Openstack liberty 云主机迁移源码分析之静态迁移1

时间:2023-02-02 02:14:18

虚拟机迁移使资源配置更加灵活,尤其是在线迁移,提供了虚拟机的可用性和可靠性。Openstack liberty中提供了两种类型的迁移实现:静态迁移(cold migration)和动态迁移(live migration)。在接下来的几篇文章中,我将详细分析两种迁移的实现过程,先来看静态迁移。

限于篇幅,静态迁移的源码分析将包含两篇文章:

  • 第一篇:主要介绍迁移过程中nova-apinova-conductor所在的工作
  • 第二篇:重点介绍nova-compute的处理过程

下面请看第一篇的内容:

发起迁移

用户可以手动通过nova CLI命令行发起云主机迁移动作:

#nova --debug migrate 52e4d485-6ccf-47f3-a754-b62649e7b256

上述命令将id=52e4d485-6ccf-47f3-a754-b62649e7b256的云主机迁移到另外一个最优的nova-compute节点上,--debug选项用来显示执行日志:

......

curl -g -i -X POST http://controller:8774/v2/eab72784b36040a186a6b88dac9ac0b2/servers/5a7d302f-f388-4ffb-af37-f1e6964b3a51/action -H "User-Agent: python-novaclient" -H "Content-Type: application/json" -H "Accept: application/json" -H "X-Auth-Token: {SHA1}8e294a111a5deaa45f6cb0f3c58a600d2b1b0493" -d '{"migrate": null}

......

上述截取的日志表明:novaclient通过http方式将迁移请求发送给nova-api并执行migrate动作(action),由nova-api启动时建立的路由映射,很容易的知道,该动作的入口函数为
nova/api/openstack/compute/migrate_server.py/MigrateServerController._migrate,下文具体分析。

源码分析

nova-api部分

如上分析,迁移入口如下:

#nova/api/openstack/compute/migrate_server.py/MigrateServerController._migrate, 省略装饰器定义
def _migrate(self, req, id, body):
"""Permit admins to migrate a server to a new host.
req 是Request对象,包含该次请求信息
id 是待迁移的云主机id 如:52e4d485-6ccf-47f3-a754-b62649e7b256
body 是该次请求的参数信息 {"migrate": null}
"""


#从Request对象提取请求上下文
context = req.environ['nova.context']
"""执行权限认证,默认会通过读取host节点/etc/nova/policy.json文件
中的权限规则完成认证,如果没有定义相关的规则,则表明认证失败抛抛异
这里对应的认证规则是:
"os_compute_api:os_migrate_server:migrate": rule:admin_api"
"""

authorize(context, action='migrate')

#从nova数据库中获取id指向的云主机信息,返回一个InstanceV2对象
instance = common.get_instance(self.compute_api, context, id)
"""省略异常处理代码
如果云主机不存在,找不到合适的目标主机,云主机处于锁定状态,
资源不足,云主机状态不对(只能是运行或者停止态)则抛异常

与‘调整云主机大小’(resize)操作一样,也是调用
`/nova/compute/api.py/API.resize`
执行迁移操作,resize是通过判断
是否指定了flavor_id参数来判断是执行‘调整云主机大小’还是‘迁移’操作,
请看下文的具体分析
"""

self.compute_api.resize(req.environ['nova.context'], instance)

---------------------------------------------------------------
#接上文:/nova/compute/api.py/API.resize, 省略装饰器定义
def resize(self, context, instance, flavor_id=None,
clean_shutdown=True,
**extra_instance_updates)
:

"""Resize (ie, migrate) a running instance.

If flavor_id is None, the process is considered a
migration, keeping the original flavor_id. If flavor_id is
not None, the instance should be migrated to a new host and
resized to the new flavor_id.

上面的注释是说:如果flavor_id = None, 则用原有的flavor(配置)执行
迁移操作。如果不为None,则应将云主机迁移到新的主机并应用flavor_id指
定的配置

conext 请求上下文
instance InstanceV2实例对象,包含云主机的详细配置信息
flavor_id 配置模板id,这里为None,因为是迁移操作
clean_shutdown = True, 静态迁移时开启关机重试,如果未能正常关闭云
主机会抛异常
"""


#检查系统磁盘的‘自动配置磁盘’功能是否打开,否则抛异常
#迁移完成后,云主机需要能够自动配置系统磁盘
self._check_auto_disk_config(instance, **extra_instance_updates)
#获取云主机配置模板信息
current_instance_type = instance.get_flavor()

# If flavor_id is not provided, only migrate the instance.
#flavor_id = None, 执行迁移操作;打印日志并将当前配置作为迁移后云主
#机的配置
if not flavor_id:
LOG.debug("flavor_id is None. Assuming migration.",
instance=instance)
new_instance_type = current_instance_type
else:
#从nova.instance_types数据表获取flavor_id指定的配置模板信息
#read_deleted="no",表示读取数据库时过滤掉已经删除的配置模板
new_instance_type = flavors.get_flavor_by_flavor_id(
flavor_id, read_deleted="no")
#如果云主机是从镜像启动的并且当前的配置模板中root_gb(根磁盘大
#小)不为0,而目标配置模板中的root_gb=0,则不支持resize操作
#因为不知道怎么分配系统磁盘大小了,抛异常
if (new_instance_type.get('root_gb') == 0 and
current_instance_type.get('root_gb') != 0 and
not self.is_volume_backed_instance(context, instance)):
reason = _('Resize to zero disk flavor is not'
'allowed.')
raise exception.CannotResizeDisk(reason=reason)

#如果没有找到指定的配置模板,抛异常
if not new_instance_type:
raise exception.FlavorNotFound(flavor_id=flavor_id)

#打印debug日志
current_instance_type_name = current_instance_type['name']
new_instance_type_name = new_instance_type['name']
LOG.debug("Old instance type %(current_instance_type_name)s, "
" new instance type %(new_instance_type_name)s",
{'current_instance_type_name':
current_instance_type_name,
'new_instance_type_name': new_instance_type_name},
instance=instance)

#判断是否是同一配置模板,迁移操作中肯定是同一配置模板
same_instance_type = (current_instance_type['id'] ==
new_instance_type['id'])

"""NOTE(sirp): We don't want to force a customer to change
their flavor when Ops is migrating off of a failed host.
"""

#如果是resize操作,新的配置模板被disable了,抛异常
if not same_instance_type and new_instance_type.get('disabled'):
raise exception.FlavorNotFound(flavor_id=flavor_id)

#默认cell关闭,cell_type = None
#这里是说resize的时候,新旧配置模板不能是相同的,因为这样做没有意义
if same_instance_type and flavor_id and
self.cell_type != 'compute':
raise exception.CannotResizeToSameFlavor()

# ensure there is sufficient headroom for upsizes
#如果是resize操作,需要先保留资源配额
if flavor_id:
#获取vcpu和memory的增量配额(如果有的话,新旧配置模板的差值)
deltas = compute_utils.upsize_quota_delta(context,
new_instance_type,
current_instance_type)
try:
#为当前用户和项目保留资源(增量)配额,更新数据库
quotas = compute_utils.reserve_quota_delta(context,
deltas,
instance)
except exception.OverQuota as exc:
#统计资源不足信息,并打印日志
quotas = exc.kwargs['quotas']
overs = exc.kwargs['overs']
usages = exc.kwargs['usages']
headroom = self._get_headroom(quotas, usages,
deltas)
(overs, reqs, total_alloweds,
useds) = self._get_over_quota_detail(headroom,
overs, quotas, deltas)
LOG.warning(_LW("%(overs)s quota exceeded for %"
"(pid)s, tried to resize instance."),
{'overs': overs, 'pid': context.project_id})
raise exception.TooManyInstances(overs=overs,
req=reqs,
used=useds,
allowed=total_alloweds)
#迁移操作,没有额外的资源需要保留
else:
quotas = objects.Quotas(context=context)

#更新与主机状态:主机状态:重建/迁移,任务状态:准备重建或者迁移
instance.task_state = task_states.RESIZE_PREP
instance.progress = 0
instance.update(extra_instance_updates)
instance.save(expected_task_state=[None])

"""为nova-scheduler生成过滤选项,
CONF.allow_resize_to_same_host = true
表示允许迁移的目的主机与源主机相同,否则过滤掉源主机
"""

filter_properties = {'ignore_hosts': []}
if not CONF.allow_resize_to_same_host:
filter_properties['ignore_hosts'].append(instance.host)

#默认cell_type = None,
if self.cell_type == 'api':
# Commit reservations early and create migration record.
self._resize_cells_support(context, quotas, instance,
current_instance_type,
new_instance_type)

#flavor_id = None, 执行迁移操作,否则执行resize
#记录实例操作,更新nova.instance_actions数据表,迁移结束后会更新数
#据库记录,反映迁移结果
if not flavor_id:
self._record_action_start(context, instance,
instance_actions.MIGRATE)
else:
self._record_action_start(context, instance,
instance_actions.RESIZE)
"""将迁移请求转发给
`/nova/conductor/api.py/ComputeTaskAPI.resize_instance`,该
方法直接调用
`nova/conductor/rpcapi.py/ComputeTaskAPI.migrate_server`处理
请求,请看下文的分析
"""

scheduler_hint = {'filter_properties': filter_properties}
self.compute_task_api.resize_instance(context, instance,
extra_instance_updates,
scheduler_hint=scheduler_hint,
flavor=new_instance_type,
reservations=quotas.reservations or [],
clean_shutdown=clean_shutdown)

------------------------------------------------------------
#接上文:`nova/conductor/rpcapi.py/ComputeTaskAPI.migrate_server`
def migrate_server(self, context, instance, scheduler_hint,
live, rebuild,
flavor, block_migration, disk_over_commit,
reservations=None, clean_shutdown=True)
:


"""输入参数如下:
live = False, 静态迁移
rebuild = false, 迁移,而不是resize
block_migration = None, 不是块迁移
disk_over_commit = None
reservations = [] 迁移操作,没有增量保留资源
"""

#生成请求参数字典
kw = {'instance': instance, 'scheduler_hint':
scheduler_hint,
'live': live, 'rebuild': rebuild, 'flavor': flavor,
'block_migration': block_migration,
'disk_over_commit': disk_over_commit,
'reservations': reservations,
'clean_shutdown': clean_shutdown}
#根据RPCClient的版本兼容性,选择客户端版本。
#在初始化rpc的时候会设置版本兼容特性
version = '1.11'
if not self.client.can_send_version(version):
del kw['clean_shutdown']
version = '1.10'
if not self.client.can_send_version(version):
kw['flavor'] = objects_base.obj_to_primitive(flavor)
version = '1.6'
if not self.client.can_send_version(version):
kw['instance'] = jsonutils.to_primitive(
objects_base.obj_to_primitive(instance))
version = '1.4'
#通过同步rpc调用将`migrate_server`消息发送给rabbitmq,
#消费者`nova-conductor`将会收到该消息
cctxt = self.client.prepare(version=version)
return cctxt.call(context, 'migrate_server', **kw)

小结:nova-api主要完成实例状态、相关条件检查, 之后更新云主机状态及添加nova.instance_actions数据库记录,最后通过同步rpc将请求转发给nova-conductor处理

nova-conductor部分

由前述的分析,我们很容易就知道nova-conductor处理迁移请求的入口:

#/nova/conductor/manager.py/ComputeTaskManager.migrate_server
def migrate_server(self, context, instance, scheduler_hint,
live, rebuild,
flavor, block_migration, disk_over_commit,
reservations=None,
clean_shutdown=True)
:

"""各输入参数来自`nova-api`,如下:
scheduler_hint 调度选项,{u'filter_properties':
{u'ignore_hosts': []}}
live = False, 静态迁移
rebuild = Flase, 迁移而不是调整云主机大小
block_migration = None, 非块迁移
disk_over_commit = None
reservations = [] ,迁移操作没有增量保留资源
"""

#如果输入的instance参数不是非法的NovaObject对象,就先从数据库获取
#云主机信息,然后生成InstanceV2对象
if instance and not isinstance(instance, nova_object.NovaObject):
# NOTE(danms): Until v2 of the RPC API, we need to tolerate
# old-world instance objects here
attrs = ['metadata', 'system_metadata', 'info_cache',
'security_groups']
instance = objects.Instance._from_db_object(
context, objects.Instance(), instance,
expected_attrs=attrs)
# NOTE: Remove this when we drop support for v1 of the RPC API
#如果输入的flavor参数不是合法的Flavor对象,就先从数据库提取指定id
#的配置模板,然后生成Flavor对象
if flavor and not isinstance(flavor, objects.Flavor):
# Code downstream may expect extra_specs to be
#populated since it is receiving an object, so lookup
#the flavor to ensure this.
flavor = objects.Flavor.get_by_id(context, flavor['id'])

#动态迁移,在另外一篇文章中详述
if live and not rebuild and not flavor:
self._live_migrate(context, instance, scheduler_hint,
block_migration, disk_over_commit)
#调用_cold_migrate执行静态迁移,下文具体分析
elif not live and not rebuild and flavor:
instance_uuid = instance.uuid
#with语句,在迁移前记录迁移事件记录到数据库
#(nova.instance_actions_events),迁移后更新数据库迁移记录
with compute_utils.EventReporter(context, 'cold_migrate',
instance_uuid):
self._cold_migrate(context, instance, flavor,
scheduler_hint['filter_properties'],
reservations, clean_shutdown)
#未知类型
else:
raise NotImplementedError()
-------------------------------------------------------------
#接上文:
def _cold_migrate(self, context, instance, flavor,
filter_properties,
reservations, clean_shutdown)
:

#从实例对象中获取所使用的镜像信息,示例如下:
"""
{u'min_disk': u'20', u'container_format': u'bare',
u'min_ram': u'0', u'disk_format': u'raw', 'properties':
{u'base_image_ref': u'e0cc468f-6501-4a85-9b19-
70e782861387'}}
"""

image = utils.get_image_from_system_metadata(
instance.system_metadata)
#通过镜像属性、云主机属性、云主机配置模板生成请求参数字典,格式如下:
"""
request_spec = {
'image': image,
'instance_properties': instance,
'instance_type': flavor,
'num_instances': 1}
"""

request_spec = scheduler_utils.build_request_spec(
context, image, [instance], instance_type=flavor)
#生成迁移任务对象
#`/nova/conductor/tasks/migrate.py/MigrationTask
task = self._build_cold_migrate_task(context, instance,
flavor,
filter_properties,
request_spec,
reservations,
clean_shutdown)

"""省略异常处理代码
如果未找到合适的目标主机,策略不合法等异常,则退出
在退出前会更新数据库,设置云主机的状态并打印日志及发送
`compute_task.migrate_server`通知
"""

#执行迁移,下文具体分析
task.execute()

---------------------------------------------------------------
#接上文:`nova/conductor/tasks/migrate.py/MigrationTask._execute
def _execute(self):
#从请求参数中获取所使用的镜像信息
image = self.request_spec.get('image')
#根据self.reservations保留配额生成配额对象,
#迁移操作没有保留配额 self.reservations = []
self.quotas = objects.Quotas.from_reservations(self.context,
self.reservations,
instance=self.instance)
#添加组(group_hosts)及组策略(group_polices)信息到过滤属性(如果有
#的话)
scheduler_utils.setup_instance_group(self.context,
self.request_spec,
self.filter_properties)
"""添加重试参数到过滤属性(如果配置的重试次数
CONF.scheduler_max_attempts 〉1的话),修改后的过滤属性如下:
{'retry': {'num_attempts': 1, 'hosts': []},
u'ignore_hosts': []}

如果是`nova-compute`发送过来的重试请求,输入的filter_properties过
滤属性中的retry字典中包含
前一次请求的异常信息,再次选择目标主机的时候会排除`hosts`中的主机,在
populate_retry过程中,会打印该条异常日志;如果重试超过了最大重试次
数,也会抛异常
"""

scheduler_utils.populate_retry(
self.filter_properties,
self.instance.uuid)
#发送请求给`nova-scheduler`,根据过滤规则选择合适的目标主机,
#如果超时会根据前文的重试参数重试。如果成功,返回合适的目标主机列表
#如果找不到合适的目标主机,抛异常
hosts = self.scheduler_client.select_destinations(
self.context, self.request_spec, self.filter_properties)
#选取第一个
host_state = hosts[0]
#添加目标主机到过滤属性的重试列表(重试的时候'hosts'中的主机被忽
略),示例如下:
"""
{'retry': {'num_attempts': 1, 'hosts': [[u'devstack',
u'devstack']]}, 'limits': {u'memory_mb': 11733.0,
u'disk_gb': 1182.0}, u'ignore_hosts': []}
"""

scheduler_utils.populate_filter_properties(
self.filter_properties,
host_state)
# context is not serializable
self.filter_properties.pop('context', None)

#通过异步rpc调用发送`prep_resize`消息到消息队列,`nova-compute`会
#处理该请求(`nova/compute/rpcapi.py/ComputeAPI`)
(host, node) = (host_state['host'], host_state['nodename'])
self.compute_rpcapi.prep_resize(
self.context, image, self.instance, self.flavor, host,
self.reservations, request_spec=self.request_spec,
filter_properties=self.filter_properties, node=node,
clean_shutdown=self.clean_shutdown)

小结:nova-conductor主要是借助nova-scheduler选择合适的目标主机,同时也会更新nova.instance_actions_events数据表,最后发起异步rpc调用将迁移请求转交给nova-compute处理

到这里静态迁移的前篇就介绍完成了,过程还是比较简单的:主要完成一些条件判断,更新数据库记录,通过nova-scheduler选主,最后将请求转交给nova-compute处理。敬请期待:
Openstack liberty 云主机迁移源码分析之静态迁移2