Login timeouts if ldapserver is unavailable

If ldap is configured but the ldap server is unreachable, the login hangs until the worker terminates.

added part::backend prio::1 source::customer type::bug labels

changed the description

assigned to @nik

created merge request !844 (merged) to address this issue

mentioned in merge request !844 (merged)

The bug is indeed reproducible by simply setting some imaginary LDAP server when running the dev server: poetry run env ALEKSIS_database__password=aleksis ALEKSIS_ldap__uri=ldap://8.8.8.8 ALEKSIS_ldap__users__search__base=dc=foo aleksis-admin runuwsgi.

Hangs upon bind here:

 --- modulename: ldapobject, funcname: simple_bind_s
ldapobject.py(248):     msgid = self.simple_bind(who,cred,serverctrls,clientctrls)
 --- modulename: ldapobject, funcname: simple_bind
ldapobject.py(242):     return self._ldap_call(self._l.simple_bind,who,cred,RequestControlTuples(serverctrls),RequestControlTuples(clientctrls))
 --- modulename: __init__, funcname: RequestControlTuples
__init__.py(113):   if ldapControls is None:
__init__.py(114):     return None
 --- modulename: __init__, funcname: RequestControlTuples
__init__.py(113):   if ldapControls is None:
__init__.py(114):     return None
 --- modulename: ldapobject, funcname: _ldap_call
ldapobject.py(114):     self._ldap_object_lock.acquire()
 --- modulename: __init__, funcname: acquire
__init__.py(85):       if _trace_level>=self._min_trace_level:
__init__.py(87):     return self._lock.acquire()
ldapobject.py(116):       if self._trace_level>=1:
ldapobject.py(125):     diagnostic_message_success = None
ldapobject.py(126):     try:
ldapobject.py(127):       try:
ldapobject.py(128):         result = func(*args,**kwargs)

closed with merge request !844 (merged)

mentioned in commit 74de127a