_specs, resume_target_sg) = (
self.configure_volume_replication(
array, volume, device_id, target_extra_specs))
model_update = {
'replication_status': rep_status,
'replication_driver_data': six.text_type(
{'device_id': rep_info_dict['target_device_id'],
'array': rep_info_dict['remote_array']})}
success, target_sg_name = self._retype_volume(
array, srp, device_id, volume, volume_name, extra_specs,
target_slo, target_workload, target_extra_specs)
# Volume is first volume in RDFG, SG needs to be protected
if rep_status == 'first_vol_in_rdf_group':
volume_name = self.utils.get_volume_element_name(volume.id)
rep_status, rdf_pair_info, tgt_device_id = (
self._post_retype_srdf_protect_storage_group(
array, target_sg_name, device_id, volume_name,
rep_extra_specs))
model_update = {
'replication_status': rep_status,
'replication_driver_data': six.text_type(
{'device_id': tgt_device_id,
'array': rdf_pair_info['remoteSymmetrixId']})}
# Scenario: Rep -> Same Rep
if was_rep_enabled and is_rep_enabled and not backend_ids_differ:
# No change in replication config, retype remote device
success = self._retype_remote_volume(
array, volume, device_id, volume_name,
rep_mode, is_rep_enabled, target_extra_specs)
if resume_target_sg:
self.rest.srdf_resume_replication(
array, rep_extra_specs['mgmt_sg_name'],
rep_extra_specs['rdf_group_no'], rep_extra_specs)
if resume_original_sg and resume_original_sg_dict:
self.rest.srdf_resume_replication(
resume_original_sg_dict[utils.ARRAY],
resume_original_sg_dict[utils.SG_NAME],
resume_original_sg_dict[utils.RDF_GROUP_NO],
resume_original_sg_dict[utils.EXTRA_SPECS])
if success:
model_update = self.update_metadata(
model_update, volume.metadata,
self.get_volume_metadata(array, device_id))
target_backend_id = None
if is_rep_enabled:
target_backend_id = target_extra_specs.get(
utils.REPLICATION_DEVICE_BACKEND_ID, 'None')
model_update['metadata']['BackendID'] = target_backend_id
if was_rep_enabled and not is_rep_enabled:
model_update = self.remove_stale_data(model_update)
self.volume_metadata.capture_retype_info(
volume, device_id, array, srp, target_slo,
target_workload, target_sg_name, is_rep_enabled, rep_mode,
target_extra_specs[utils.DISABLECOMPRESSION],
target_backend_id)
return success, model_update
def _retype_volume(
self, array, srp, device_id, volume, volume_name, extra_specs,
target_slo, target_workload, target_extra_specs, remote=False):
"""Retype a volume from one volume type to another.
The target storage group ID is returned so the next phase in the
calling function can SRDF protect it if required.
:param array: the array serial number
:param srp: the storage resource pool name
:param device_id: the device ID to be retyped
:param volume: the volume object
:param volume_name: the volume name
:param extra_specs: source extra specs
:param target_slo: target service level id
:param target_workload: target workload id
:param target_extra_specs: target extra specs
:param remote: if the volume being retyped is on a remote replication
target
:returns: retype success, target storage group -- bool, str
"""
is_re, rep_mode, mgmt_sg_name = False, None, None
parent_sg = None
if self.utils.is_replication_enabled(target_extra_specs):
is_re, rep_mode = True, target_extra_specs['rep_mode']
mgmt_sg_name = self.utils.get_rdf_management_group_name(
target_extra_specs[utils.REP_CONFIG])
device_info = self.rest.get_volume(array, device_id)
target_extra_specs[utils.PORTGROUPNAME] = (
extra_specs.get(utils.PORTGROUPNAME, None))
disable_compression = target_extra_specs[utils.DISABLECOMPRESSION]
source_sg_list = device_info['storageGroupId']
if mgmt_sg_name in source_sg_list:
source_sg_list.remove(mgmt_sg_name)
source_sg_name = source_sg_list[0]
# If volume is attached set up the parent/child SGs if not already
# present on array
if volume.attach_status == 'attached' and not remote:
attached_host = self.utils.get_volume_attached_hostname(
volume)
if not attached_host:
LOG.error(
"There was an issue retrieving attached host from volume "
"%(volume_name)s, aborting storage-assisted migration.",
{'volume_name': device_id})
return False, None
port_group_label = self.utils.get_port_name_label(
target_extra_specs[utils.PORTGROUPNAME],
self.powermax_port_group_name_template)
target_sg_name, __, __ = self.utils.get_child_sg_name(
attached_host, target_extra_specs, port_group_label)
target_sg = self.rest.get_storage_group(array, target_sg_name)
if not target_sg:
self.provision.create_storage_group(
array, target_sg_name, srp, target_slo, target_workload,
target_extra_specs, disable_compression)
source_sg = self.rest.get_storage_group(array, source_sg_name)
parent_sg = source_sg.get('parent_storage_group', None)
if parent_sg:
parent_sg = parent_sg[0]
self.masking.add_child_sg_to_parent_sg(
array, target_sg_name, parent_sg, target_extra_specs)
# Else volume is not attached or is remote volume, default SGs are used
else:
target_sg_name = (
self.masking.get_or_create_default_storage_group(
array, srp, target_slo, target_workload, extra_specs,
disable_compression, is_re, rep_mode))
# Move the volume from the source to target storage group
self.masking.move_volume_between_storage_groups(
array, device_id, source_sg_name, target_sg_name, extra_specs,
force=True, parent_sg=parent_sg)
# Check if volume should be member of GVG
self.masking.return_volume_to_volume_group(
array, volume, device_id, volume_name, extra_specs)
# Check the move was successful
success = self.rest.is_volume_in_storagegroup(
array, device_id, target_sg_name)
if not success:
LOG.error(
"Volume: %(volume_name)s has not been "
"added to target storage group %(storageGroup)s.",
{'volume_name': device_id,
'storageGroup': target_sg_name})
return False, None
else:
LOG.info("Move successful: %(success)s", {'success': success})
return success, target_sg_name
def remove_stale_data(self, model_update):
"""Remove stale RDF data
:param model_update: the model
:returns: model_update -- dict
"""
new_metadata = model_update.get('metadata')
if isinstance(new_metadata, dict):
keys = ['R2-DeviceID', 'R2-ArrayID', 'R2-ArrayModel',
'ReplicationMode', 'RDFG-Label', 'R1-RDFG', 'R2-RDFG',
'BackendID']
for k in keys:
new_metadata.pop(k, None)
return model_update
def _post_retype_srdf_protect_storage_group(
self, array, local_sg_name, device_id, volume_name,
rep_extra_specs):
"""SRDF protect SG if first volume in SG after retype operation.
:param array: the array serial number
:param local_sg_name: the local storage group name
:param device_id: the local device ID
:param volume_name: the volume name
:param rep_extra_specs: replication info dictionary
:returns: replication enables status, device pair info,
remote device id -- str, dict, str
"""
rep_mode = rep_extra_specs['rep_mode']
remote_array = rep_extra_specs['array']
rdf_group_no = rep_extra_specs['rdf_group_no']
service_level = rep_extra_specs['slo']
remote_sg_name = self.utils.derive_default_sg_from_extra_specs(
rep_extra_specs, rep_mode)
self.rest.srdf_protect_storage_group(
array, remote_array, rdf_group_no, rep_mode, local_sg_name,
service_level, rep_extra_specs, target_sg=remote_sg_name)
pair_info = self.rest.get_rdf_pair_volume(
array, rdf_group_no, device_id)
r2_device_id = pair_info['remoteVolumeName']
self.rest.rename_volume(remote_array, r2_device_id, volume_name)
if rep_mode in [utils.REP_ASYNC, utils.REP_METRO]:
self._add_volume_to_rdf_management_group(
array, device_id, volume_name, remote_array,
r2_device_id, rep_extra_specs)
return REPLICATION_ENABLED, pair_info, r2_device_id
def _retype_remote_volume(self, array, volume, device_id,
volume_name, rep_mode, is_re, extra_specs):
"""Retype the remote volume.
:param array: the array serial number
:param volume: the volume object
:param device_id: the device id
:param volume_name: the volume name
:param rep_mode: the replication mode
:param is_re: replication enabled
:param extra_specs: the target extra specs
:returns: bool
"""
success = True
rep_config = extra_specs[utils.REP_CONFIG]
rep_extra_specs = self._get_replication_extra_specs(
extra_specs, rep_config)
target_device = self.rest.get_rdf_pair_volume(
array, rep_extra_specs['rdf_group_no'], device_id)
target_device_id = target_device['remoteVolumeName']
remote_array = rep_extra_specs['array']
rep_compr_disabled = self.utils.is_compression_disabled(
rep_extra_specs)
remote_sg_name = self.masking.get_or_create_default_storage_group(
remote_array, rep_extra_specs[utils.SRP],
rep_extra_specs[utils.SLO], rep_extra_specs[utils.WORKLOAD],
rep_extra_specs, rep_compr_disabled,
is_re=is_re, rep_mode=rep_mode)
found_storage_group_list = self.rest.get_storage_groups_from_volume(
remote_array, target_device_id)
move_rqd = True
for found_storage_group_name in found_storage_group_list:
# Check if remote volume is already in the correct sg
if found_storage_group_name == remote_sg_name:
move_rqd = False
break
if move_rqd:
success, __ = self._retype_volume(
remote_array, rep_extra_specs[utils.SRP],
target_device_id, volume, volume_name, rep_extra_specs,
extra_specs[utils.SLO], extra_specs[utils.WORKLOAD],
extra_specs, remote=True)
return success
def _is_valid_for_storage_assisted_migration(
self, device_id, host, source_array, source_srp, volume_name,
do_change_compression, do_change_replication):
"""Check if volume is suitable for storage assisted (pool) migration.
:param device_id: the volume device id
:param host: the host dict
:param source_array: the volume's current array serial number
:param source_srp: the volume's current pool name
:param volume_name: the name of the volume to be migrated
:param do_change_compression: do change compression
:param do_change_replication: flag indicating replication change
:returns: boolean -- True/False
:returns: string -- targetSlo
:returns: string -- targetWorkload
"""
false_ret = (False, None, None)
host_info = host['host']
LOG.debug("Target host is : %(info)s.", {'info': host_info})
try:
info_detail = host_info.split('#')
pool_details = info_detail[1].split('+')
if len(pool_details) == 4:
target_slo = pool_details[0]
if pool_details[1].lower() == 'none':
target_workload = 'NONE'
else:
target_workload = pool_details[1]
target_srp = pool_details[2]
target_array_serial = pool_details[3]
elif len(pool_details) == 3:
target_slo = pool_details[0]
target_srp = pool_details[1]
target_array_serial = pool_details[2]
target_workload = 'NONE'
else:
raise IndexError
if target_slo.lower() == 'none':
target_slo = None
if self.rest.is_next_gen_array(target_array_serial):
target_workload = 'NONE'
except IndexError:
LOG.error("Error parsing array, pool, SLO and workload.")
return false_ret
if target_array_serial not in source_array:
LOG.error(
"The source array: %(source_array)s does not "
"match the target array: %(target_array)s - "
"skipping storage-assisted migration.",
{'source_array': source_array,
'target_array': target_array_serial})
return false_ret
if target_srp not in source_srp:
LOG.error(
"Only SLO/workload migration within the same SRP Pool is "
"supported in this version. The source pool: "
"%(source_pool_name)s does not match the target array: "
"%(target_pool)s. Skipping storage-assisted migration.",
{'source_pool_name': source_srp,
'target_pool': target_srp})
return false_ret
found_storage_group_list = self.rest.get_storage_groups_from_volume(
source_array, device_id)
if not found_storage_group_list:
LOG.warning("Volume: %(volume_name)s does not currently "
"belong to any storage groups.",
{'volume_name': volume_name})
else:
for found_storage_group_name in found_storage_group_list:
emc_fast_setting = (
self.provision.
get_slo_workload_settings_from_storage_group(
source_array, found_storage_group_name))
target_combination = ("%(targetSlo)s+%(targetWorkload)s"
% {'targetSlo': target_slo,
'targetWorkload': target_workload})
if target_combination == emc_fast_setting:
# Check if migration is to change compression
# or replication types
action_rqd = (True if do_change_compression
or do_change_replication else False)
if not action_rqd:
LOG.warning(
"No action required. Volume: %(volume_name)s is "
"already part of slo/workload combination: "
"%(targetCombination)s.",
{'volume_name': volume_name,
'targetCombination': target_combination})
return false_ret
return True, target_slo, target_workload
def configure_volume_replication(self, array, volume, device_id,
extra_specs):
"""Configure volume replication for a source device.
:param array: the array serial number
:param volume: the volume object
:param device_id: the device id
:param extra_specs: volume extra specifications
:returns: replication status, device pair info, replication info,
resume rdf -- str, dict, dict, bool
"""
# Set session attributes
LOG.debug('Starting replication setup for volume %(vol)s',
{'vol': volume.name})
resume_rdf, mgmt_sg_name = False, None
disable_compression = self.utils.is_compression_disabled(
extra_specs)
rep_config = extra_specs[utils.REP_CONFIG]
rdf_group_no, remote_array = self.get_rdf_details(
array, rep_config)
rep_extra_specs = self._get_replication_extra_specs(
extra_specs, rep_config)
rep_mode = rep_extra_specs['rep_mode']
rep_extra_specs['mgmt_sg_name'] = None
group_details = self.rest.get_rdf_group(array, rdf_group_no)
if group_details['numDevices'] == 0:
rep_info = {
'remote_array': remote_array, 'rdf_group_no': rdf_group_no,
'rep_mode': rep_mode, 'slo': rep_extra_specs['slo'],
'extra_specs': rep_extra_specs, 'target_device_id': None}
return ('first_vol_in_rdf_group', None, rep_info,
rep_extra_specs, False)
if group_details['numDevices'] > 0 and (
rep_mode in [utils.REP_ASYNC, utils.REP_METRO]):
mgmt_sg_name = self.utils.get_rdf_management_group_name(
rep_config)
self.rest.srdf_suspend_replication(
array, mgmt_sg_name, rdf_group_no, rep_extra_specs)
rep_extra_specs['mgmt_sg_name'] = mgmt_sg_name
resume_rdf = True
pair_info = self.rest.srdf_create_device_pair(
array, rdf_group_no, rep_mode, device_id, rep_extra_specs,
self.next_gen)
r2_device_id = pair_info['tgt_device']
device_uuid = self.utils.get_volume_element_name(volume.id)
self.rest.rename_volume(remote_array, r2_device_id, device_uuid)
tgt_sg_name = self.masking.get_or_create_default_storage_group(
remote_array, rep_extra_specs['srp'], rep_extra_specs['slo'],
rep_extra_specs['workload'], rep_extra_specs, disable_compression,
is_re=True, rep_mode=rep_mode)
self.rest.add_vol_to_sg(remote_array, tgt_sg_name, r2_device_id,
rep_extra_specs, force=True)
if rep_mode in [utils.REP_ASYNC, utils.REP_METRO]:
self._add_volume_to_rdf_management_group(
array, device_id, device_uuid, remote_array, r2_device_id,
extra_specs)
rep_status = REPLICATION_ENABLED
target_name = self.utils.get_volume_element_name(volume.id)
rep_info_dict = self.volume_metadata.gather_replication_info(
volume.id, 'replication', False,
rdf_group_no=rdf_group_no, target_name=target_name,
remote_array=remote_array, target_device_id=r2_device_id,
replication_status=rep_status, rep_mode=rep_mode,
rdf_group_label=rep_config['rdf_group_label'],
target_array_model=rep_extra_specs['target_array_model'],
mgmt_sg_name=rep_extra_specs['mgmt_sg_name'])
return (rep_status, pair_info, rep_info_dict, rep_extra_specs,
resume_rdf)
def _add_volume_to_rdf_management_group(
self, array, device_id, volume_name, remote_array,
target_device_id, extra_specs):
"""Add a volume to its rdf management group.
:param array: the array serial number
:param device_id: the device id
:param volume_name: the volume name
:param remote_array: the remote array
:param target_device_id: the target device id
:param extra_specs: the extra specifications
:raises: VolumeBackendAPIException
"""
grp_name = self.utils.get_rdf_management_group_name(
extra_specs[utils.REP_CONFIG])
try:
self.provision.get_or_create_group(array, grp_name, extra_specs)
self.masking.add_volume_to_storage_group(
array, device_id, grp_name, volume_name, extra_specs,
force=True)
# Add remote volume
self.provision.get_or_create_group(
remote_array, grp_name, extra_specs)
self.masking.add_volume_to_storage_group(
remote_array, target_device_id, grp_name, volume_name,
extra_specs, force=True)
except Exception as e:
exception_message = (
_('Exception occurred adding volume %(vol)s to its '
'rdf management group - the exception received was: %(e)s')
% {'vol': volume_name, 'e': six.text_type(e)})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
def break_rdf_device_pair_session(self, array, device_id, volume_name,
extra_specs):
"""Delete RDF device pair deleting R2 volume but leaving R1 in place.
:param array: the array serial number
:param device_id: the device id
:param volume_name: the volume name
:param extra_specs: the volume extra specifications
:returns: replication extra specs, resume rdf -- dict, bool
"""
LOG.debug('Starting replication cleanup for RDF pair source device: '
'%(d_id)s.', {'d_id': device_id})
# Set session attributes
resume_rdf, mgmt_sg_name = True, None
rep_config = extra_specs[utils.REP_CONFIG]
rep_extra_specs = self._get_replication_extra_specs(
extra_specs, rep_config)
remote_array = rep_extra_specs['array']
rdfg_no = rep_extra_specs['rdf_group_no']
remote_device = self.rest.get_rdf_pair_volume(
array, rdfg_no, device_id)
remote_device_id = remote_device['remoteVolumeName']
extra_specs['force_vol_remove'] = True
rep_extra_specs['force_vol_remove'] = True
# Get the names of the SGs associated with the volume on the R2 array
# before any operations are carried out - this will be used later for
# remove vol operations
r1_sg_names = self.rest.get_storage_groups_from_volume(
array, device_id)
r2_sg_names = self.rest.get_storage_groups_from_volume(
remote_array, remote_device_id)
if rep_config['mode'] in [utils.REP_ASYNC, utils.REP_METRO]:
mgmt_sg_name = self.utils.get_rdf_management_group_name(rep_config)
sg_name = mgmt_sg_name
rdf_group_state = self.rest.get_storage_group_rdf_group_state(
array, sg_name, rdfg_no)
if len(rdf_group_state) > 1 or (
rdf_group_state[0] not in utils.RDF_SYNCED_STATES):
self.rest.wait_for_rdf_group_sync(
array, sg_name, rdfg_no, rep_extra_specs)
else:
sg_name = r1_sg_names[0]
rdf_pair = self.rest.get_rdf_pair_volume(
array, rdfg_no, device_id)
rdf_pair_state = rdf_pair['rdfpairState']
if rdf_pair_state.lower() not in utils.RDF_SYNCED_STATES:
self.rest.wait_for_rdf_pair_sync(
array, rdfg_no, device_id, rep_extra_specs)
self.rest.srdf_suspend_replication(
array, sg_name, rdfg_no, rep_extra_specs)
self.rest.srdf_delete_device_pair(array, rdfg_no, device_id)
# Remove the volume from the R1 RDFG mgmt SG (R1)
if rep_config['mode'] in [utils.REP_ASYNC, utils.REP_METRO]:
self.masking.remove_volume_from_sg(
array, device_id, volume_name, mgmt_sg_name, extra_specs)
# Remove volume from R2 replication SGs
for r2_sg_name in r2_sg_names:
self.masking.remove_volume_from_sg(
remote_array, remote_device_id, volume_name, r2_sg_name,
rep_extra_specs)
if mgmt_sg_name:
if not self.rest.get_volumes_in_storage_group(array, mgmt_sg_name):
resume_rdf = False
else:
if not self.rest.get_volumes_in_storage_group(array, sg_name):
resume_rdf = False
if resume_rdf:
rep_extra_specs['mgmt_sg_name'] = sg_name
self._delete_from_srp(remote_array, remote_device_id, volume_name,
extra_specs)
return rep_extra_specs, resume_rdf
@coordination.synchronized('emc-{rdf_group}-rdf')
def _cleanup_remote_target(
self, array, volume, remote_array, device_id, target_device,
rdf_group, volume_name, rep_extra_specs):
"""Clean-up remote replication target after exception or on deletion.
:param array: the array serial number
:param volume: the volume object
:param remote_array: the remote array serial number
:param device_id: the source device id
:param target_device: the target device id
:param rdf_group: the RDF group
:param volume_name: the volume name
:param rep_extra_specs: replication extra specifications
"""
are_vols_paired, __, pair_state = (
self.rest.are_vols_rdf_paired(
array, remote_array, device_id, target_device))
if are_vols_paired:
async_grp = None
rep_mode = rep_extra_specs['rep_mode']
if rep_mode in [utils.REP_ASYNC, utils.REP_METRO]:
async_grp = self.utils.get_rdf_management_group_name(
rep_extra_specs[utils.REP_CONFIG])
sg_name = self.rest.get_storage_groups_from_volume(
array, device_id)
self.provision.break_rdf_relationship(
array, device_id, sg_name, rdf_group,
rep_extra_specs, pair_state)
self.masking.remove_and_reset_members(
remote_array, volume, target_device, volume_name,
rep_extra_specs, sg_name)
if async_grp:
self.masking.remove_and_reset_members(
remote_array, volume, target_device, volume_name,
rep_extra_specs, async_grp)
rdfg_details = self.rest.get_rdf_group(array, rdf_group)
if rdfg_details and int(rdfg_details.get('numDevices', 0)):
self.rest.srdf_resume_replication(
array, sg_name, rdf_group, rep_extra_specs)
self._delete_from_srp(
remote_array, target_device, volume_name, rep_extra_specs)
def _cleanup_replication_source(
self, array, volume, volume_name, volume_dict, extra_specs):
"""Cleanup a remote replication source volume on failure.
If replication setup fails at any stage on a new volume create,
we must clean-up the source instance as the cinder database won't
be updated with the provider_location. This means the volume cannot
be properly deleted from the array by cinder.
:param array: the array serial number
:param volume: the volume object
:param volume_name: the name of the volume
:param volume_dict: the source volume dictionary
:param extra_specs: the extra specifications
"""
LOG.warning(
"Replication failed. Cleaning up the source volume. "
"Volume name: %(sourceName)s ",
{'sourceName': volume_name})
device_id = volume_dict['device_id']
# Check if volume is snap target (e.g. if clone volume)
self._sync_check(array, device_id, extra_specs)
# Remove from any storage groups and cleanup replication
self._remove_vol_and_cleanup_replication(
array, device_id, volume_name, extra_specs, volume)
self._delete_from_srp(
array, device_id, volume_name, extra_specs)
def get_rdf_details(self, array, rep_config):
"""Retrieves an SRDF group instance.
:param array: the array serial number
:param rep_config: rep config to get details of
:returns: rdf_group_no, remote_array
"""
if not self.rep_configs:
exception_message = (_("Replication is not configured on "
"backend: %(backend)s.") %
{'backend': self.configuration.safe_get(
'volume_backend_name')})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
remote_array = rep_config['array']
rdf_group_label = rep_config['rdf_group_label']
LOG.info("Replication group: %(RDFGroup)s.",
{'RDFGroup': rdf_group_label})
rdf_group_no = self.rest.get_rdf_group_number(array, rdf_group_label)
if rdf_group_no is None:
exception_message = (_("Cannot find replication group: "
"%(RDFGroup)s. Please check the name "
"and the array") %
{'RDFGroup': rdf_group_label})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
LOG.info("Found RDF group number: %(RDFGroup)s.",
{'RDFGroup': rdf_group_no})
return rdf_group_no, remote_array
def failover_host(self, volumes, secondary_id=None, groups=None):
"""Fails over the volumes on a host back and forth.
Driver needs to update following info for failed-over volume:
1. provider_location: update array details
2. replication_status: new status for replication-enabled volume
:param volumes: the list of volumes to be failed over
:param secondary_id: the target backend
:param groups: replication groups
:returns: secondary_id, volume_update_list, group_update_list
:raises: VolumeBackendAPIException
"""
is_valid, msg = self.utils.validate_failover_request(
self.failover, secondary_id, self.rep_configs)
if not is_valid:
LOG.error(msg)
raise exception.InvalidReplicationTarget(msg)
group_fo = None
if not self.failover:
self.failover = True
self.active_backend_id = secondary_id if secondary_id else None
else:
self.failover = False
group_fo = 'default'
volume_update_list, group_update_list = (
self._populate_volume_and_group_update_lists(
volumes, groups, group_fo))
LOG.info("Failover host complete.")
return secondary_id, volume_update_list, group_update_list
def _populate_volume_and_group_update_lists(
self, volumes, groups, group_fo):
"""Populate volume and group update lists
:param volumes: the list of volumes to be failed over
:param groups: replication groups
:param group_fo: group fail over
:returns: volume_update_list, group_update_list
"""
volume_update_list = []
group_update_list = []
if groups:
for group in groups:
vol_list = []
for index, vol in enumerate(volumes):
if vol.group_id == group.id:
vol_list.append(volumes.pop(index))
grp_update, vol_updates = (
self.failover_replication(
None, group, vol_list, group_fo, host=True))
group_update_list.append({'group_id': group.id,
'updates': grp_update})
volume_update_list += vol_updates
non_rep_vol_list, sync_vol_dict, async_vol_dict, metro_vol_list = (
[], {}, {}, [])
for volume in volumes:
array = ast.literal_eval(volume.provider_location)['array']
extra_specs = self._initial_setup(volume)
extra_specs[utils.ARRAY] = array
if self.utils.is_replication_enabled(extra_specs):
device_id = self._find_device_on_array(volume, extra_specs)
self._sync_check(array, device_id, extra_specs)
rep_mode = extra_specs.get(utils.REP_MODE, utils.REP_SYNC)
backend_id = self._get_replicated_volume_backend_id(
volume)
rep_config = self.utils.get_rep_config(
backend_id, self.rep_configs)
if rep_mode == utils.REP_SYNC:
key = rep_config['rdf_group_label']
sync_vol_dict.setdefault(key, []).append(volume)
elif rep_mode == utils.REP_ASYNC:
vol_grp_name = self.utils.get_rdf_management_group_name(
rep_config)
async_vol_dict.setdefault(vol_grp_name, []).append(volume)
else:
metro_vol_list.append(volume)
else:
non_rep_vol_list.append(volume)
if len(sync_vol_dict) > 0:
for key, sync_vol_list in sync_vol_dict.items():
vol_updates = (
self._update_volume_list_from_sync_vol_list(
sync_vol_list, group_fo))
volume_update_list += vol_updates
if len(async_vol_dict) > 0:
for vol_grp_name, async_vol_list in async_vol_dict.items():
__, vol_updates = self._failover_replication(
async_vol_list, None, vol_grp_name,
secondary_backend_id=group_fo, host=True)
volume_update_list += vol_updates
if len(metro_vol_list) > 0:
__, vol_updates = (
self._failover_replication(
metro_vol_list, None, None, secondary_backend_id=group_fo,
host=True, is_metro=True))
volume_update_list += vol_updates
if len(non_rep_vol_list) > 0:
if self.failover:
# Since the array has been failed-over,
# volumes without replication should be in error.
for vol in non_rep_vol_list:
volume_update_list.append({
'volume_id': vol.id,
'updates': {'status': 'error'}})
return volume_update_list, group_update_list
def _update_volume_list_from_sync_vol_list(
self, sync_vol_list, group_fo):
"""Update the volume update list from the synced volume list
:param sync_vol_list: synced volume list
:param group_fo: group fail over
:returns: vol_updates
"""
extra_specs = self._initial_setup(sync_vol_list[0])
array = ast.literal_eval(
sync_vol_list[0].provider_location)['array']
extra_specs[utils.ARRAY] = array
temp_grp_name = self.utils.get_temp_failover_grp_name(
extra_specs[utils.REP_CONFIG])
self.provision.create_volume_group(
array, temp_grp_name, extra_specs)
device_ids = self._get_volume_device_ids(sync_vol_list, array)
self.masking.add_volumes_to_storage_group(
array, device_ids, temp_grp_name, extra_specs)
__, vol_updates = (
self._failover_replication(
sync_vol_list, None, temp_grp_name,
secondary_backend_id=group_fo, host=True))
self.rest.delete_storage_group(array, temp_grp_name)
return vol_updates
def _get_replication_extra_specs(self, extra_specs, rep_config):
"""Get replication extra specifications.
Called when target array operations are necessary -
on create, extend, etc and when volume is failed over.
:param extra_specs: the extra specifications
:param rep_config: the replication configuration
:returns: repExtraSpecs - dict
"""
if not self.utils.is_replication_enabled(extra_specs):
# Skip this if the volume is not replicated
return
rep_extra_specs = deepcopy(extra_specs)
rep_extra_specs[utils.ARRAY] = rep_config['array']
rep_extra_specs[utils.SRP] = rep_config['srp']
rep_extra_specs[utils.PORTGROUPNAME] = rep_config['portgroup']
# Get the RDF Group label & number
rep_extra_specs['rdf_group_label'] = rep_config['rdf_group_label']
rdf_group_no, __ = self.get_rdf_details(
extra_specs['array'], rep_config)
rep_extra_specs['rdf_group_no'] = rdf_group_no
# Get the SRDF wait/retries settings
rep_extra_specs['sync_retries'] = rep_config['sync_retries']
rep_extra_specs['sync_interval'] = rep_config['sync_interval']
if rep_config['mode'] == utils.REP_METRO:
exempt = True if self.next_gen else False
rep_extra_specs[utils.RDF_CONS_EXEMPT] = exempt
bias = True if rep_config.get(utils.METROBIAS) else False
rep_extra_specs[utils.METROBIAS] = bias
# If disable compression is set, check if target array is all flash
do_disable_compression = self.utils.is_compression_disabled(
extra_specs)
if do_disable_compression:
if not self.rest.is_compression_capable(
rep_extra_specs[utils.ARRAY]):
rep_extra_specs.pop(utils.DISABLECOMPRESSION, None)
# Check to see if SLO and Workload are configured on the target array.
rep_extra_specs['target_array_model'], next_gen = (
self.rest.get_array_model_info(rep_config['array']))
if extra_specs[utils.SLO]:
is_valid_slo, is_valid_workload = (
self.provision.verify_slo_workload(
rep_extra_specs[utils.ARRAY],
extra_specs[utils.SLO],
rep_extra_specs[utils.WORKLOAD], next_gen,
rep_extra_specs['target_array_model']))
if not is_valid_slo:
LOG.warning("The target array does not support the "
"storage pool setting for SLO %(slo)s, "
"setting to NONE.",
{'slo': extra_specs[utils.SLO]})
rep_extra_specs[utils.SLO] = None
if not is_valid_workload:
LOG.warning("The target array does not support the "
"storage pool setting for workload "
"%(workload)s, setting to NONE.",
{'workload': extra_specs[utils.WORKLOAD]})
rep_extra_specs[utils.WORKLOAD] = None
return rep_extra_specs
@staticmethod
def get_secondary_stats_info(rep_config, array_info):
"""On failover, report on secondary array statistics.
:param rep_config: the replication configuration
:param array_info: the array info
:returns: secondary_info - dict
"""
secondary_info = array_info.copy()
secondary_info['SerialNumber'] = six.text_type(rep_config['array'])
secondary_info['srpName'] = rep_config['srp']
return secondary_info
def create_group(self, context, group):
"""Creates a generic volume group.
:param context: the context
:param group: the group object to be created
:returns: dict -- modelUpdate
:raises: VolumeBackendAPIException, NotImplementedError, InvalidInput
"""
if (not volume_utils.is_group_a_cg_snapshot_type(group)
and not group.is_replicated):
raise NotImplementedError()
# If volume types are added during creation, validate replication
# extra_spec consistency across volume types.
extra_specs_list = list()
for volume_type_id in group.get('volume_type_ids'):
vt_extra_specs = self.utils.get_volumetype_extra_specs(
None, volume_type_id)
extra_specs_list.append(vt_extra_specs)
if group.is_replicated:
self.utils.validate_replication_group_config(
self.rep_configs, extra_specs_list)
else:
self.utils.validate_non_replication_group_config(extra_specs_list)
model_update = {'status': fields.GroupStatus.AVAILABLE}
LOG.info("Create generic volume group: %(group)s.",
{'group': group.id})
vol_grp_name = self.utils.update_volume_group_name(group)
try:
array, interval_retries_dict = self._get_volume_group_info(group)
self.provision.create_volume_group(
array, vol_grp_name, interval_retries_dict)
if group.is_replicated:
LOG.debug("Group: %(group)s is a replication group.",
{'group': group.id})
target_backend_id = extra_specs_list[0].get(
utils.REPLICATION_DEVICE_BACKEND_ID,
utils.BACKEND_ID_LEGACY_REP)
target_rep_config = self.utils.get_rep_config(
target_backend_id, self.rep_configs)
# Create remote group
__, remote_array = self.get_rdf_details(
array, target_rep_config)
self.provision.create_volume_group(
remote_array, vol_grp_name, interval_retries_dict)
model_update.update({
'replication_status': fields.ReplicationStatus.ENABLED})
except Exception:
exception_message = (_("Failed to create generic volume group:"
" %(volGrpName)s.")
% {'volGrpName': vol_grp_name})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
return model_update
def delete_group(self, context, group, volumes):
"""Deletes a generic volume group.
:param context: the context
:param group: the group object to be deleted
:param volumes: the list of volumes in the generic group to be deleted
:returns: dict -- modelUpdate
:returns: list -- list of volume model updates
:raises: NotImplementedError
"""
LOG.info("Delete generic volume group: %(group)s.",
{'group': group.id})
if (not volume_utils.is_group_a_cg_snapshot_type(group)
and not group.is_replicated):
raise NotImplementedError()
model_update, volumes_model_update = self._delete_group(
group, volumes)
return model_update, volumes_model_update
def _delete_group(self, group, volumes):
"""Helper function to delete a volume group.
:param group: the group object
:param volumes: the member volume objects
:returns: model_update, volumes_model_update
"""
volumes_model_update = []
array, interval_retries_dict = self._get_volume_group_info(group)
vol_grp_name = None
volume_group = self._find_volume_group(
array, group)
if volume_group is None:
LOG.error("Cannot find generic volume group %(volGrpName)s.",
{'volGrpName': group.id})
model_update = {'status': fields.GroupStatus.DELETED}
volumes_model_update = self.utils.update_volume_model_updates(
volumes_model_update, volumes, group.id, status='deleted')
return model_update, volumes_model_update
if 'name' in volume_group:
vol_grp_name = volume_group['name']
volume_device_ids = self._get_members_of_volume_group(
array, vol_grp_name)
deleted_volume_device_ids = []
# Remove replication for group, if applicable
if group.is_replicated:
vt_extra_specs = self.utils.get_volumetype_extra_specs(
None, group.get('volume_types')[0]['id'])
target_backend_id = vt_extra_specs.get(
utils.REPLICATION_DEVICE_BACKEND_ID,
utils.BACKEND_ID_LEGACY_REP)
target_rep_config = self.utils.get_rep_config(
target_backend_id, self.rep_configs)
self._cleanup_group_replication(
array, vol_grp_name, volume_device_ids,
interval_retries_dict, target_rep_config)
try:
if volume_device_ids:
# First remove all the volumes from the SG
self.masking.remove_volumes_from_storage_group(
array, volume_device_ids, vol_grp_name,
interval_retries_dict)
for vol in volumes:
extra_specs = self._initial_setup(vol)
device_id = self._find_device_on_array(
vol, extra_specs)
if device_id in volume_device_ids:
self._clone_check(
array, device_id, extra_specs, force_unlink=True)
self.masking.remove_and_reset_members(
array, vol, device_id, vol.name,
extra_specs, False)
self._delete_from_srp(
array, device_id, "group vol", extra_specs)
else:
LOG.debug("Volume not found on the array.")
# Add the device id to the deleted list
deleted_volume_device_ids.append(device_id)
# Once all volumes are deleted then delete the SG
self.rest.delete_storage_group(array, vol_grp_name)
model_update = {'status': fields.GroupStatus.DELETED}
volumes_model_update = self.utils.update_volume_model_updates(
volumes_model_update, volumes, group.id, status='deleted')
except Exception as e:
LOG.error("Error deleting volume group."
"Error received: %(e)s", {'e': e})
model_update = {'status': fields.GroupStatus.ERROR_DELETING}
volumes_model_update = self._handle_delete_group_exception(
deleted_volume_device_ids, volume_device_ids, group.id, array,
vol_grp_name, interval_retries_dict, volumes_model_update)
return model_update, volumes_model_update
def _handle_delete_group_exception(
self, deleted_volume_device_ids, volume_device_ids, group_id,
array, vol_grp_name, interval_retries_dict, volumes_model_update):
"""Handle delete group exception and update volume model
:param deleted_volume_device_ids: deleted volume device ids
:param volume_device_ids: volume device ids
:param group_id: group id
:param array: array serial number
:param vol_grp_name: volume group name
:param interval_retries_dict: intervals and retries dict
:param volumes_model_update: volume model update dict
:returns: volumes_model_update
"""
# Update the volumes_model_update
if deleted_volume_device_ids:
LOG.debug("Device ids: %(dev)s are deleted.",
{'dev': deleted_volume_device_ids})
volumes_not_deleted = []
for vol in volume_device_ids:
if vol not in deleted_volume_device_ids:
volumes_not_deleted.append(vol)
if not deleted_volume_device_ids:
volumes_model_update = self.utils.update_volume_model_updates(
volumes_model_update, deleted_volume_device_ids,
group_id, status='deleted')
if not volumes_not_deleted:
volumes_model_update = self.utils.update_volume_model_updates(
volumes_model_update, volumes_not_deleted,
group_id, status='error_deleting')
# As a best effort try to add back the undeleted volumes to sg
# Don't throw any exception in case of failure
try:
if not volumes_not_deleted:
self.masking.add_volumes_to_storage_group(
array, volumes_not_deleted,
vol_grp_name, interval_retries_dict)
except Exception as ex:
LOG.error("Error in rollback - %(ex)s. "
"Failed to add back volumes to sg %(sg_name)s",
{'ex': ex, 'sg_name': vol_grp_name})
return volumes_model_update
def _cleanup_group_replication(
self, array, vol_grp_name, volume_device_ids, extra_specs,
rep_config):
"""Cleanup remote replication.
Break and delete the rdf replication relationship and
delete the remote storage group and member devices.
:param array: the array serial number
:param vol_grp_name: the volume group name
:param volume_device_ids: the device ids of the local volumes
:param extra_specs: the extra specifications
:param rep_config: the rep config to use for rdf operations
"""
rdf_group_no, remote_array = self.get_rdf_details(array, rep_config)
# Delete replication for group, if applicable
group_details = self.rest.get_storage_group_rep(
array, vol_grp_name)
if group_details and group_details.get('rdf', False):
if extra_specs['rep_mode'] in [utils.REP_ASYNC, utils.REP_METRO]:
self.rest.srdf_suspend_replication(
array, vol_grp_name, rdf_group_no, extra_specs)
if volume_device_ids:
LOG.debug("Deleting remote replication for group %(sg)s", {
'sg': vol_grp_name})
self.rest.delete_storagegroup_rdf(array, vol_grp_name,
rdf_group_no)
remote_device_ids = self._get_members_of_volume_group(
remote_array, vol_grp_name)
# Remove volumes from remote replication group
if remote_device_ids:
self.masking.remove_volumes_from_storage_group(
remote_array, remote_device_ids, vol_grp_name, extra_specs)
for device_id in remote_device_ids:
# Make sure they are not members of any other storage groups
self.masking.remove_and_reset_members(
remote_array, None, device_id, 'target_vol',
extra_specs, False)
self._delete_from_srp(
remote_array, device_id, "group vol", extra_specs)
# Once all volumes are deleted then delete the SG
self.rest.delete_storage_group(remote_array, vol_grp_name)
def create_group_snapshot(self, context, group_snapshot, snapshots):
"""Creates a generic volume group snapshot.
:param context: the context
:param group_snapshot: the group snapshot to be created
:param snapshots: snapshots
:returns: dict -- modelUpdate
:returns: list -- list of snapshots
:raises: VolumeBackendAPIException, NotImplementedError
"""
grp_id = group_snapshot.group_id
source_group = group_snapshot.get('group')
if not volume_utils.is_group_a_cg_snapshot_type(source_group):
raise NotImplementedError()
snapshots_model_update = []
LOG.info(
"Create snapshot for %(grpId)s "
"group Snapshot ID: %(group_snapshot)s.",
{'group_snapshot': group_snapshot.id,
'grpId': grp_id})
try:
snap_name = self.utils.truncate_string(group_snapshot.id, 19)
self._create_group_replica(source_group, snap_name)
except Exception as e:
exception_message = (_("Failed to create snapshot for group: "
"%(volGrpName)s. Exception received: %(e)s")
% {'volGrpName': grp_id,
'e': six.text_type(e)})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
for snapshot in snapshots:
src_dev_id = self._get_src_device_id_for_group_snap(snapshot)
extra_specs = self._initial_setup(snapshot.volume)
array = extra_specs['array']
snapshot_model_dict = {
'id': snapshot.id,
'provider_location': six.text_type(
{'source_id': src_dev_id, 'snap_name': snap_name}),
'status': fields.SnapshotStatus.AVAILABLE}
snapshot_model_dict = self.update_metadata(
snapshot_model_dict, snapshot.metadata,
self.get_snapshot_metadata(
array, src_dev_id, snap_name))
snapshots_model_update.append(snapshot_model_dict)
model_update = {'status': fields.GroupStatus.AVAILABLE}
return model_update, snapshots_model_update
def _get_src_device_id_for_group_snap(self, snapshot):
"""Get the source device id for the provider_location.
:param snapshot: the snapshot object
:returns: src_device_id
"""
volume = snapshot.volume
extra_specs = self._initial_setup(volume)
return self._find_device_on_array(volume, extra_specs)
def _create_group_replica(
self, source_group, snap_name):
"""Create a group replica.
This can be a group snapshot or a cloned volume group.
:param source_group: the group object
:param snap_name: the name of the snapshot
"""
array, interval_retries_dict = self._get_volume_group_info(
source_group)
vol_grp_name = None
volume_group = (
self._find_volume_group(array, source_group))
if volume_group:
if 'name' in volume_group:
vol_grp_name = volume_group['name']
if vol_grp_name is None:
exception_message = (
_("Cannot find generic volume group %(group_id)s.") %
{'group_id': source_group.id})
raise exception.VolumeBackendAPIException(
message=exception_message)
self.provision.create_group_replica(
array, vol_grp_name,
snap_name, interval_retries_dict)
def delete_group_snapshot(self, context, group_snapshot, snapshots):
"""Delete a volume group snapshot.
:param context: the context
:param group_snapshot: the volume group snapshot to be deleted
:param snapshots: the snapshot objects
:returns: model_update, snapshots_model_update
"""
model_update, snapshots_model_update = self._delete_group_snapshot(
group_snapshot, snapshots)
return model_update, snapshots_model_update
def _delete_group_snapshot(self, group_snapshot, snapshots):
"""Helper function to delete a group snapshot.
:param group_snapshot: the group snapshot object
:param snapshots: the snapshot objects
:returns: model_update, snapshots_model_update
:raises: VolumeBackendApiException, NotImplementedError
"""
snapshots_model_update = []
source_group = group_snapshot.get('group')
grp_id = group_snapshot.group_id
if not volume_utils.is_group_a_cg_snapshot_type(source_group):
raise NotImplementedError()
LOG.info("Delete snapshot grpSnapshotId: %(grpSnapshotId)s"
" for source group %(grpId)s",
{'grpSnapshotId': group_snapshot.id,
'grpId': grp_id})
snap_name = self.utils.truncate_string(group_snapshot.id, 19)
vol_grp_name = None
try:
# Get the array serial
array, extra_specs = self._get_volume_group_info(
source_group)
# Get the volume group dict for getting the group name
volume_group = (self._find_volume_group(array, source_group))
if volume_group and volume_group.get('name'):
vol_grp_name = volume_group['name']
if vol_grp_name is None:
exception_message = (
_("Cannot find generic volume group %(grp_id)s.") %
{'group_id': source_group.id})
raise exception.VolumeBackendAPIException(
message=exception_message)
self.provision.delete_group_replica(
array, snap_name, vol_grp_name)
model_update = {'status': fields.GroupSnapshotStatus.DELETED}
for snapshot in snapshots:
snapshots_model_update.append(
{'id': snapshot.id,
'status': fields.SnapshotStatus.DELETED})
except Exception as e:
LOG.error("Error deleting volume group snapshot."
"Error received: %(e)s", {'e': e})
model_update = {
'status': fields.GroupSnapshotStatus.ERROR_DELETING}
return model_update, snapshots_model_update
def _get_snap_src_dev_list(self, array, snapshots):
"""Get the list of source devices for a list of snapshots.
:param array: the array serial number
:param snapshots: the list of snapshot objects
:returns: src_dev_ids
"""
src_dev_ids = []
for snap in snapshots:
src_dev_id, snap_name = self._parse_snap_info(array, snap)
if snap_name:
src_dev_ids.append(src_dev_id)
return src_dev_ids
def _find_volume_group(self, array, group):
"""Finds a volume group given the group.
:param array: the array serial number
:param group: the group object
:returns: volume group dictionary
"""
group_name = self.utils.update_volume_group_name(group)
volume_group = self.rest.get_storage_group_rep(array, group_name)
if not volume_group:
LOG.warning("Volume group %(group_id)s cannot be found",
{'group_id': group_name})
return None
return volume_group
def _get_members_of_volume_group(self, array, group_name):
"""Get the members of a volume group.
:param array: the array serial number
:param group_name: the storage group name
:returns: list -- member_device_ids
"""
member_device_ids = self.rest.get_volumes_in_storage_group(
array, group_name)
if not member_device_ids:
LOG.info("No member volumes found in %(group_id)s",
{'group_id': group_name})
return member_device_ids
def update_group(self, group, add_volumes, remove_volumes):
"""Updates LUNs in generic volume group.
:param group: storage configuration service instance
:param add_volumes: the volumes uuids you want to add to the vol grp
:param remove_volumes: the volumes uuids you want to remove from
the CG
:returns: model_update
:raises: VolumeBackendAPIException, NotImplementedError
"""
LOG.info("Update generic volume Group: %(group)s. "
"This adds and/or removes volumes from "
"a generic volume group.",
{'group': group.id})
if (not volume_utils.is_group_a_cg_snapshot_type(group)
and not group.is_replicated):
raise NotImplementedError()
array, interval_retries_dict = self._get_volume_group_info(group)
model_update = {'status': fields.GroupStatus.AVAILABLE}
add_vols = [vol for vol in add_volumes] if add_volumes else []
add_device_ids = self._get_volume_device_ids(add_vols, array)
remove_vols = [vol for vol in remove_volumes] if remove_volumes else []
remove_device_ids = self._get_volume_device_ids(remove_vols, array)
vol_grp_name = None
try:
volume_group = self._find_volume_group(array, group)
if volume_group:
if 'name' in volume_group:
vol_grp_name = volume_group['name']
if vol_grp_name is None:
raise exception.GroupNotFound(group_id=group.id)
# Add volume(s) to the group
if add_device_ids:
self.utils.check_rep_status_enabled(group)
for vol in add_vols:
extra_specs = self._initial_setup(vol)
self.utils.check_replication_matched(vol, extra_specs)
self.masking.add_volumes_to_storage_group(
array, add_device_ids, vol_grp_name, interval_retries_dict)
if group.is_replicated:
# Add remote volumes to remote storage group
self.masking.add_remote_vols_to_volume_group(
add_vols, group, interval_retries_dict)
# Remove volume(s) from the group
if remove_device_ids:
if group.is_replicated:
# Need force flag when manipulating RDF enabled SGs
interval_retries_dict[utils.FORCE_VOL_REMOVE] = True
# Check if the volumes exist in the storage group
temp_list = deepcopy(remove_device_ids)
for device_id in temp_list:
if not self.rest.is_volume_in_storagegroup(
array, device_id, vol_grp_name):
remove_device_ids.remove(device_id)
if remove_device_ids:
self.masking.remove_volumes_from_storage_group(
array, remove_device_ids,
vol_grp_name, interval_retries_dict)
if group.is_replicated:
# Remove remote volumes from the remote storage group
self._remove_remote_vols_from_volume_group(
array, remove_vols, group, interval_retries_dict)
except exception.GroupNotFound:
raise
except Exception as ex:
exception_message = (_("Failed to update volume group:"
" %(volGrpName)s. Exception: %(ex)s.")
% {'volGrpName': group.id,
'ex': ex})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
self.volume_metadata.capture_modify_group(
vol_grp_name, group.id, add_vols, remove_volumes, array)
return model_update, None, None
def _remove_remote_vols_from_volume_group(
self, array, volumes, group, extra_specs):
"""Remove the remote volumes from their volume group.
:param array: the array serial number
:param volumes: list of volumes
:param group: the id of the group
:param extra_specs: the extra specifications
"""
remote_device_list = []
backend_id = self._get_replicated_volume_backend_id(volumes[0])
rep_config = self.utils.get_rep_config(backend_id, self.rep_configs)
__, remote_array = self.get_rdf_details(array, rep_config)
for vol in volumes:
remote_loc = ast.literal_eval(vol.replication_driver_data)
founddevice_id = self.rest.check_volume_device_id(
remote_array, remote_loc['device_id'], vol.id)
if founddevice_id is not None:
remote_device_list.append(founddevice_id)
group_name = self.provision.get_or_create_volume_group(
array, group, extra_specs)
self.masking.remove_volumes_from_storage_group(
remote_array, remote_device_list, group_name, extra_specs)
LOG.info("Removed volumes from remote volume group.")
def _get_volume_device_ids(self, volumes, array):
"""Get volume device ids from volume.
:param volumes: volume objects
:returns: device_ids
"""
device_ids = []
for volume in volumes:
specs = {utils.ARRAY: array}
device_id = self._find_device_on_array(volume, specs)
if device_id is None:
LOG.error("Volume %(name)s not found on the array.",
{'name': volume['name']})
else:
device_ids.append(device_id)
return device_ids
def create_group_from_src(self, context, group, volumes,
group_snapshot, snapshots, source_group,
source_vols):
"""Creates the volume group from source.
:param context: the context
:param group: the volume group object to be created
:param volumes: volumes in the consistency group
:param group_snapshot: the source volume group snapshot
:param snapshots: snapshots of the source volumes
:param source_group: the source volume group
:param source_vols: the source vols
:returns: model_update, volumes_model_update
model_update is a dictionary of cg status
volumes_model_update is a list of dictionaries of volume
update
:raises: VolumeBackendAPIException, NotImplementedError
"""
if not volume_utils.is_group_a_cg_snapshot_type(group):
raise NotImplementedError()
create_snapshot = False
volumes_model_update = []
if group_snapshot:
source_id = group_snapshot.id
actual_source_grp = group_snapshot.get('group')
elif source_group:
source_id = source_group.id
actual_source_grp = source_group
create_snapshot = True
else:
exception_message = (_("Must supply either group snapshot or "
"a source group."))
raise exception.VolumeBackendAPIException(
message=exception_message)
tgt_name = self.utils.update_volume_group_name(group)
rollback_dict = {}
array, interval_retries_dict = self._get_volume_group_info(group)
source_sg = self._find_volume_group(array, actual_source_grp)
if source_sg is not None:
src_grp_name = (source_sg['name']
if 'name' in source_sg else None)
rollback_dict['source_group_name'] = src_grp_name
else:
error_msg = (_("Cannot retrieve source volume group %(grp_id)s "
"from the array.")
% {'grp_id': actual_source_grp.id})
LOG.error(error_msg)
raise exception.VolumeBackendAPIException(message=error_msg)
LOG.debug("Enter PowerMax/VMAX create_volume group_from_src. Group "
"to be created: %(grpId)s, Source : %(SourceGrpId)s.",
{'grpId': group.id, 'SourceGrpId': source_id})
try:
self.provision.create_volume_group(
array, tgt_name, interval_retries_dict)
rollback_dict.update({
'target_group_name': tgt_name, 'volumes': [],
'device_ids': [], 'list_volume_pairs': [],
'interval_retries_dict': interval_retries_dict})
model_update = {'status': fields.GroupStatus.AVAILABLE}
# Create the target devices
list_volume_pairs = []
for volume in volumes:
volumes_model_update, rollback_dict, list_volume_pairs = (
self. _create_vol_and_add_to_group(
volume, group.id, tgt_name, rollback_dict,
source_vols, snapshots, list_volume_pairs,
volumes_model_update))
snap_name, rollback_dict = (
self._create_group_replica_and_get_snap_name(
group.id, actual_source_grp, source_id, source_sg,
rollback_dict, create_snapshot))
# Link and break the snapshot to the source group
self.provision.link_and_break_replica(
array, src_grp_name, tgt_name, snap_name,
interval_retries_dict, list_volume_pairs,
delete_snapshot=create_snapshot)
# Update the replication status
if group.is_replicated:
backend = self._get_replicated_volume_backend_id(volumes[0])
rep_config = self.utils.get_rep_config(
backend, self.rep_configs)
interval_retries_dict[utils.REP_CONFIG] = rep_config
volumes_model_update = self._replicate_group(
array, volumes_model_update,
tgt_name, interval_retries_dict)
model_update.update({
'replication_status': fields.ReplicationStatus.ENABLED})
except Exception:
exception_message = (_("Failed to create vol grp %(volGrpName)s"
" from source %(grpSnapshot)s.")
% {'volGrpName': group.id,
'grpSnapshot': source_id})
LOG.error(exception_message)
if array is not None:
LOG.info("Attempting rollback for the create group from src.")
self._rollback_create_group_from_src(array, rollback_dict)
raise exception.VolumeBackendAPIException(
message=exception_message)
return model_update, volumes_model_update
def _create_group_replica_and_get_snap_name(
self, group_id, actual_source_grp, source_id, source_sg,
rollback_dict, create_snapshot):
"""Create group replica and get snap name
:param group_id: the group id
:param actual_source_grp: the source group
:param source_id: source id
:param source_sg: source storage goup
:param rollback_dict: rollback dict
:param create_snapshot: boolean
:returns: snap_name, rollback_dict
"""
if create_snapshot is True:
# We have to create a snapshot of the source group
snap_name = self.utils.truncate_string(group_id, 19)
self._create_group_replica(actual_source_grp, snap_name)
rollback_dict['snap_name'] = snap_name
else:
# We need to check if the snapshot exists
snap_name = self.utils.truncate_string(source_id, 19)
if ('snapVXSnapshots' in source_sg and
snap_name in source_sg['snapVXSnapshots']):
LOG.info("Snapshot is present on the array")
else:
error_msg = (_("Cannot retrieve source snapshot %(snap_id)s "
"from the array.") % {'snap_id': source_id})
LOG.error(error_msg)
raise exception.VolumeBackendAPIException(
message=error_msg)
return snap_name, rollback_dict
def _create_vol_and_add_to_group(
self, volume, group_id, tgt_name, rollback_dict, source_vols,
snapshots, list_volume_pairs, volumes_model_update):
"""Creates the volume group from source.
:param volume: volume object
:param group_id: the group id
:param tgt_name: target name
:param rollback_dict: rollback dict
:param source_vols: source volumes
:param snapshots: snapshot objects
:param list_volume_pairs: volume pairs list
:param volumes_model_update: volume model update
:returns: volumes_model_update, rollback_dict, list_volume_pairs
"""
src_dev_id, extra_specs, vol_size, tgt_vol_name = (
self._get_clone_vol_info(
volume, source_vols, snapshots))
volume_dict, __, __, = self._create_volume(
volume, tgt_vol_name, vol_size, extra_specs)
device_id = volume_dict['device_id']
# Add the volume to the volume group SG
self.masking.add_volume_to_storage_group(
extra_specs[utils.ARRAY], device_id, tgt_name,
tgt_vol_name, extra_specs)
# Record relevant information
list_volume_pairs.append((src_dev_id, device_id))
# Add details to rollback dict
rollback_dict['device_ids'].append(device_id)
rollback_dict['list_volume_pairs'].append(
(src_dev_id, device_id))
rollback_dict['volumes'].append(
(device_id, extra_specs, volume))
volumes_model_update.append(
self.utils.get_grp_volume_model_update(
volume, volume_dict, group_id,
meta=self.get_volume_metadata(volume_dict['array'],
volume_dict['device_id'])))
return volumes_model_update, rollback_dict, list_volume_pairs
def _get_clone_vol_info(self, volume, source_vols, snapshots):
"""Get the clone volume info.
:param volume: the new volume object
:param source_vols: the source volume list
:param snapshots: the source snapshot list
:returns: src_dev_id, extra_specs, vol_size, tgt_vol_name
"""
src_dev_id, vol_size = None, None
extra_specs = self._initial_setup(volume)
if not source_vols:
for snap in snapshots:
if snap.id == volume.snapshot_id:
src_dev_id, __ = self._parse_snap_info(
extra_specs[utils.ARRAY], snap)
vol_size = snap.volume_size
else:
for src_vol in source_vols:
if src_vol.id == volume.source_volid:
src_extra_specs = self._initial_setup(src_vol)
src_dev_id = self._find_device_on_array(
src_vol, src_extra_specs)
vol_size = src_vol.size
tgt_vol_name = self.utils.get_volume_element_name(volume.id)
return src_dev_id, extra_specs, vol_size, tgt_vol_name
def _rollback_create_group_from_src(self, array, rollback_dict):
"""Performs rollback for create group from src in case of failure.
:param array: the array serial number
:param rollback_dict: dict containing rollback details
"""
try:
# Delete the snapshot if required
if rollback_dict.get("snap_name"):
try:
self.provision.delete_group_replica(
array, rollback_dict["snap_name"],
rollback_dict["source_group_name"])
except Exception as e:
LOG.debug("Failed to delete group snapshot. Attempting "
"further rollback. Exception received: %(e)s.",
{'e': e})
if rollback_dict.get('volumes'):
# Remove any devices which were added to the target SG
if rollback_dict['device_ids']:
self.masking.remove_volumes_from_storage_group(
array, rollback_dict['device_ids'],
rollback_dict['target_group_name'],
rollback_dict['interval_retries_dict'])
# Delete all the volumes
for dev_id, extra_specs, volume in rollback_dict['volumes']:
self._remove_vol_and_cleanup_replication(
array, dev_id, "group vol", extra_specs, volume)
self._delete_from_srp(
array, dev_id, "group vol", extra_specs)
# Delete the target SG
if rollback_dict.get("target_group_name"):
self.rest.delete_storage_group(
array, rollback_dict['target_group_name'])
LOG.info("Rollback completed for create group from src.")
except Exception as e:
LOG.error("Rollback failed for the create group from src. "
"Exception received: %(e)s.", {'e': e})
def _replicate_group(self, array, volumes_model_update,
group_name, extra_specs):
"""Replicate a cloned volume group.
:param array: the array serial number
:param volumes_model_update: the volumes model updates
:param group_name: the group name
:param extra_specs: the extra specs
:returns: volumes_model_update
"""
ret_volumes_model_update = []
rdf_group_no, remote_array = self.get_rdf_details(
array, extra_specs[utils.REP_CONFIG])
self.rest.replicate_group(
array, group_name, rdf_group_no, remote_array, extra_specs)
# Need to set SRP to None for remote generic volume group - Not set
# automatically, and a volume can only be in one storage group
# managed by FAST
self.rest.set_storagegroup_srp(
remote_array, group_name, "None", extra_specs)
for volume_model_update in volumes_model_update:
vol_id = volume_model_update['id']
loc = ast.literal_eval(volume_model_update['provider_location'])
src_device_id = loc['device_id']
rdf_vol_details = self.rest.get_rdf_group_volume(
array, src_device_id)
tgt_device_id = rdf_vol_details['remoteDeviceID']
element_name = self.utils.get_volume_element_name(vol_id)
self.rest.rename_volume(remote_array, tgt_device_id, element_name)
rep_update = {'device_id': tgt_device_id, 'array': remote_array}
volume_model_update.update(
{'replication_driver_data': six.text_type(rep_update),
'replication_status': fields.ReplicationStatus.ENABLED})
volume_model_update = self.update_metadata(
volume_model_update, None, self.get_volume_metadata(
array, src_device_id))
ret_volumes_model_update.append(volume_model_update)
return ret_volumes_model_update
def enable_replication(self, context, group, volumes):
"""Enable replication for a group.
Replication is enabled on replication-enabled groups by default.
:param context: the context
:param group: the group object
:param volumes: the list of volumes
:returns: model_update, None
"""
if not group.is_replicated:
raise NotImplementedError()
model_update = {}
if not volumes:
# Return if empty group
return model_update, None
try:
vol_grp_name = None
extra_specs = self._initial_setup(volumes[0])
array = extra_specs[utils.ARRAY]
volume_group = self._find_volume_group(array, group)
if volume_group:
if 'name' in volume_group:
vol_grp_name = volume_group['name']
if vol_grp_name is None:
raise exception.GroupNotFound(group_id=group.id)
rdf_group_no, __ = self.get_rdf_details(
array, extra_specs[utils.REP_CONFIG])
self.rest.srdf_resume_replication(
array, vol_grp_name, rdf_group_no, extra_specs)
model_update.update({
'replication_status': fields.ReplicationStatus.ENABLED})
except Exception as e:
model_update.update({
'replication_status': fields.ReplicationStatus.ERROR})
LOG.error("Error enabling replication on group %(group)s. "
"Exception received: %(e)s.",
{'group': group.id, 'e': e})
return model_update, None
def disable_replication(self, context, group, volumes):
"""Disable replication for a group.
:param context: the context
:param group: the group object
:param volumes: the list of volumes
:returns: model_update, None
"""
if not group.is_replicated:
raise NotImplementedError()
model_update = {}
if not volumes:
# Return if empty group
return model_update, None
try:
vol_grp_name = None
extra_specs = self._initial_setup(volumes[0])
array = extra_specs[utils.ARRAY]
volume_group = self._find_volume_group(array, group)
if volume_group:
if 'name' in volume_group:
vol_grp_name = volume_group['name']
if vol_grp_name is None:
raise exception.GroupNotFound(group_id=group.id)
rdf_group_no, __ = self.get_rdf_details(
array, extra_specs[utils.REP_CONFIG])
self.rest.srdf_suspend_replication(
array, vol_grp_name, rdf_group_no, extra_specs)
model_update.update({
'replication_status': fields.ReplicationStatus.DISABLED})
except Exception as e:
model_update.update({
'replication_status': fields.ReplicationStatus.ERROR})
LOG.error("Error disabling replication on group %(group)s. "
"Exception received: %(e)s.",
{'group': group.id, 'e': e})
return model_update, None
def failover_replication(self, context, group, volumes,
secondary_backend_id=None, host=False):
"""Failover replication for a group.
:param context: the context
:param group: the group object
:param volumes: the list of volumes
:param secondary_backend_id: the secondary backend id - default None
:param host: flag to indicate if whole host is being failed over
:returns: model_update, vol_model_updates
"""
return self._failover_replication(
volumes, group, None,
secondary_backend_id=secondary_backend_id, host=host)
def _failover_replication(
self, volumes, group, vol_grp_name,
secondary_backend_id=None, host=False, is_metro=False):
"""Failover replication for a group.
:param volumes: the list of volumes
:param group: the group object
:param vol_grp_name: the group name
:param secondary_backend_id: the secondary backend id - default None
:param host: flag to indicate if whole host is being failed over
:returns: model_update, vol_model_updates
"""
model_update, vol_model_updates = dict(), list()
if not volumes:
# Return if empty group
return model_update, vol_model_updates
extra_specs = self._initial_setup(volumes[0])
array = ast.literal_eval(volumes[0].provider_location)['array']
extra_specs[utils.ARRAY] = array
failover = False if secondary_backend_id == 'default' else True
try:
rdf_group_no, __ = self.get_rdf_details(
array, extra_specs[utils.REP_CONFIG])
if group:
volume_group = self._find_volume_group(array, group)
if volume_group:
if 'name' in volume_group:
vol_grp_name = volume_group['name']
if vol_grp_name is None:
raise exception.GroupNotFound(group_id=group.id)
if not is_metro:
if failover:
self.rest.srdf_failover_group(
array, vol_grp_name, rdf_group_no, extra_specs)
else:
self.rest.srdf_failback_group(
array, vol_grp_name, rdf_group_no, extra_specs)
if failover:
model_update.update({
'replication_status':
fields.ReplicationStatus.FAILED_OVER})
vol_rep_status = fields.ReplicationStatus.FAILED_OVER
else:
model_update.update({
'replication_status': fields.ReplicationStatus.ENABLED})
vol_rep_status = fields.ReplicationStatus.ENABLED
except Exception as e:
model_update.update({
'replication_status': fields.ReplicationStatus.ERROR})
vol_rep_status = fields.ReplicationStatus.ERROR
LOG.error("Error failover replication on group %(group)s. "
"Exception received: %(e)s.",
{'group': vol_grp_name, 'e': e})
for vol in volumes:
loc = vol.provider_location
rep_data = vol.replication_driver_data
if vol_rep_status != fields.ReplicationStatus.ERROR:
loc = vol.replication_driver_data
rep_data = vol.provider_location
local = ast.literal_eval(loc)
remote = ast.literal_eval(rep_data)
self.volume_metadata.capture_failover_volume(
vol, local['device_id'], local['array'], rdf_group_no,
remote['device_id'], remote['array'], extra_specs,
failover, vol_grp_name, vol_rep_status, utils.REP_ASYNC)
update = {'id': vol.id,
'replication_status': vol_rep_status,
'provider_location': loc,
'replication_driver_data': rep_data}
if host:
update = {'volume_id': vol.id, 'updates': update}
vol_model_updates.append(update)
LOG.debug("Volume model updates: %s", vol_model_updates)
return model_update, vol_model_updates
def get_attributes_from_cinder_config(self):
"""Get all attributes from the configuration file
:returns: kwargs
"""
kwargs = None
username = self.configuration.safe_get(utils.VMAX_USER_NAME)
password = self.configuration.safe_get(utils.VMAX_PASSWORD)
if username and password:
serial_number = self._get_configuration_value(
utils.VMAX_ARRAY, utils.POWERMAX_ARRAY)
if serial_number is None:
LOG.error("Array Serial Number must be set in cinder.conf")
srp_name = self._get_configuration_value(
utils.VMAX_SRP, utils.POWERMAX_SRP)
if srp_name is None:
LOG.error("SRP Name must be set in cinder.conf")
slo = self._get_configuration_value(
utils.VMAX_SERVICE_LEVEL, utils.POWERMAX_SERVICE_LEVEL)
workload = self.configuration.safe_get(utils.VMAX_WORKLOAD)
port_groups = self._get_configuration_value(
utils.VMAX_PORT_GROUPS, utils.POWERMAX_PORT_GROUPS)
random_portgroup = None
if port_groups:
random_portgroup = random.choice(port_groups)
kwargs = (
{'RestServerIp': self.configuration.safe_get(
utils.VMAX_SERVER_IP),
'RestServerPort': self._get_unisphere_port(),
'RestUserName': username,
'RestPassword': password,
'SerialNumber': serial_number,
'srpName': srp_name,
'PortGroup': random_portgroup})
if self.configuration.safe_get('driver_ssl_cert_verify'):
if self.configuration.safe_get('driver_ssl_cert_path'):
kwargs.update({'SSLVerify': self.configuration.safe_get(
'driver_ssl_cert_path')})
else:
kwargs.update({'SSLVerify': True})
else:
kwargs.update({'SSLVerify': False})
if slo:
kwargs.update({'ServiceLevel': slo, 'Workload': workload})
return kwargs
def _get_configuration_value(self, first_key, second_key):
"""Get the configuration value of the first or second key
:param first_key: the first key
:param second_key: the second key
:returns: value
"""
return_value = None
if (self.configuration.safe_get(first_key)
and self.configuration.safe_get(second_key)):
LOG.error("Cannot specifiy both %(first_key)s. "
"and %(second_key)s.",
{'first_key': first_key, 'second_key': second_key})
else:
return_value = self.configuration.safe_get(first_key)
if return_value is None:
return_value = self.configuration.safe_get(second_key)
return return_value
def _get_volume_group_info(self, group):
"""Get the volume group array, retries and intervals
:param group: the group object
:returns: array -- str
interval_retries_dict -- dict
"""
array, interval_retries_dict = self.utils.get_volume_group_utils(
group, self.interval, self.retries)
if not array:
array = self._get_configuration_value(
utils.VMAX_ARRAY, utils.POWERMAX_ARRAY)
if not array:
exception_message = _(
"Cannot get the array serial_number")
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
return array, interval_retries_dict
def _get_unlink_configuration_value(self, first_key, second_key):
"""Get the configuration value of snapvx_unlink_limit
This will give back the value of the default snapvx_unlink_limit
unless either powermax_snapvx_unlink_limit or vmax_snapvx_unlink_limit
is set to something else
:param first_key: the first key
:param second_key: the second key
:returns: value
"""
return_value = self.configuration.safe_get(second_key)
if return_value == 3:
return_value = self.configuration.safe_get(first_key)
return return_value
def _get_unisphere_port(self):
"""Get unisphere port from the configuration file
:returns: unisphere port
"""
if self.configuration.safe_get(utils.U4P_SERVER_PORT):
return self.configuration.safe_get(utils.U4P_SERVER_PORT)
else:
LOG.debug("PowerMax/VMAX port is not set, using default port: %s",
utils.DEFAULT_PORT)
return utils.DEFAULT_PORT
def revert_to_snapshot(self, volume, snapshot):
"""Revert volume to snapshot.
:param volume: the volume object
:param snapshot: the snapshot object
"""
extra_specs = self._initial_setup(volume)
if self.utils.is_replication_enabled(extra_specs):
exception_message = (_(
"Volume is replicated - revert to snapshot feature is not "
"supported for replicated volumes."))
LOG.error(exception_message)
raise exception.VolumeDriverException(message=exception_message)
array = extra_specs[utils.ARRAY]
sourcedevice_id, snap_name = self._parse_snap_info(
array, snapshot)
if not sourcedevice_id or not snap_name:
LOG.error("No snapshot found on the array")
exception_message = (_(
"Failed to revert the volume to the snapshot"))
raise exception.VolumeDriverException(message=exception_message)
self._clone_check(array, sourcedevice_id, extra_specs)
try:
LOG.info("Reverting device: %(deviceid)s "
"to snapshot: %(snapname)s.",
{'deviceid': sourcedevice_id, 'snapname': snap_name})
self.provision.revert_volume_snapshot(
array, sourcedevice_id, snap_name, extra_specs)
# Once the restore is done, we need to check if it is complete
restore_complete = self.provision.is_restore_complete(
array, sourcedevice_id, snap_name, extra_specs)
if not restore_complete:
LOG.debug("Restore couldn't complete in the specified "
"time interval. The terminate restore may fail")
LOG.debug("Terminating restore session")
# This may throw an exception if restore_complete is False
self.provision.delete_volume_snap(
array, snap_name, sourcedevice_id, restored=True, generation=0)
# Revert volume to snapshot is successful if termination was
# successful - possible even if restore_complete was False
# when we checked last.
LOG.debug("Restored session was terminated")
LOG.info("Reverted the volume to snapshot successfully")
except Exception as e:
exception_message = (_(
"Failed to revert the volume to the snapshot. "
"Exception received was %(e)s") % {'e': six.text_type(e)})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
def update_metadata(
self, model_update, existing_metadata, new_metadata):
"""Update volume metadata in model_update.
:param model_update: existing model
:param existing_metadata: existing metadata
:param new_metadata: new object metadata
:returns: dict -- updated model
"""
if existing_metadata:
self._is_dict(existing_metadata, 'existing metadata')
else:
existing_metadata = dict()
if model_update:
self._is_dict(model_update, 'existing model')
if 'metadata' in model_update:
model_update['metadata'].update(existing_metadata)
else:
model_update.update({'metadata': existing_metadata})
else:
model_update = {}
model_update.update({'metadata': existing_metadata})
if new_metadata:
self._is_dict(new_metadata, 'new object metadata')
model_update['metadata'].update(new_metadata)
return model_update
def _is_dict(self, input, description):
"""Check that the input is a dict
:param input: object for checking
:raises: VolumeBackendAPIException
"""
if not isinstance(input, dict):
exception_message = (_(
"Input %(desc)s is not a dict.") % {'desc': description})
LOG.error(exception_message)
raise exception.VolumeBackendAPIException(
message=exception_message)
def get_volume_metadata(self, array, device_id):
"""Get volume metadata for model_update.
:param array: the array ID
:param device_id: the device ID
:returns: dict -- volume metadata
"""
vol_info = self.rest._get_private_volume(array, device_id)
vol_header = vol_info['volumeHeader']
array_model, __ = self.rest.get_array_model_info(array)
sl = (vol_header['serviceLevel'] if
vol_header.get('serviceLevel') else 'None')
wl = vol_header['workload'] if vol_header.get('workload') else 'None'
cd = 'False' if vol_header.get('compressionEnabled') else 'True'
metadata = {'DeviceID': device_id,
'DeviceLabel': vol_header['userDefinedIdentifier'],
'ArrayID': array, 'ArrayModel': array_model,
'ServiceLevel': sl, 'Workload': wl,
'Emulation': vol_header['emulationType'],
'Configuration': vol_header['configuration'],
'CompressionDisabled': cd}
is_rep_enabled = vol_info['rdfInfo']['RDF']
if is_rep_enabled:
rdf_info = vol_info['rdfInfo']
rdf_session = rdf_info['RDFSession'][0]
rdf_num = rdf_session['SRDFGroupNumber']
rdfg_info = self.rest.get_rdf_group(array, str(rdf_num))
r2_array_model, __ = self.rest.get_array_model_info(
rdf_session['remoteSymmetrixID'])
metadata.update(
{'ReplicationEnabled': 'True',
'R2-DeviceID': rdf_session['remoteDeviceID'],
'R2-ArrayID': rdf_session['remoteSymmetrixID'],
'R2-ArrayModel': r2_array_model,
'ReplicationMode': rdf_session['SRDFReplicationMode'],
'RDFG-Label': rdfg_info['label'],
'R1-RDFG': rdf_session['SRDFGroupNumber'],
'R2-RDFG': rdf_session['SRDFRemoteGroupNumber']})
if metadata.get('ReplicationMode') == utils.RDF_ACTIVE.title():
metadata['ReplicationMode'] = utils.REP_METRO
else:
metadata['ReplicationEnabled'] = 'False'
return metadata
def get_snapshot_metadata(self, array, device_id, snap_name):
"""Get snapshot metadata for model_update.
:param array: the array ID
:param device_id: the device ID
:param snap_name: the snapshot name
:returns: dict -- volume metadata
"""
snap_info = self.rest.get_volume_snap_info(array, device_id)
device_name = snap_info['deviceName']
device_label = device_name.split(':')[1]
metadata = {'SnapshotLabel': snap_name,
'SourceDeviceID': device_id,
'SourceDeviceLabel': device_label}
return metadata
def _check_and_add_tags_to_storage_array(
self, serial_number, array_tag_list, extra_specs):
"""Add tags to a storage group.
:param serial_number: the array serial number
:param array_tag_list: the array tag list
:param extra_specs: the extra specifications
"""
if array_tag_list:
existing_array_tags = self.rest.get_array_tags(serial_number)
new_tag_list = self.utils.get_new_tags(
self.utils.convert_list_to_string(array_tag_list),
self.utils.convert_list_to_string(existing_array_tags))
if not new_tag_list:
LOG.warning("No new tags to add. Existing tags "
"associated with %(array)s are "
"%(tags)s.",
{'array': serial_number,
'tags': existing_array_tags})
else:
self._validate_array_tag_list(new_tag_list)
LOG.info("Adding the tags %(tag_list)s to %(array)s",
{'tag_list': new_tag_list,
'array': serial_number})
try:
self.rest.add_storage_array_tags(
serial_number, new_tag_list, extra_specs)
except Exception as ex:
LOG.warning("Unexpected error: %(ex)s. If you still "
"want to add tags to this storage array, "
"please do so on the Unisphere UI.",
{'ex': ex})
def prepare_replication_details(self, extra_specs):
"""Prepare details required for initialising replication.
:param extra_specs: extra sepcifications
:returns: first volume in SG, replication extra specs, replication info
dict -- bool, dict, dict
"""
rep_info_dict, rep_first_vol, rdfg_empty = dict(), True, True
# Get volume type replication extra specs
rep_extra_specs = self._get_replication_extra_specs(
extra_specs, extra_specs[utils.REP_CONFIG])
# Get the target SG name for the current volume create op
sg_name = self.utils.derive_default_sg_from_extra_specs(
extra_specs, rep_mode=extra_specs['rep_mode'])
rep_extra_specs['sg_name'] = sg_name
# Check if the RDFG has volume in it regardless of target SG state
rdf_group_details = self.rest.get_rdf_group(
extra_specs['array'], rep_extra_specs['rdf_group_no'])
rdfg_device_count = rdf_group_details['numDevices']
if rdfg_device_count > 0:
rdfg_empty = False
# Check if there are any volumes in the SG, will return 0 if the SG
# does not exist
if self.rest.get_num_vols_in_sg(extra_specs['array'], sg_name):
# Volumes exist, not first volume in SG
rep_first_vol = False
# Get the list of the current devices in the SG, this will help
# with determining the new device added because no device ID is
# returned
local_device_list = self.rest.get_volume_list(
extra_specs['array'],
{'storageGroupId': sg_name})
# Set replication info that we will need for creating volume in
# existing SG, these are not required for new SGs as the only
# additional step required is to SRDF protect the SG
rep_info_dict.update({
'local_array': extra_specs['array'],
'remote_array': rep_extra_specs['array'],
'rdf_group_no': rep_extra_specs['rdf_group_no'],
'rep_mode': extra_specs['rep_mode'],
'sg_name': sg_name,
'service_level': extra_specs['slo'],
'initial_device_list': local_device_list,
'sync_interval': rep_extra_specs['sync_interval'],
'sync_retries': rep_extra_specs['sync_retries']})
return rep_first_vol, rep_extra_specs, rep_info_dict, rdfg_empty
def srdf_protect_storage_group(self, extra_specs, rep_extra_specs,
volume_dict):
"""SRDF protect a storage group.
:param extra_specs: source extra specs
:param rep_extra_specs: replication extra specs
:param volume_dict: volume details dict
"""
self.rest.srdf_protect_storage_group(
extra_specs['array'], rep_extra_specs['array'],
rep_extra_specs['rdf_group_no'], extra_specs['rep_mode'],
volume_dict['storage_group'], rep_extra_specs['slo'], extra_specs)
def get_and_set_remote_device_uuid(
self, extra_specs, rep_extra_specs, volume_dict):
"""Get a remote device id and set device UUID.
:param extra_specs: source extra specs
:param rep_extra_specs: replication extra specs
:param volume_dict: volume details dict
:returns: remote device ID -- str
"""
rdf_pair = self.rest.get_rdf_pair_volume(
extra_specs['array'], rep_extra_specs['rdf_group_no'],
volume_dict['device_id'])
self.rest.rename_volume(rep_extra_specs['array'],
rdf_pair['remoteVolumeName'],
volume_dict['device_uuid'])
return rdf_pair['remoteVolumeName']
def gather_replication_updates(self, extra_specs, rep_extra_specs,
volume_dict):
"""Gather replication updates for returns.
:param extra_specs: extra specs
:param rep_extra_specs: replication extra specs
:param volume_dict: volume info dict
:returns: replication status, replication info -- str, dict
"""
replication_update = (
{'replication_status': REPLICATION_ENABLED,
'replication_driver_data': six.text_type(
{'array': rep_extra_specs['array'],
'device_id': volume_dict['remote_device_id']})})
rep_info_dict = self.volume_metadata.gather_replication_info(
volume_dict['device_uuid'], 'replication', False,
local_array=extra_specs['array'],
remote_array=rep_extra_specs['array'],
target_device_id=volume_dict['remote_device_id'],
target_name=volume_dict['device_uuid'],
rdf_group_no=rep_extra_specs['rdf_group_no'],
rep_mode=extra_specs['rep_mode'],
replication_status=REPLICATION_ENABLED,
rdf_group_label=rep_extra_specs['rdf_group_label'],
target_array_model=rep_extra_specs['target_array_model'],
backend_id=rep_extra_specs[
utils.REP_CONFIG].get(utils.BACKEND_ID, None))
return replication_update, rep_info_dict
def _cleanup_non_rdf_volume_create_post_failure(
self, volume, volume_name, extra_specs, device_ids):
"""Delete lingering volumes that exist in an non-RDF SG post exception.
:param volume: Cinder volume -- Volume
:param volume_name: Volume name -- str
:param extra_specs: Volume extra specs -- dict
:param device_ids: Devices ids to be deleted -- list
"""
array = extra_specs[utils.ARRAY]
for device_id in device_ids:
self.masking.remove_and_reset_members(
array, volume, device_id, volume_name, extra_specs, False)
self._delete_from_srp(
array, device_id, volume_name, extra_specs)
def _cleanup_rdf_volume_create_post_failure(
self, volume, volume_name, extra_specs, device_ids):
"""Delete lingering volumes that exist in an RDF SG post exception.
:param volume: Cinder volume -- Volume
:param volume_name: Volume name -- str
:param extra_specs: Volume extra specs -- dict
:param device_ids: Devices ids to be deleted -- list
"""
__, rep_extra_specs, __, __ = self.prepare_replication_details(
extra_specs)
array = extra_specs[utils.ARRAY]
srp = extra_specs['srp']
slo = extra_specs['slo']
workload = extra_specs['workload']
do_disable_compression = self.utils.is_compression_disabled(
extra_specs)
rep_mode = extra_specs['rep_mode']
rdf_group = rep_extra_specs['rdf_group_no']
rep_config = extra_specs[utils.REP_CONFIG]
if rep_mode is utils.REP_SYNC:
storagegroup_name = self.utils.get_default_storage_group_name(
srp, slo, workload, do_disable_compression, True, rep_mode)
else:
storagegroup_name = self.utils.get_rdf_management_group_name(
rep_config)
self.rest.srdf_resume_replication(
array, storagegroup_name, rdf_group, rep_extra_specs)
for device_id in device_ids:
__, __, vol_is_rdf = self.rest.is_vol_in_rep_session(
array, device_id)
if vol_is_rdf:
self.cleanup_rdf_device_pair(array, rdf_group, device_id,
extra_specs)
else:
self.masking.remove_and_reset_members(
array, volume, device_id, volume_name, extra_specs, False)
self._delete_from_srp(
array, device_id, volume_name, extra_specs)
def _validate_rdfg_status(self, array, extra_specs):
"""Validate RDF group states before and after various operations
:param array: array serial number -- str
:param extra_specs: volume extra specs -- dict
"""
rep_extra_specs = self._get_replication_extra_specs(
extra_specs, extra_specs[utils.REP_CONFIG])
rep_mode = extra_specs['rep_mode']
rdf_group_no = rep_extra_specs['rdf_group_no']
# Get default storage group for volume
disable_compression = self.utils.is_compression_disabled(extra_specs)
storage_group_name = self.utils.get_default_storage_group_name(
extra_specs['srp'], extra_specs['slo'], extra_specs['workload'],
disable_compression, True, extra_specs['rep_mode'])
# Check for storage group. Will be unavailable for first vol create
storage_group_details = self.rest.get_storage_group(
array, storage_group_name)
storage_group_available = storage_group_details is not None
if storage_group_available:
is_rep = self._validate_storage_group_is_replication_enabled(
array, storage_group_name)
is_exclusive = self._validate_rdf_group_storage_group_exclusivity(
array, storage_group_name)
is_valid_states = self._validate_storage_group_rdf_states(
array, storage_group_name, rdf_group_no, rep_mode)
if not (is_rep and is_exclusive and is_valid_states):
msg = (_('RDF validation for storage group %s failed. Please '
'see logged error messages for specific details.'
) % storage_group_name)
raise exception.VolumeBackendAPIException(msg)
# Perform checks against Async or Metro management storage groups
if rep_mode is not utils.REP_SYNC:
management_sg_name = self.utils.get_rdf_management_group_name(
extra_specs['rep_config'])
management_sg_details = self.rest.get_storage_group(
array, management_sg_name)
management_sg_available = management_sg_details is not None
if management_sg_available:
is_rep = self._validate_storage_group_is_replication_enabled(
array, management_sg_name)
is_excl = self._validate_rdf_group_storage_group_exclusivity(
array, management_sg_name)
is_valid_states = self._validate_storage_group_rdf_states(
array, management_sg_name, rdf_group_no, rep_mode)
is_cons = self._validate_management_group_volume_consistency(
array, management_sg_name, rdf_group_no)
if not (is_rep and is_excl and is_valid_states and is_cons):
msg = (_(
'RDF validation for storage group %s failed. Please '
'see logged error messages for specific details.')
% management_sg_name)
raise exception.VolumeBackendAPIException(msg)
def _validate_storage_group_is_replication_enabled(
self, array, storage_group_name):
"""Validate that a storage groups is marked as RDF enabled
:param array: array serial number -- str
:param storage_group_name: name of the storage group -- str
:returns: consistency validation checks passed -- boolean
"""
is_valid = True
sg_details = self.rest.get_storage_group_rep(array, storage_group_name)
sg_rdf_enabled = sg_details.get('rdf', False)
if not sg_rdf_enabled:
LOG.error('Storage group %s is expected to be RDF enabled but '
'is not. Please check that all volumes in this storage '
'group are RDF enabled and part of the same RDFG.',
storage_group_name)
is_valid = False
return is_valid
def _validate_storage_group_rdf_states(
self, array, storage_group_name, rdf_group_no, rep_mode):
"""Validate that the RDF states found for storage groups are valid.
:param array: array serial number -- str
:param storage_group_name: name of the storage group -- str
:param rep_mode: replication mode being used -- str
:returns: consistency validation checks passed -- boolean
"""
is_valid = True
sg_rdf_states = self.rest.get_storage_group_rdf_group_state(
array, storage_group_name, rdf_group_no)
# Verify Async & Metro modes only have a single state
if rep_mode is not utils.REP_SYNC:
if len(sg_rdf_states) > 1:
sg_states_str = (', '.join(sg_rdf_states))
LOG.error('More than one RDFG state found for storage group '
'%s. We expect a single state for all volumes when '
'using %s replication mode. Found %s states.',
storage_group_name, rep_mode, sg_states_str)
is_valid = False
# Determine which list of valid states to use
if rep_mode is utils.REP_SYNC:
valid_states = utils.RDF_VALID_STATES_SYNC
elif rep_mode is utils.REP_ASYNC:
valid_states = utils.RDF_VALID_STATES_ASYNC
else:
valid_states = utils.RDF_VALID_STATES_METRO
# Validate storage group states
for state in sg_rdf_states:
if state.lower() not in valid_states:
valid_states_str = (', '.join(valid_states))
LOG.error('Invalid RDF state found for storage group %s. '
'Found state %s. Valid states are %s.',
storage_group_name, state, valid_states_str)
is_valid = False
return is_valid
def _validate_rdf_group_storage_group_exclusivity(
self, array, storage_group_name):
"""Validate that a storage group only has one RDF group.
:param array: array serial number -- str
:param storage_group_name: name of storage group -- str
:returns: consistency validation checks passed -- boolean
"""
is_valid = True
sg_rdf_groups = self.rest.get_storage_group_rdf_groups(
array, storage_group_name)
if len(sg_rdf_groups) > 1:
rdf_groups_str = ', '.join(sg_rdf_groups)
LOG.error('Detected more than one RDF group associated with '
'storage group %s. Only one RDFG should be associated '
'with a storage group. Found RDF groups %s',
storage_group_name, rdf_groups_str)
is_valid = False
return is_valid
def _validate_management_group_volume_consistency(
self, array, management_sg_name, rdf_group_number):
"""Validate volume consistency between management SG and RDF group
:param array: array serial number -- str
:param management_sg_name: name of storage group -- str
:param rdf_group_number: rdf group number to check -- str
:returns: consistency validation checks passed -- boolean
"""
is_valid = True
rdfg_volumes = self.rest.get_rdf_group_volume_list(
array, rdf_group_number)
sg_volumes = self.rest.get_volumes_in_storage_group(
array, management_sg_name)
missing_volumes = list()
for rdfg_volume in rdfg_volumes:
if rdfg_volume not in sg_volumes:
missing_volumes.append(rdfg_volume)
if missing_volumes:
missing_volumes_str = ', '.join(missing_volumes)
LOG.error(
'Inconsistency found between management group %s and RDF '
'group %s. The following volumes are not in the management '
'storage group %s. All Asynchronous and Metro volumes must '
'be managed together.',
management_sg_name, rdf_group_number, missing_volumes_str)
is_valid = False
return is_valid
View Git Blame Copy Permalink