OpenStack_Swift源码分析——创建Ring及添加设备源码算法详细分析

1 创建Ring 代码详细分析

在OpenStack_Swift——Ring组织架构中我们详细分析了Ring的具体工作过程，下面就Ring中增加设备，删除设备，已经重新平衡的实现过程作详细的介绍。

首先看RingBuilder类

def __init__(self, part_power, replicas, min_part_hours):
#why 最大 2**32
if part_power > 32:
raise ValueError("part_power must be at most 32 (was %d)"
% (part_power,))
if replicas < 1:
raise ValueError("replicas must be at least 1 (was %.6f)"
% (replicas,))
if min_part_hours < 0:
raise ValueError("min_part_hours must be non-negative (was %d)"
% (min_part_hours,))

self.part_power = part_power
self.replicas = replicas
self.min_part_hours = min_part_hours
self.parts = 2 ** self.part_power          #分区数量即虚拟节点数量
self.devs = []                             #用来存放设备的添加的设备都会加入到devs里
self.devs_changed = False
self.version = 0

self._last_part_moves_epoch = None
self._last_part_moves = None

self._last_part_gather_start = 0
self._remove_devs = []
self._ring = None

其中devs里面没一个dev含有的属性为： dev = {'weight': 100.0, 'zone': 0, 'ip': '127.0.0.1', 'region': 0, 'parts': 0, 'id': 0, 'meta': 'some meta data', 'device': 'sda1', 'parts_wanted': 96, 'port': 6000}

其中part_wanted 是在添加设备后设备需要的虚拟节点数，parts为设备已经被分配的虚拟节点。

在swift-ring-builder object.builder create 18 3 1

会调用Command类的create方法，具体看此方法：

def create(self):
"""
swift-ring-builder object.builder create 18 3 1
swift-ring-builder <builder_file> create <part_power> <replicas>
<min_part_hours>
Creates <builder_file> with 2^<part_power> partitions and <replicas>.
<min_part_hours> is number of hours to restrict（限制） moving a partition more
than once（不止一次）.
"""
if len(argv) < 6:
print Commands.create.__doc__.strip()
exit(EXIT_ERROR)
builder = RingBuilder(int(argv[3]), float(argv[4]), int(argv[5]))
backup_dir = pathjoin(dirname(argv[1]), 'backups')
try:
mkdir(backup_dir)
except OSError as err:
if err.errno != EEXIST:
raise
builder.save(pathjoin(backup_dir, '%d.' % time() + basename(argv[1])))
builder.save(argv[1])             #序列化
exit(EXIT_SUCCESS)

在第一次创建时，首先会判断/etc/swift下是否有object.builder文件，没有需要创建此文件，并根据命令中个的 18 3 1 来实例化RingBuilder，最后将builder序列化。

2 为Ring添加设备

创建了.builder文件后，就需要添加设备，下面为添加一个设备的命令：swift-ring-builder object.builder add z2-127.0.0.1:6020/sdb2 100

其中 add 指明为添加设备 z2 为zone2，127.0.0.1:6020/sdb2为设备ip地址，绑定的端口以及设备中存储文件的磁盘，100为设备的权重

下面看添加设备的具体实现：

这个方法在main方法中药判断/etc/swift下是否有object.builder文件，在第一步create中我们已经创建了此文件，现在只需要将他反序列化回来并用存的数据来实例化RingBuilder类,然后调用Commands类的add方法：

def add(self):
"""
swift-ring-builder <builder_file> add
[r<region>]z<zone>-<ip>:<port>[R<r_ip>:<r_port>]/<device_name>_<meta>
<weight>
[[r<region>]z<zone>-<ip>:<port>[R<r_ip>:<r_port>]/<device_name>_<meta>
<weight>] ...

Where <r_ip> and <r_port> are replication ip and port.

or

swift-ring-builder <builder_file> add
[--region <region>] --zone <zone> --ip <ip> --port <port>
--replication-ip <r_ip> --replication-port <r_port>
--device <device_name> --meta <meta> --weight <weight>

Adds devices to the ring with the given information. No partitions will be
assigned to the new device until after running 'rebalance'. This is so you
can make multiple device changes and rebalance them all just once.
"""
if len(argv) < 5 or len(argv) % 2 != 1:
print Commands.add.__doc__.strip()
exit(EXIT_ERROR)

for new_dev in _parse_add_values(argv[3:]):
for dev in builder.devs:
if dev is None:
continue
if dev['ip'] == new_dev['ip'] and \
dev['port'] == new_dev['port'] and \
dev['device'] == new_dev['device']:
print 'Device %d already uses %s:%d/%s.' % \
(dev['id'], dev['ip'], dev['port'], dev['device'])
print "The on-disk ring builder is unchanged.\n"
exit(EXIT_ERROR)
dev_id = builder.add_dev(new_dev)         #调用RingBuilder类的add_dev方法
print('Device %s with %s weight got id %s' %
(format_device(new_dev), new_dev['weight'], dev_id))

builder.save(argv[1])
exit(EXIT_SUCCESS)

将新添加的设备会加入到devs[]中，其中会调用add_dev,下面看此方法的具体实现

def add_dev(self, dev):
"""
Add a device to the ring. This device dict should have a minimum of the
following keys:

======  ===============================================================
id      unique integer identifier amongst devices. Defaults to the next
id if the 'id' key is not provided in the dict
weight  a float of the relative weight of this device as compared to
others; this indicates how many partitions the builder will try
to assign to this device
region  integer indicating which region the device is in
zone    integer indicating which zone the device is in; a given
partition will not be assigned to multiple devices within the
same (region, zone) pair if there is any alternative
ip      the ip address of the device
port    the tcp port of the device
device  the device's name on disk (sdb1, for example)
meta    general use 'extra' field; for example: the online date, the
hardware description
======  ===============================================================

.. note::
This will not rebalance the ring immediately as you may want to
make multiple changes for a single rebalance.

:param dev: device dict

:returns: id of device
dev  =  {'weight': 100.0, 'zone': 0, 'ip': '127.0.0.1', 'region': 0, 'parts': 0, 'id': 0, 'meta': 'some meta data', 'device': 'sda1', 'parts_wanted': 96, 'port': 6000}
"""
if 'id' not in dev:
dev['id'] = 0
if self.devs:
dev['id'] = max(d['id'] for d in self.devs if d) + 1
if dev['id'] < len(self.devs) and self.devs[dev['id']] is not None:
raise exceptions.DuplicateDeviceError(
'Duplicate device id: %d' % dev['id'])
# Add holes to self.devs to ensure self.devs[dev['id']] will be the dev
while dev['id'] >= len(self.devs):
self.devs.append(None)
dev['weight'] = float(dev['weight'])
dev['parts'] = 0
self.devs[dev['id']] = dev
self._set_parts_wanted()         #设置part_wanted
self.devs_changed = True
self.version += 1
return dev['id']

此方法主要是获得weight,并为设备dev中的id赋值，以及part_wanted赋值。其中part_wanted值的计算为：

((self.parts * self.replicas / sum(d['weight'] for d in self._iter_devs())) * dev['weight']) - dev['parts']

parts=2**power(此处在create方法中的18) replicas为备份数，此处为3

即part_wanted=（(虚拟节点数*备份数/所有已添加的节点的weight和)*当前设备的weight）-已被分配给此设备的虚拟节点数

至此 创建Ring和为Ring添加设备讲解完毕，再完成这两个步骤后，就需要平衡环，并为replica2part2dev（备份到分区到设备的映射）赋值，平衡算法是Ring 的核心算法，在下一篇文章中具体讲解。

由于本人水平有限，文中难免出现理解错误，敬请指正、交流，谢谢！