如何通过 SyncManager 远程访问 multiprocessing.Manager().Namespace 元素？
How to access multiprocessing.Manager().Namespace element remotely via SyncManager?
无论我多么努力地阅读文档，我都无法理解如何使用 SyncManager 远程访问多处理命名空间的项目。
请注意，下面的代码是最短的代码，用于演示哪些有效，哪些无效。我的问题的症结在于对 register 的调用，我未能正确理解如何使用它。
下面的示例由两部分组成（都在同一个文件中mp-example.py）：服务器和客户端。为了让客户端完成它的工作，服务器部分必须 运行ning（在另一个 terminal/screen 中）。
首先，有效的代码：
#!/usr/bin/env python3
import os
import sys
import multiprocessing as mp
import multiprocessing.managers as mpm
import queue
import socket
import contextlib
from contextlib import contextmanager
def get_ip():
with contextlib.closing(socket.socket(socket.AF_INET, socket.SOCK_DGRAM)) as s:
s.connect(('192.168.17.1', 80))
return s.getsockname()[0]
def get_free_port():
with contextlib.closing(socket.socket(socket.AF_INET, socket.SOCK_STREAM)) as s:
s.bind(('', 0))
s.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
return s.getsockname()[1]
def startserver(name):
# Find a free port for the SyncManager
newportnumber = get_free_port()
Server = {'address': (get_ip(), newportnumber), 'authkey': b'mptest'}
# Save data to file
serverdata_filename = f'mpSERVERDATA.txt'
with open(serverdata_filename, 'wt', encoding="utf-8") as serverdata_file:
print(Server, file=serverdata_file)
print(Server)
# On the server, we do not need our own IP address
Server['address'] = ('', newportnumber)
# Set up server:
# define a multiprocessing manager
# define a multipprocessing namespace
# define a Value item in the namespace
# define a Value item outside the namespace
# define a multiprocessing syncmanager
# start the syncmanager
# register both the namespace and the value outside the namespace
# set values
# initialise a different process
# print the values
# run the other process & wait for completion
# print the values
# read standard input & print the values (waiting for a remote process connection)
m = mp.Manager()
m.name = 'mp.Manager: ' + name
with m:
ns = m.Namespace()
ns.name = 'm.Namespace: ' + name
ns.yv = m.Value(int, 0)
yv = m.Value(int, 0)
bm = mpm.SyncManager(**Server)
bm.name = 'mpm.SyncManager: ' + name
#bm.register('ns', callable=lambda: ns, proxytype=mpm.NamespaceProxy,
# exposed=('__getattribute__', '__setattr__', '__delattr__', 'yv')
# )
# The above has not made any difference
bm.register('ns', callable=lambda: ns, proxytype=mpm.NamespaceProxy)
bm.register('yv', callable=lambda: yv, proxytype=mpm.ValueProxy)
with bm:
yv.value = 99
ns.yv = 999
def f(n, y):
n.yv += 222
y.value += 33
proc = mp.Process(target=f, args=(ns, yv))
print(f'yv={yv}, ns.yv={ns.yv}')
proc.start()
proc.join()
proc.terminate()
print(f'yv={yv}, ns.yv={ns.yv}, ns={ns}')
print('Waiting for input (waiting for client), press CTRL-D to end')
for line in sys.stdin:
print(f'yv={yv}, ns.yv={ns.yv}, ns={ns}')
print('Waiting for input (waiting for client), press CTRL-D to end')
elif sys.argv[1] == 'server' and len(sys.argv) == 3:
startserver(sys.argv[2])
elif sys.argv[1] == 'client' and len(sys.argv) == 3:
startclient(sys.argv[2])
上面的代码在 运行 单独执行时符合我的预期：
> python3 mp-example.py server test
{'address': ('192.168.17.10', 50793), 'authkey': b'mptest'}
yv=Value(<class 'int'>, 99), ns.yv=999
yv=Value(<class 'int'>, 132), ns.yv=1221, ns=Namespace(name='m.Namespace: test', yv=1221)
Waiting for input (waiting for client), press CTRL-D to end
接下来，不起作用的位：
下面的代码在一定程度上起作用，但在尝试访问命名空间中的值项 yv 时失败。
代码：
def startclient(name):
# retrieve server identification from file
serverdata_filename = f'mpSERVERDATA.txt'
with open(serverdata_filename, 'rt', encoding="utf-8") as serverdata_file:
Server = eval(serverdata_file.read())
print(Server)
bm = mpm.SyncManager(**Server)
bm.register('yv', proxytype=mpm.ValueProxy)
bm.register('ns', proxytype=mpm.NamespaceProxy)
bm.connect()
yv = bm.yv()
print(f'yv={yv.value}')
yv.value += 55
print(f'yv={yv.value}')
ns = bm.ns()
print(f'ns={ns}')
print(f'ns.yv={ns.yv}')
ns.yv += 666
print(f'ns.yv={ns.yv}')
输出：
> python3 mp-example.py client test
{'address': ('192.168.17.10', 50793), 'authkey': b'mptest'}
yv=627
yv=682
['_Client', '__class__', '__deepcopy__', '__delattr__', '__dict__', '__dir__', '__doc__', '__eq__', '__format__', '__ge__', '__getattr__', '__getattribute__', '__gt__', '__hash__', '__init__', '__init_subclass__', '__le__', '__lt__', '__module__', '__ne__', '__new__', '__reduce__', '__reduce_ex__', '__repr__', '__setattr__', '__sizeof__', '__str__', '__subclasshook__', '__weakref__', '_address_to_local', '_after_fork', '_authkey', '_callmethod', '_close', '_connect', '_decref', '_exposed_', '_getvalue', '_id', '_idset', '_incref', '_manager', '_mutex', '_owned_by_manager', '_serializer', '_tls', '_token']
Help on method _getvalue in module multiprocessing.managers:
_getvalue() method of multiprocessing.managers.NamespaceProxy instance
Get a copy of the value of the referent
None
ns=<NamespaceProxy object, typeid 'Namespace' at 0x7f5d05689a20>
Traceback (most recent call last):
File "mp-example.py", line 333, in <module>
startclient(sys.argv[2])
File "mp-example.py", line 282, in startclient
print(f'ns.yv={ns.yv}')
File "/usr/lib/python3.6/multiprocessing/managers.py", line 1060, in __getattr__
return callmethod('__getattribute__', (key,))
File "/usr/lib/python3.6/multiprocessing/managers.py", line 772, in _callmethod
raise convert_to_error(kind, result)
AttributeError: 'NamespaceProxy' object has no attribute 'yv'
问题：
我需要做什么才能在客户端访问ns.yv？
我也为此苦苦挣扎——很难按照您所遵循的思路想出任何既不复杂又低效的解决方案。顺便说一句，您正在尝试 return 共享内存值的代理，该值已经可以共享而无需代理，并且如果您的目标只是提供一个单例值，则不需要它在共享内存中。
一个更简单的方法，它不需要您创建两个管理器进程，只是为了创建一个代理对象，然后在您启动它之前向另一个管理器注册，就是定义您自己的自定义 类.您通常会创建自己的 multiprocessing.managers.BaseManager 子类，您将在其中注册新的托管 类（当然，如果您的客户还需要已经在 SyncManager 中注册的 类，例如 dict，那么您可以将其子类化）。
以下程序展示了如何创建单例 Namespace 等效类型。我把你的IP地址注释掉了，用127.0.0.1来测试。我还用更安全的替代方法替换了您对 eval 的调用。
#!/usr/bin/env python3
import os
import sys
import multiprocessing as mp
from multiprocessing.managers import BaseManager, NamespaceProxy
import socket
import contextlib
from ast import literal_eval
from threading import Thread
def get_ip():
with contextlib.closing(socket.socket(socket.AF_INET, socket.SOCK_DGRAM)) as s:
#s.connect(('192.168.17.1', 80))
s.connect(('127.0.0.1', 80))
return s.getsockname()[0]
def get_free_port():
with contextlib.closing(socket.socket(socket.AF_INET, socket.SOCK_STREAM)) as s:
s.bind(('', 0))
s.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
return s.getsockname()[1]
class MyNamespaceManager(BaseManager):
pass
class MyNamespace:
pass
# Singleton:
singleton = MyNamespace()
def get_singleton():
return singleton
def run_server(server):
server.serve_forever()
def startserver():
# Find a free port for the SyncManager
newportnumber = get_free_port()
Server = {'address': (get_ip(), newportnumber), 'authkey': b'mptest'}
# Save data to file
serverdata_filename = f'mpSERVERDATA.txt'
with open(serverdata_filename, 'wt', encoding="utf-8") as serverdata_file:
print(Server, file=serverdata_file)
print(Server)
# On the server, we do not need our own IP address
Server['address'] = ('', newportnumber)
# Singleton
MyNamespaceManager.register('MyNamespace', callable=get_singleton, proxytype=NamespaceProxy)
manager = MyNamespaceManager(**Server)
"""
manager.start()
input('Hit enter to end: ')
manager.shutdown()
"""
server = manager.get_server()
Thread(target=run_server, args=(server,)).start()
try:
input('Hit enter to end: ')
except KeyboardInterrupt:
pass
os.unlink(serverdata_filename)
server.stop_event.set() # The server stops and calls sys.exit()
def startclient():
# retrieve server identification from file
serverdata_filename = f'mpSERVERDATA.txt'
with open(serverdata_filename, 'rt', encoding="utf-8") as serverdata_file:
Server = literal_eval(serverdata_file.read())
print(Server)
MyNamespaceManager.register('MyNamespace', proxytype=NamespaceProxy)
manager = MyNamespaceManager(**Server)
manager.connect()
yv1 = manager.MyNamespace()
yv1.value = 99
yv1.value += 55
yv1.x = 9
yv2 = manager.MyNamespace()
print(f'yv2.value={yv2.value}')
print(f'yv2.x={yv2.x}')
if __name__ == '__main__':
if len(sys.argv) == 2 and sys.argv[1] == 'server':
startserver()
else:
startclient()
客户端打印：
{'address': ('127.0.0.1', 51211), 'authkey': b'mptest'}
yv2.value=154
yv2.x=9
编辑-综合解决方案：
我通过编写一个新的 RemoteSyncManager class 来封装 multiprocessing.managers.SyncManager() 实例来解决我的大部分挫折。
我的这个库确实简化了 SyncManager 本地和远程进程的使用。
前往我的 GitHub 存储库 PythonLibraries-Multiprocessing-RemoteSyncManager 了解详细信息。
我的第一个解决方案：
我终于设法设置了一个与远程连接同步的托管多处理命名空间。
但是这个设置拒绝使用 Managed dict，它的工作方式与上面 Booboo 的建议类似。
有什么方法可以在远程共享命名空间中定义和保持同步整个字典，而无需在每次操作时重新分配字典？
代码：
if __name__ == '__main__':
if sys.argv[1] == 'Server' and len(sys.argv) == 2:
#localmanager = mp.Manager()
#with localmanager:
q1 = mp.Queue(1)
q2 = mp.Queue(1)
localnamespace = mpm.Namespace()
class syncmanager(mpm.SyncManager):
pass
syncmanager.register('get_q1', callable=lambda: q1)
syncmanager.register('get_q2', callable=lambda: q2)
syncmanager.register('get_ns', callable=lambda: localnamespace, proxytype=mpm.NamespaceProxy)
remotemanager = syncmanager(address = ('', 56789), authkey = b'mypassword')
remotemanager.start()
with remotemanager:
ns = remotemanager.get_ns()
#ns.sampledict = remotemanager.dict()
ns.sampledict = {}
#ns.sampledict = localmanager.dict()
ns.sampledict['testabc'] = 1234567
ns.sampledict['TESTABC'] = 9876
d = {}
d['testabc'] = 1234567
d['TESTABC'] = 9876
ns.sampledict2 = d
print(f'Server ns = {ns}')
print(q2.get())
q1.put(f'Server ns = {ns}')
ns.sampledict['testabc'] = 1
ns.sampleserverint = 98789
print(f'Server ns = {ns}')
print(q2.get())
q1.put(f'Server ns = {ns}')
elif sys.argv[1] == 'Client' and len(sys.argv) == 2:
class syncmanager(mpm.SyncManager):
pass
syncmanager.register('get_q1')
syncmanager.register('get_q2')
syncmanager.register('get_ns', proxytype=mpm.NamespaceProxy)
remotemanager = syncmanager(address = ('', 56789), authkey = b'mypassword')
remotemanager.connect()
q1 = remotemanager.get_q1()
q2 = remotemanager.get_q2()
ns = remotemanager.get_ns()
q2.put('ready')
msg = q1.get()
print(msg)
print(f'Client ns = {ns}')
ns.sampledict['TESTABC'] = 9
ns.sampleclientint = 7887
q2.put('next.1')
msg = q1.get()
print(msg)
print(f'Client ns = {ns}')
无论我多么努力地阅读文档，我都无法理解如何使用 SyncManager 远程访问多处理命名空间的项目。
请注意，下面的代码是最短的代码，用于演示哪些有效，哪些无效。我的问题的症结在于对 register 的调用，我未能正确理解如何使用它。
下面的示例由两部分组成（都在同一个文件中mp-example.py）：服务器和客户端。为了让客户端完成它的工作，服务器部分必须 运行ning（在另一个 terminal/screen 中）。
首先，有效的代码：
#!/usr/bin/env python3
import os
import sys
import multiprocessing as mp
import multiprocessing.managers as mpm
import queue
import socket
import contextlib
from contextlib import contextmanager
def get_ip():
with contextlib.closing(socket.socket(socket.AF_INET, socket.SOCK_DGRAM)) as s:
s.connect(('192.168.17.1', 80))
return s.getsockname()[0]
def get_free_port():
with contextlib.closing(socket.socket(socket.AF_INET, socket.SOCK_STREAM)) as s:
s.bind(('', 0))
s.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
return s.getsockname()[1]
def startserver(name):
# Find a free port for the SyncManager
newportnumber = get_free_port()
Server = {'address': (get_ip(), newportnumber), 'authkey': b'mptest'}
# Save data to file
serverdata_filename = f'mpSERVERDATA.txt'
with open(serverdata_filename, 'wt', encoding="utf-8") as serverdata_file:
print(Server, file=serverdata_file)
print(Server)
# On the server, we do not need our own IP address
Server['address'] = ('', newportnumber)
# Set up server:
# define a multiprocessing manager
# define a multipprocessing namespace
# define a Value item in the namespace
# define a Value item outside the namespace
# define a multiprocessing syncmanager
# start the syncmanager
# register both the namespace and the value outside the namespace
# set values
# initialise a different process
# print the values
# run the other process & wait for completion
# print the values
# read standard input & print the values (waiting for a remote process connection)
m = mp.Manager()
m.name = 'mp.Manager: ' + name
with m:
ns = m.Namespace()
ns.name = 'm.Namespace: ' + name
ns.yv = m.Value(int, 0)
yv = m.Value(int, 0)
bm = mpm.SyncManager(**Server)
bm.name = 'mpm.SyncManager: ' + name
#bm.register('ns', callable=lambda: ns, proxytype=mpm.NamespaceProxy,
# exposed=('__getattribute__', '__setattr__', '__delattr__', 'yv')
# )
# The above has not made any difference
bm.register('ns', callable=lambda: ns, proxytype=mpm.NamespaceProxy)
bm.register('yv', callable=lambda: yv, proxytype=mpm.ValueProxy)
with bm:
yv.value = 99
ns.yv = 999
def f(n, y):
n.yv += 222
y.value += 33
proc = mp.Process(target=f, args=(ns, yv))
print(f'yv={yv}, ns.yv={ns.yv}')
proc.start()
proc.join()
proc.terminate()
print(f'yv={yv}, ns.yv={ns.yv}, ns={ns}')
print('Waiting for input (waiting for client), press CTRL-D to end')
for line in sys.stdin:
print(f'yv={yv}, ns.yv={ns.yv}, ns={ns}')
print('Waiting for input (waiting for client), press CTRL-D to end')
elif sys.argv[1] == 'server' and len(sys.argv) == 3:
startserver(sys.argv[2])
elif sys.argv[1] == 'client' and len(sys.argv) == 3:
startclient(sys.argv[2])
上面的代码在 运行 单独执行时符合我的预期：
> python3 mp-example.py server test
{'address': ('192.168.17.10', 50793), 'authkey': b'mptest'}
yv=Value(<class 'int'>, 99), ns.yv=999
yv=Value(<class 'int'>, 132), ns.yv=1221, ns=Namespace(name='m.Namespace: test', yv=1221)
Waiting for input (waiting for client), press CTRL-D to end
接下来，不起作用的位：
下面的代码在一定程度上起作用，但在尝试访问命名空间中的值项 yv 时失败。
代码：
def startclient(name):
# retrieve server identification from file
serverdata_filename = f'mpSERVERDATA.txt'
with open(serverdata_filename, 'rt', encoding="utf-8") as serverdata_file:
Server = eval(serverdata_file.read())
print(Server)
bm = mpm.SyncManager(**Server)
bm.register('yv', proxytype=mpm.ValueProxy)
bm.register('ns', proxytype=mpm.NamespaceProxy)
bm.connect()
yv = bm.yv()
print(f'yv={yv.value}')
yv.value += 55
print(f'yv={yv.value}')
ns = bm.ns()
print(f'ns={ns}')
print(f'ns.yv={ns.yv}')
ns.yv += 666
print(f'ns.yv={ns.yv}')
输出：
> python3 mp-example.py client test
{'address': ('192.168.17.10', 50793), 'authkey': b'mptest'}
yv=627
yv=682
['_Client', '__class__', '__deepcopy__', '__delattr__', '__dict__', '__dir__', '__doc__', '__eq__', '__format__', '__ge__', '__getattr__', '__getattribute__', '__gt__', '__hash__', '__init__', '__init_subclass__', '__le__', '__lt__', '__module__', '__ne__', '__new__', '__reduce__', '__reduce_ex__', '__repr__', '__setattr__', '__sizeof__', '__str__', '__subclasshook__', '__weakref__', '_address_to_local', '_after_fork', '_authkey', '_callmethod', '_close', '_connect', '_decref', '_exposed_', '_getvalue', '_id', '_idset', '_incref', '_manager', '_mutex', '_owned_by_manager', '_serializer', '_tls', '_token']
Help on method _getvalue in module multiprocessing.managers:
_getvalue() method of multiprocessing.managers.NamespaceProxy instance
Get a copy of the value of the referent
None
ns=<NamespaceProxy object, typeid 'Namespace' at 0x7f5d05689a20>
Traceback (most recent call last):
File "mp-example.py", line 333, in <module>
startclient(sys.argv[2])
File "mp-example.py", line 282, in startclient
print(f'ns.yv={ns.yv}')
File "/usr/lib/python3.6/multiprocessing/managers.py", line 1060, in __getattr__
return callmethod('__getattribute__', (key,))
File "/usr/lib/python3.6/multiprocessing/managers.py", line 772, in _callmethod
raise convert_to_error(kind, result)
AttributeError: 'NamespaceProxy' object has no attribute 'yv'
问题：
我需要做什么才能在客户端访问ns.yv？
我也为此苦苦挣扎——很难按照您所遵循的思路想出任何既不复杂又低效的解决方案。顺便说一句，您正在尝试 return 共享内存值的代理，该值已经可以共享而无需代理，并且如果您的目标只是提供一个单例值，则不需要它在共享内存中。
一个更简单的方法，它不需要您创建两个管理器进程，只是为了创建一个代理对象，然后在您启动它之前向另一个管理器注册，就是定义您自己的自定义 类.您通常会创建自己的 multiprocessing.managers.BaseManager 子类，您将在其中注册新的托管 类（当然，如果您的客户还需要已经在 SyncManager 中注册的 类，例如 dict，那么您可以将其子类化）。
以下程序展示了如何创建单例 Namespace 等效类型。我把你的IP地址注释掉了，用127.0.0.1来测试。我还用更安全的替代方法替换了您对 eval 的调用。
#!/usr/bin/env python3
import os
import sys
import multiprocessing as mp
from multiprocessing.managers import BaseManager, NamespaceProxy
import socket
import contextlib
from ast import literal_eval
from threading import Thread
def get_ip():
with contextlib.closing(socket.socket(socket.AF_INET, socket.SOCK_DGRAM)) as s:
#s.connect(('192.168.17.1', 80))
s.connect(('127.0.0.1', 80))
return s.getsockname()[0]
def get_free_port():
with contextlib.closing(socket.socket(socket.AF_INET, socket.SOCK_STREAM)) as s:
s.bind(('', 0))
s.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
return s.getsockname()[1]
class MyNamespaceManager(BaseManager):
pass
class MyNamespace:
pass
# Singleton:
singleton = MyNamespace()
def get_singleton():
return singleton
def run_server(server):
server.serve_forever()
def startserver():
# Find a free port for the SyncManager
newportnumber = get_free_port()
Server = {'address': (get_ip(), newportnumber), 'authkey': b'mptest'}
# Save data to file
serverdata_filename = f'mpSERVERDATA.txt'
with open(serverdata_filename, 'wt', encoding="utf-8") as serverdata_file:
print(Server, file=serverdata_file)
print(Server)
# On the server, we do not need our own IP address
Server['address'] = ('', newportnumber)
# Singleton
MyNamespaceManager.register('MyNamespace', callable=get_singleton, proxytype=NamespaceProxy)
manager = MyNamespaceManager(**Server)
"""
manager.start()
input('Hit enter to end: ')
manager.shutdown()
"""
server = manager.get_server()
Thread(target=run_server, args=(server,)).start()
try:
input('Hit enter to end: ')
except KeyboardInterrupt:
pass
os.unlink(serverdata_filename)
server.stop_event.set() # The server stops and calls sys.exit()
def startclient():
# retrieve server identification from file
serverdata_filename = f'mpSERVERDATA.txt'
with open(serverdata_filename, 'rt', encoding="utf-8") as serverdata_file:
Server = literal_eval(serverdata_file.read())
print(Server)
MyNamespaceManager.register('MyNamespace', proxytype=NamespaceProxy)
manager = MyNamespaceManager(**Server)
manager.connect()
yv1 = manager.MyNamespace()
yv1.value = 99
yv1.value += 55
yv1.x = 9
yv2 = manager.MyNamespace()
print(f'yv2.value={yv2.value}')
print(f'yv2.x={yv2.x}')
if __name__ == '__main__':
if len(sys.argv) == 2 and sys.argv[1] == 'server':
startserver()
else:
startclient()
客户端打印：
{'address': ('127.0.0.1', 51211), 'authkey': b'mptest'}
yv2.value=154
yv2.x=9
编辑-综合解决方案：
我通过编写一个新的 RemoteSyncManager class 来封装 multiprocessing.managers.SyncManager() 实例来解决我的大部分挫折。
我的这个库确实简化了 SyncManager 本地和远程进程的使用。
前往我的 GitHub 存储库 PythonLibraries-Multiprocessing-RemoteSyncManager 了解详细信息。
我的第一个解决方案：
我终于设法设置了一个与远程连接同步的托管多处理命名空间。
但是这个设置拒绝使用 Managed dict，它的工作方式与上面 Booboo 的建议类似。
有什么方法可以在远程共享命名空间中定义和保持同步整个字典，而无需在每次操作时重新分配字典？
代码：
if __name__ == '__main__':
if sys.argv[1] == 'Server' and len(sys.argv) == 2:
#localmanager = mp.Manager()
#with localmanager:
q1 = mp.Queue(1)
q2 = mp.Queue(1)
localnamespace = mpm.Namespace()
class syncmanager(mpm.SyncManager):
pass
syncmanager.register('get_q1', callable=lambda: q1)
syncmanager.register('get_q2', callable=lambda: q2)
syncmanager.register('get_ns', callable=lambda: localnamespace, proxytype=mpm.NamespaceProxy)
remotemanager = syncmanager(address = ('', 56789), authkey = b'mypassword')
remotemanager.start()
with remotemanager:
ns = remotemanager.get_ns()
#ns.sampledict = remotemanager.dict()
ns.sampledict = {}
#ns.sampledict = localmanager.dict()
ns.sampledict['testabc'] = 1234567
ns.sampledict['TESTABC'] = 9876
d = {}
d['testabc'] = 1234567
d['TESTABC'] = 9876
ns.sampledict2 = d
print(f'Server ns = {ns}')
print(q2.get())
q1.put(f'Server ns = {ns}')
ns.sampledict['testabc'] = 1
ns.sampleserverint = 98789
print(f'Server ns = {ns}')
print(q2.get())
q1.put(f'Server ns = {ns}')
elif sys.argv[1] == 'Client' and len(sys.argv) == 2:
class syncmanager(mpm.SyncManager):
pass
syncmanager.register('get_q1')
syncmanager.register('get_q2')
syncmanager.register('get_ns', proxytype=mpm.NamespaceProxy)
remotemanager = syncmanager(address = ('', 56789), authkey = b'mypassword')
remotemanager.connect()
q1 = remotemanager.get_q1()
q2 = remotemanager.get_q2()
ns = remotemanager.get_ns()
q2.put('ready')
msg = q1.get()
print(msg)
print(f'Client ns = {ns}')
ns.sampledict['TESTABC'] = 9
ns.sampleclientint = 7887
q2.put('next.1')
msg = q1.get()
print(msg)
print(f'Client ns = {ns}')