Closed GoogleCodeExporter closed 8 years ago
That attribute indicates that the next operation performed on the connection
must be a password change.
However, it does not affect the initial bind.
From your logs:
> Sep 24 10:54:40 pupubuntu slapd[5207]: conn=1014 op=0 BIND
dn="uid=test,ou=people,dc=vbox,dc=it" mech=SIMPLE ssf=0
> Sep 24 10:54:40 pupubuntu slapd[5207]: conn=1014 op=0 RESULT tag=97 err=0
text=
err=0 means success
Since authentication only performs a single bind and then closes, you never get
the error you are looking for.
From a practical sense, how could someone change their password if they cannot
bind?
Your application logic will need to detect the CHANGE_AFTER_RESET and then take
the user to your IDM password change interface.
Original comment by dfis...@gmail.com
on 24 Sep 2013 at 3:26
Do you have any further comments or questions on this issue?
Original comment by dfis...@gmail.com
on 30 Sep 2013 at 4:04
Hello Fisher,
you're absolutely right, we implemented a solution which is able to solve our
specific problem (yet it is not probably general purpose for all).
We extended the JASS module LdapLoginModule in order to verify the presence of
the attribute CHANGE_AFTER_RESET.
In this case we deny access, and return to the login page a specific error
message so that login.jsp can provide a proper message to the user.
This solution is good for use since our IDM password change interface do not
use Java JAAS module to authenticate users. So it is perfectly good for us to
deny access to users with CHANGE_AFTER_RESET.
To make our solution more general purpose, we may decide to change our
implementation and permit user login to users with CHANGE_AFTER_RESET and raise
"something" that could be used as a flag for the login.jsp to verify if the
user is obliged to reset his password.
For the moment we put in place our solution (that blocks logins for users with
CHANGE_AFTER_RESET flag in LDAP).
Regards,
Marco
Original comment by marco.ma...@gmail.com
on 1 Oct 2013 at 7:58
Thanks for the feedback Marco.
Original comment by dfis...@gmail.com
on 1 Oct 2013 at 4:47
Original issue reported on code.google.com by
marco.ma...@gmail.com
on 24 Sep 2013 at 12:03Attachments: