parallel (concurrent) eventlet

David Gabriel davidgab283 at gmail.com
Mon Jan 18 06:03:13 EST 2016


Dears,

I have an issue when I use eventlet Api to create parallel threads.
In fact, when I run the below code, only the program dealing with the
synchronozation with ldap data base is working and is continuously blocking
the others to run.
But, when I use the 'thread' Api the program is working fine without any
blocking issue. However, I can not use thread Api and I must to use
eventlet.
So I am wondering how to get the thread Api behavior using the eventlet Api
?

Could you please inform me how to fix this issue ?

Kindly find below my code.
But you need some configurations regarding ldap server/client.

#!/usr/bin/python
# -*- coding: utf-8 -*-
"""
This script implements a syncrepl consumer which syncs data from an OpenLDAP
server to a local (shelve) database.
Notes:
The bound user needs read access to the attributes entryDN and entryCSN.
This needs the following software:
Python
pyasn1 0.1.4+
pyasn1-modules
python-ldap 2.4.10+
"""

# Import the python-ldap modules
import ldap,ldapurl
# Import specific classes from python-ldap
from ldap.ldapobject import ReconnectLDAPObject
from ldap.syncrepl import SyncreplConsumer

# Import modules from Python standard lib
import shelve,signal,time,sys,logging
import eventlet
#import thread
eventlet.monkey_patch()

# Global state
watcher_running = True
ldap_connection = False



class SyncReplConsumer(ReconnectLDAPObject,SyncreplConsumer):
    """
    Syncrepl Consumer interface
    """
    def __init__(self,db_path,*args,**kwargs):
        # Initialise the LDAP Connection first
        ldap.ldapobject.ReconnectLDAPObject.__init__(self, *args, **kwargs)
        # Now prepare the data store
        self.__data = shelve.open(db_path, 'c')
        # We need this for later internal use
        self.__presentUUIDs = dict()

    def __del__(self):
            # Close the data store properly to avoid corruption
            self.__data.close()

    def syncrepl_get_cookie(self):
        if 'cookie' in self.__data:
            return self.__data['cookie']

    def syncrepl_set_cookie(self,cookie):
        self.__data['cookie'] = cookie

    def syncrepl_entry(self,dn,attributes,uuid):

        # First we determine the type of change we have here (and store
away the previous data for later if needed)
        previous_attributes = dict()
        if uuid in self.__data:
            change_type = 'modify'
            previous_attributes = self.__data[uuid]
        else:
            change_type = 'add'
        # Now we store our knowledge of the existence of this entry
(including the DN as an attribute for convenience)
        attributes['dn'] = dn
        self.__data[uuid] = attributes
        # Debugging
        print 'Detected', change_type, 'of entry:', dn
        # If we have a cookie then this is not our first time being run, so
it must be a change
        if 'ldap_cookie' in self.__data:
                self.perform_application_sync(dn, attributes,
previous_attributes)

    def syncrepl_delete(self,uuids):
        # Make sure we know about the UUID being deleted, just in case...
        uuids = [uuid for uuid in uuids if uuid in self.__data]
        # Delete all the UUID values we know of
        for uuid in uuids:
            print 'Detected deletion of entry:', self.__data[uuid]['dn']
            del self.__data[uuid]

    def syncrepl_present(self,uuids,refreshDeletes=False):
        # If we have not been given any UUID values, then we have recieved
all the present controls...
        if uuids is None:
            # We only do things if refreshDeletes is false as the syncrepl
extension will call syncrepl_delete instead when it detects a delete notice
            if refreshDeletes is False:
                deletedEntries = [uuid for uuid in self.__data.keys() if
uuid not in self.__presentUUIDs and uuid != 'ldap_cookie']
                self.syncrepl_delete( deletedEntries )
            # Phase is now completed, reset the list
            self.__presentUUIDs = {}
        else:
            # Note down all the UUIDs we have been sent
            for uuid in uuids:
                    self.__presentUUIDs[uuid] = True

    def perform_application_sync(self,dn,attributes,previous_attributes):
        print 'Performing application sync for:', dn
        return True


# Shutdown handler
#def commenceShutdown(signum, stack):
def commenceShutdown():
    # Declare the needed global variables
    global watcher_running, ldap_connection
    print 'Shutting down!'

    # We are no longer running
    watcher_running = False

    # Tear down the server connection
    if( ldap_connection ):
            del ldap_connection

    # Shutdown
    sys.exit(0)

def mainOfSyncrepl(threadName):
     # Time to actually begin execution
     # Install our signal handlers
#     signal.signal(signal.SIGTERM,commenceShutdown)
#     signal.signal(signal.SIGINT,commenceShutdown)
     try:
       ldap_url =
ldapurl.LDAPUrl('ldap://localhost/dc=example,dc=org?*?sub?(objectClass=*)?bindname=cn=admin%2cdc=test%2cdc=com,X-BINDPW=myPassword')#ldapurl.LDAPUrl(sys.argv[1])
     #  ldap_url = ldapurl.LDAPUrl(link)
       database_path = 'test.com'#sys.argv[2]
     #  database_path = pathName
     except IndexError,e:
       print 'Usage: syncrepl-client.py <LDAP URL> <pathname of database>'
       sys.exit(1)
     except ValueError,e:
       print 'Error parsing command-line arguments:',str(e)
       sys.exit(1)

     while watcher_running:
         print 'Connecting to LDAP server now...'
         # Prepare the LDAP server connection (triggers the connection as
well)
         ldap_connection =
SyncReplConsumer(database_path,ldap_url.initializeUrl())

         # Now we login to the LDAP server
         try:
             ldap_connection.simple_bind_s(ldap_url.who,ldap_url.cred)
         except ldap.INVALID_CREDENTIALS, e:
             print 'Login to LDAP server failed: ', str(e)
             sys.exit(1)
         except ldap.SERVER_DOWN:
             continue

         # Commence the syncing
         print 'Commencing sync process'
         ldap_search = ldap_connection.syncrepl_search(
           ldap_url.dn or '',
           ldap_url.scope or ldap.SCOPE_SUBTREE,
           mode = 'refreshAndPersist',
           filterstr = ldap_url.filterstr or '(objectClass=*)')
         print 'After syncrepl_search.'
         try:
             while ldap_connection.syncrepl_poll( all = 1, msgid =
ldap_search):
                  pass
         except KeyboardInterrupt:
          # User asked to exit
             commenceShutdown()
             pass
         except Exception, e:
          # Handle any exception
             if watcher_running:
                 print 'Encountered a problem, going to retry. Error:',
str(e)
                 eventlet.sleep(5)
             pass

# Define a function for the 2nd thread
def print_time(ThreadName):
     count = 0
     delay = 3
     while 1:#count < 5:
         count += 1
         print "%s: %s" % (ThreadName, time.ctime(time.time()) )
         eventlet.sleep(delay)



print 'Before call threads'

evt1 = eventlet.spawn(mainOfSyncrepl, "Thread-1",)
evt2 = eventlet.spawn(print_time, "Thread-2",)
evt3 = eventlet.spawn(print_time, "Thread-3",)

print 'After call threads'

evt1.wait()
evt2.wait()
evt3.wait()

print 'After wait'



2016-01-12 7:20 GMT-08:00 Ned Batchelder <ned at nedbatchelder.com>:

> David,
>
> We aren't going to be able to debug code that we can't see.  Please post a
> link to the *actual* code that you are running.
>
> --Ned.
>
>
> On 1/12/16 7:00 AM, David Gabriel wrote:
>
> Dears
>
> For more details, I am using this code
> <https://github.com/rbarrois/python-ldap/blob/master/Demo/pyasn1/syncrepl.py>in
> order to ensure the updates from my data base.
> However, when I create an eventlet basing on this code, my program is
> blocked there and is not running other eventlets !!!
> Please advise me how to fix this issue ?
>
> Thanks in advance.
> Regards
>
> 2016-01-11 7:29 GMT-08:00 David Gabriel <davidgab283 at gmail.com>:
>
>> Thanks so much John
>> In fact your proposal works fine for this simple example but when I use
>> it for a complex code (a data base client that receives all updates from
>> the db), my program is continously running this db client and not other
>> programs.
>>
>> Any suggestions
>> Thanks in advance
>> regards
>>
>> 2016-01-11 5:50 GMT-08:00 John Eskew < <john.eskew at gmail.com>
>> john.eskew at gmail.com>:
>>
>>> Add this line below your imports:
>>>
>>> eventlet.monkey_patch()
>>>
>>> Here's why that line should fix things:
>>>
>>> http://eventlet.net/doc/patching.html#greening-the-world
>>>
>>> On Mon, Jan 11, 2016 at 6:27 AM, David Gabriel < <davidgab283 at gmail.com>
>>> davidgab283 at gmail.com> wrote:
>>>
>>>> Dears,
>>>> It is the first time I am developping with python and I want to execute
>>>> parallel threads using eventlet.When I run the below code, only one thread
>>>> is executed and not both.Please could you tell me how to fix this issue ?
>>>> Please advise me how to ensure a concurrent behavior between evt1 and
>>>> evt2.
>>>>
>>>> import eventlet
>>>> import time
>>>>
>>>> def print_time(ThreadName):
>>>>         count = 0
>>>>         delay = 3
>>>>         while 1:#count < 5:
>>>>             count += 1
>>>>             print "%s: %s" % (ThreadName, time.ctime(time.time()) )
>>>>             time.sleep(delay)
>>>>
>>>> print 'Before call threads'
>>>>
>>>> evt1 = eventlet.spawn(print_time, "Thread-1",)
>>>> evt2 = eventlet.spawn(print_time, "Thread-2",)
>>>>
>>>> print 'After call threads'
>>>>
>>>> evt1.wait()
>>>> evt2.wait()
>>>>
>>>>
>>>>
>>>> Any answer is welcome.
>>>> Thanks in advance.
>>>> Regards.
>>>>
>>>> _______________________________________________
>>>> Boston mailing list
>>>> Boston at python.org
>>>> https://mail.python.org/mailman/listinfo/boston
>>>>
>>>>
>>>
>>
>
>
> _______________________________________________
> Boston mailing listBoston at python.orghttps://mail.python.org/mailman/listinfo/boston
>
>
>
> _______________________________________________
> Boston mailing list
> Boston at python.org
> https://mail.python.org/mailman/listinfo/boston
>
>



More information about the Python-list mailing list