Re: cmusaslsecretPLAIN attribute

2007-07-04 Thread TechnoSophos
On 7/3/07, John Burian [EMAIL PROTECTED] wrote: [...] if you want a SASL bind with PLAIN mechanism and TLS, the ldapwhoami should look something like $ ldapwhoami -Y PLAIN -U burianj -ZZ -H ldap://localhost I'm not having a problem getting TLS to work. ldapwhoami is connecting over port

reference to other entry - question

2007-07-04 Thread Marcin Giedz
Is there any way to do something like this without copying all data many times. 1) there is entry: uid=user1,ou=people,dc=xx,dc=x The entry has objectClass = person, posixAccount etc.. So there is attribute userPassword. 2) there are entries: [EMAIL PROTECTED],ou=domains,dc=xx,dc=x [EMAIL

Lock is no longer valid / deferring operation

2007-07-04 Thread Toby Blake
Hi all, For largely historical reasons we run slapd servers on most clients (this will probably change in the future - I'm just giving this information as background). We're seeing problems when some of these machines are busy, particularly, it seems, with memory intensive activity, although

Re: index_param failed

2007-07-04 Thread JOYDEEP
Luca Scamoni wrote: JOYDEEP ha scritto: Dear list, from the log I am getting these errors like == = bdb_equality_candidates: (uid) index_param failed (18) = bdb_equality_candidates: (uid) index_param failed (18) = bdb_equality_candidates: (uidNumber) index_param failed

No such object error after converting from 2.0.27 to 2.3.32

2007-07-04 Thread Brian Gaber
Took the slapcat output from version 2.0.27 (ldbm) to version 2.3.32 (bdm). Used /usr/local/bin/slapadd on 2.3.32 and am using Berkeley 4.5.20. The slapadd works fine. Then I issued chown ldap:ldap on the /var/lib/ldap-2.3.32 directory and files. Any type of ldapsearch results in a 32 no such

Re: cmusaslsecretPLAIN attribute

2007-07-04 Thread John M. Burian
Buchan Milne wrote: But, SASL authentication does not use a DN, but a username (as provided in the example Dieter gave you above). And you would need to have configured slapd to map a SASL identity to a DN for the bind to succeed. I have an authz-regexp that maps SASL's

Re: force use of start_tls: how?

2007-07-04 Thread Hallvard B Furuseth
Andreas Hasenack writes: I'm trying to avoid mistakes and configure a server and/or client to force the use of start tls. So, if someone binds to the server and accidentally forgets to configure start_tls on the client, the connection is rejected. The problem is that the rejection happens

Re: No such object error after converting from 2.0.27 to 2.3.32

2007-07-04 Thread Buchan Milne
On Wednesday, 4 July 2007, Brian Gaber wrote: Took the slapcat output from version 2.0.27 (ldbm) to version 2.3.32 (bdm). Used /usr/local/bin/slapadd on 2.3.32 and am using Berkeley 4.5.20. The slapadd works fine. Then I issued chown ldap:ldap on the /var/lib/ldap-2.3.32 directory and files.

Re: force use of start_tls: how?

2007-07-04 Thread Andreas Hasenack
On Wed, Jul 04, 2007 at 05:53:24PM +0200, Hallvard B Furuseth wrote: The problem is that the rejection happens too late: the client password was already sent to the server in clear test. If you want to ensure it on the server side, all you can do is not listen for ldap:// connections since

Re: cmusaslsecretPLAIN attribute

2007-07-04 Thread Dieter Kluenter
John M. Burian [EMAIL PROTECTED] writes: Buchan Milne wrote: But, SASL authentication does not use a DN, but a username (as provided in the example Dieter gave you above). And you would need to have configured slapd to map a SASL identity to a DN for the bind to succeed. I have an

Re: force use of start_tls: how?

2007-07-04 Thread Hallvard B Furuseth
Andreas Hasenack writes: URI ldaps://fully.qualified.server-hostname/ TLS_CACERT file with the CA-certificate which signed the server cert TLS_REQCERT demand The only problem is that I really want start_tls, and not ldaps (which is deprecated, right?). Don't know. It's

Re: reference to other entry - question

2007-07-04 Thread Gavin Henry
quote who=Marcin Giedz Is there any way to do something like this without copying all data many times. 1) there is entry: uid=user1,ou=people,dc=xx,dc=x The entry has objectClass = person, posixAccount etc.. So there is attribute userPassword. 2) there are entries: [EMAIL

Re: Lock is no longer valid / deferring operation

2007-07-04 Thread Gavin Henry
quote who=Toby Blake Hi all, Hi Toby. For largely historical reasons we run slapd servers on most clients (this will probably change in the future - I'm just giving this information as background). Why? We're seeing problems when some of these machines are busy, particularly, it seems,

Re: reference to other entry - question

2007-07-04 Thread Marcin Giedz
Gavin Henry napisaƂ(a): quote who=Marcin Giedz Is there any way to do something like this without copying all data many times. 1) there is entry: uid=user1,ou=people,dc=xx,dc=x The entry has objectClass = person, posixAccount etc.. So there is attribute userPassword. 2) there are entries:

Re: Lock is no longer valid / deferring operation

2007-07-04 Thread Quanah Gibson-Mount
--On Wednesday, July 04, 2007 8:40 PM +0100 Gavin Henry [EMAIL PROTECTED] wrote: We're running openldap 2.3.35 with ITS#4924 and ITS#4925 patches with a bdb backend running 4.2.52 with all 6 recommended patches. I hope you mean 5, as there are only 5 listed on the Oracle site. There are 6

Re: Lock is no longer valid / deferring operation

2007-07-04 Thread Gabriel Stein
And about your DBCONFIG? Is everything ok? Cheers. On 7/4/07, Quanah Gibson-Mount [EMAIL PROTECTED] wrote: --On Wednesday, July 04, 2007 8:40 PM +0100 Gavin Henry [EMAIL PROTECTED] wrote: We're running openldap 2.3.35 with ITS#4924 and ITS#4925 patches with a bdb backend running 4.2.52

Re: force use of start_tls: how?

2007-07-04 Thread Philip Guenther
On Wed, 4 Jul 2007, Andreas Hasenack wrote: ... The only problem is that I really want start_tls, and not ldaps (which is deprecated, right?). Can't be done. The problem is that LDAP does not mandate that clients perform any sort of capability negotiation before performing a bind. Ergo,