plone site wedged when ldap server not reachable

[blackbronze]

     We have Plone 4.3.1 with PloneHotfix20130618 applied. With version Products.LDAPUserFolder-2.26-py2.7 when only one LDAP server is configured and isn't reachable plone upon startup (via ./bin/plonectl start) will results in a initially non-responsive site.

     When that LDAP server is unreachable anytime the site has been successfully running the site becomes completely unresponsive and requires a plonectl restart. IMHO a timeout shouldn't lock the site up.

     Let me know what additional information I need to provide.

2013-08-15T13:49:54 CRITICAL event.LDAPDelegate Failure connecting, last attempted server: ldaps://ldap.mydomain.com:636 ({'desc': "Can't contact LDAP server"})
Traceback (most recent call last):
  File "/usr/local/Plone/buildout-cache/eggs/Products.LDAPUserFolder-2.26-py2.7.egg/Products/LDAPUserFolder/LDAPDelegate.py", line 240, in connect
    , op_timeout=server['op_timeout']
  File "/usr/local/Plone/buildout-cache/eggs/Products.LDAPUserFolder-2.26-py2.7.egg/Products/LDAPUserFolder/LDAPDelegate.py", line 342, in _connect
    connection.simple_bind_s(user_dn, user_pwd)
  File "/usr/local/Plone/buildout-cache/eggs/python_ldap-2.4.13-py2.7-linux-x86_64.egg/ldap/ldapobject.py", line 834, in simple_bind_s
    res = self._apply_method_s(SimpleLDAPObject.simple_bind_s,*args,**kwargs)
  File "/usr/local/Plone/buildout-cache/eggs/python_ldap-2.4.13-py2.7-linux-x86_64.egg/ldap/ldapobject.py", line 820, in _apply_method_s
    self.reconnect(self._uri)
  File "/usr/local/Plone/buildout-cache/eggs/python_ldap-2.4.13-py2.7-linux-x86_64.egg/ldap/ldapobject.py", line 787, in reconnect
    self._apply_last_bind()
  File "/usr/local/Plone/buildout-cache/eggs/python_ldap-2.4.13-py2.7-linux-x86_64.egg/ldap/ldapobject.py", line 759, in _apply_last_bind
    SimpleLDAPObject.simple_bind_s(self,'','')
  File "/usr/local/Plone/buildout-cache/eggs/python_ldap-2.4.13-py2.7-linux-x86_64.egg/ldap/ldapobject.py", line 207, in simple_bind_s
    msgid = self.simple_bind(who,cred,serverctrls,clientctrls)
  File "/usr/local/Plone/buildout-cache/eggs/python_ldap-2.4.13-py2.7-linux-x86_64.egg/ldap/ldapobject.py", line 201, in simple_bind
    return self._ldap_call(self._l.simple_bind,who,cred,RequestControlTuples(serverctrls),RequestControlTuples(clientctrls))
  File "/usr/local/Plone/buildout-cache/eggs/python_ldap-2.4.13-py2.7-linux-x86_64.egg/ldap/ldapobject.py", line 99, in _ldap_call
    result = func(*args,**kwargs)
SERVER_DOWN: {'desc': "Can't contact LDAP server"}

[Dylan Jay (t-launchpad-dylanjay-com)]

This is related to a bug in python-ldap that is still being fixed.

[blackbronze (blackbronze979)]

Thanks Dylan! I looked for the bug report and couldn't find it. Would you mind posting where you found it at?

[gwarah (lhmp1967)]

I've got a similar problem. Our ldap server has changed IP, so the webserver downs after plone-ldap try to connect ldap server. Nothing works even ZMI. Is there another way than ZMI to reconfigure the new ldap server? Thanks for any tip.

[Jens Vagelpohl (dataflake-deactivatedaccount-deactivatedaccount)]

Please see the previous comments. Also, have you reduced the timeout values set for the respective LDAP connection in the LDAPUserFolder server settings?

[gwarah (lhmp1967)]

I solve the bug related on my last post just deleting ldap plugin directly on instance and activating again on site.

more details, see stack overflow question #37681319