测试环境：3个计算节点，分别属于3个zone
[[email protected] ~(keystone_admin)]# nova availability-zone-list
+-----------------------+----------------------------------------+
| Name | Status |
+-----------------------+----------------------------------------+
| internal | available |
| |- controller2 | |
| | |- nova-conductor | enabled :-) 2016-08-20T14:57:07.000000 |
| | |- nova-scheduler | enabled :-) 2016-08-20T14:57:06.000000 |
| | |- nova-consoleauth | enabled :-) 2016-08-20T14:57:08.000000 |
| | |- nova-cert | enabled :-) 2016-08-20T14:57:07.000000 |
| nova | available |
| |- controller3 | |
| | |- nova-compute | enabled :-) 2016-08-20T14:57:04.000000 |
| ag1 | available |
| |- controller1 | |
| | |- nova-compute | enabled :-) 2016-08-19T23:41:45.000000 |
| ag2 | available |
| |- controller2 | |
| | |- nova-compute | enabled :-) 2016-08-20T14:57:06.000000 |
+-----------------------+----------------------------------------+
测试方式： 启动虚拟机的时候选择Booting from image(creates a new volume)测试
产生的原因： cinder无法识别nova的多zone，cinder能获取的zone信息有：
1、cinder-volume所在的zone
2、cinder.conf配置文件中的两个参数storage_availability_zone = nova和default_availability_zone = nova
大致过下代码调用过程：(从nova那边调用cinder那部分开始)
1、nova/virt/block_device.py -> class DriverImageBlockDevice def attach vol = volume_api.create
2、nova/volume/cinder.py -> class API def create item = client.volumes.create
3、cinder/api/v2/volumes.py -> class VolumeController def create new_volume = self.volume_api.create
4、cinder/volume/api.py -> class API def create flow_engine = create_volume.get_flow
在create函数中cinder获取能获得到的zone的信息：
raw_zones = self.list_availability_zones(enable_cache=True)
availability_zones = set([az[‘name‘] for az in raw_zones])
if CONF.storage_availability_zone:
availability_zones.add(CONF.storage_availability_zone)
def list_availability_zones services = objects.ServiceList.get_all_by_topic ->
cinder/objects/service.py（def get_all_by_topic） db.service_get_all_by_topic ->
cinder/db/api.py（def service_get_all_by_topic） IMPL.service_get_all_by_topic ->
cinder/db/sqlalchemy/api.py（def service_get_all_by_topic）
@require_admin_context def service_get_all_by_topic(context, topic, disabled=None): query = model_query( context, models.Service, read_deleted="no").\ # models在这里cinder/db/sqlalchemy/models.py filter_by(topic=topic) # topic默认传过来的是cinder-volume if disabled is not None: query = query.filter_by(disabled=disabled) return query.all()
查询数据库找出cinder-volume所在的zone
5、cinder/volume/flows/api/create_volume.py -> def get_flow
# 这里是taskflow，只需关注add方法里面的东东
def get_flow(db_api, image_service_api, availability_zones, create_what,
scheduler_rpcapi=None, volume_rpcapi=None):
"""Constructs and returns the api entrypoint flow.
This flow will do the following:
1. Inject keys & values for dependent tasks.
2. Extracts and validates the input keys & values.
3. Reserves the quota (reverts quota on any failures).
4. Creates the database entry.
5. Commits the quota.
6. Casts to volume manager or scheduler for further processing.
"""
flow_name = ACTION.replace(":", "_") + "_api"
api_flow = linear_flow.Flow(flow_name)
api_flow.add(ExtractVolumeRequestTask(
image_service_api,
availability_zones,
rebind={‘size‘: ‘raw_size‘,
‘availability_zone‘: ‘raw_availability_zone‘,
‘volume_type‘: ‘raw_volume_type‘}))
api_flow.add(QuotaReserveTask(),
EntryCreateTask(db_api),
QuotaCommitTask())
if scheduler_rpcapi and volume_rpcapi:
# This will cast it out to either the scheduler or volume manager via
# the rpc apis provided.
api_flow.add(VolumeCastTask(scheduler_rpcapi, volume_rpcapi, db_api))
# Now load (but do not run) the flow using the provided initial data.
return taskflow.engines.load(api_flow, store=create_what)
我们关心的在class ExtractVolumeRequestTask里
先看下入口execute方法，为什么看execute方法，你翻下taskflow怎么用的，就懂了
def execute(self, context, size, snapshot, image_id, source_volume,
availability_zone, volume_type, metadata, key_manager,
source_replica, consistencygroup, cgsnapshot):
utils.check_exclusive_options(snapshot=snapshot,
imageRef=image_id,
source_volume=source_volume)
policy.enforce_action(context, ACTION)
# TODO(harlowja): what guarantee is there that the snapshot or source
# volume will remain available after we do this initial verification??
snapshot_id = self._extract_snapshot(snapshot)
source_volid = self._extract_source_volume(source_volume)
source_replicaid = self._extract_source_replica(source_replica)
size = self._extract_size(size, source_volume, snapshot)
consistencygroup_id = self._extract_consistencygroup(consistencygroup)
cgsnapshot_id = self._extract_cgsnapshot(cgsnapshot)
self._check_image_metadata(context, image_id, size)
availability_zone = self._extract_availability_zone(availability_zone, # 关心的在这里
snapshot,
source_volume)
# _extract_availability_zone函数
def _extract_availability_zone(self, availability_zone, snapshot,
source_volume):
"""Extracts and returns a validated availability zone.
This function will extract the availability zone (if not provided) from
the snapshot or source_volume and then performs a set of validation
checks on the provided or extracted availability zone and then returns
the validated availability zone.
"""
# Try to extract the availability zone from the corresponding snapshot
# or source volume if either is valid so that we can be in the same
# availability zone as the source.
if availability_zone is None:
if snapshot:
try:
availability_zone = snapshot[‘volume‘][‘availability_zone‘]
except (TypeError, KeyError):
pass
if source_volume and availability_zone is None:
try:
availability_zone = source_volume[‘availability_zone‘]
except (TypeError, KeyError):
pass
if availability_zone is None:
if CONF.default_availability_zone: # default_availability_zone判断
availability_zone = CONF.default_availability_zone
else:
# For backwards compatibility use the storage_availability_zone
availability_zone = CONF.storage_availability_zone
if availability_zone not in self.availability_zones: # self.availability_zones就是上述cinder-volume的zone外加cinder.conf中两个配置参数
if CONF.allow_availability_zone_fallback: # allow_availability_zone_fallback这个很关键
original_az = availability_zone
availability_zone = (
CONF.default_availability_zone or
CONF.storage_availability_zone)
LOG.warning(_LW("Availability zone ‘%(s_az)s‘ "
"not found, falling back to "
"‘%(s_fallback_az)s‘."),
{‘s_az‘: original_az,
‘s_fallback_az‘: availability_zone})
else:
msg = _("Availability zone ‘%(s_az)s‘ is invalid.")
msg = msg % {‘s_az‘: availability_zone}
raise exception.InvalidInput(reason=msg) # 没启动allow_availability_zone_fallback的话，就抛异常了