openldap-technical

openldap-technical@openldap.org

26 participants
9849 discussions

Getting details for "TLS trace: SSL3 alert read:fatal:unsupported certificate"
by Windl, Ulrich 06 Mar '25

06 Mar '25

Hi! After "playing" significant time with certificate authentication, I managed to authenticate one user. However when I tried to authenticate a different user with a similar certificate, I see a TLS trace: SSL3 alert read:fatal:unsupported certificate Error. Can I can some details about the "usupportedness" of my certificate? The only thing I could think if is that uid of the newer certificate has a CN that is three characters longer than the one that worked. A more complete trace for ldapwhoami woul look like this: ... ldap_parse_result ber_scanf fmt ({iAA) ber: ber_scanf fmt (}) ber: ldap_msgfree TLS trace: SSL_connect:before SSL initialization TLS trace: SSL_connect:SSLv3/TLS write client hello TLS trace: SSL_connect:SSLv3/TLS write client hello TLS trace: SSL_connect:SSLv3/TLS read server hello TLS trace: SSL_connect:TLSv1.3 read encrypted extensions TLS trace: SSL_connect:SSLv3/TLS read server certificate request TLS certificate verification: depth: 2, err: 0, subject: /.... Root-CA (2018), issuer: /... Root-CA (2018) TLS certificate verification: depth: 1, err: 0, subject: /... Host-CA (2018), issuer: /... Root-CA (2018) TLS certificate verification: depth: 0, err: 0, subject: /... FQHN, issuer: /... Host-CA (2018) TLS trace: SSL_connect:SSLv3/TLS read server certificate TLS trace: SSL_connect:TLSv1.3 read server certificate verify TLS trace: SSL_connect:SSLv3/TLS read finished TLS trace: SSL_connect:SSLv3/TLS write change cipher spec TLS trace: SSL_connect:SSLv3/TLS write client certificate TLS trace: SSL_connect:SSLv3/TLS write certificate verify TLS trace: SSL_connect:SSLv3/TLS write finished ldap_sasl_interactive_bind: user selected: EXTERNAL ldap_int_sasl_bind: EXTERNAL ldap_int_sasl_open: host=FQHN SASL/EXTERNAL authentication started ldap_sasl_bind ldap_send_initial_request ldap_send_server_request ber_scanf fmt ({it) ber: ber_scanf fmt ({i) ber: ber_flush2: 26 bytes to sd 3 ldap_msgfree ldap_result ld 0x56432476ac30 msgid 2 wait4msg ld 0x56432476ac30 msgid 2 (infinite timeout) wait4msg continue ld 0x56432476ac30 msgid 2 all 1 ** ld 0x56432476ac30 Connections: * host: FQHN port: 389 (default) * from: IP=172.20.16.36:57868 refcnt: 2 status: Connected last used: Wed Mar 5 15:42:03 2025 ** ld 0x56432476ac30 Outstanding Requests: * msgid 2, origid 2, status InProgress outstanding referrals 0, parent count 0 ld 0x56432476ac30 request count 1 (abandoned 0) ** ld 0x56432476ac30 Response Queue: Empty ld 0x56432476ac30 response count 0 ldap_chkResponseList ld 0x56432476ac30 msgid 2 all 1 ldap_chkResponseList returns ld 0x56432476ac30 NULL ldap_int_select read1msg: ld 0x56432476ac30 msgid 2 all 1 ber_get_next TLS trace: SSL3 alert read:fatal:unsupported certificate ber_get_next failed, errno=0. ldap_err2string ldap_sasl_interactive_bind: Can't contact LDAP server (-1) ... Kind regards, Ulrich Windl

3 5

Q: Uniqueness of syncrepl rid
by Windl, Ulrich 04 Mar '25

04 Mar '25

Hi! I have a question on the uniqueness of rid in syncrepl: Do they have to be globally unique, or just unique within a database (olcDatabase config)? Kind regards, Ulrich Windl

4 5

OlcLimits : sizelimit exceeded (openldap 2.5.16)
by bourguijl＠gmail.com 04 Mar '25

04 Mar '25

Dears, I've an issue with one user for which I set its limits to "unlimited" but which get err=4 (sizelimit exceeded). This is in the context of a META ldap fronted to a ldap backend. If I do the ldapsearch of this user in CLI directly on the ldap backend where the OlcLimits are set for it, this works perfectly, no err=4 errors but if this ldapsearch comes from the META ldap frontend with the same user which is used in the olcDbIDAssertBind as Binddn, it's failling with err=4 !!?? I did check ldap logs at back end and they are identical in both cases except the err=4. I don't see what can be the cause ?? Any help will be appreciate !! Thx in advance, Jean-Luc.

2 1

How to start debugging olcAuthzRegexp?
by Windl, Ulrich 04 Mar '25

04 Mar '25

Hi! I tried to remove the credentials from my syncrepl configuration using certificate authentication instead. To do so I added a user certificate for my own user and tried ldapwhoami to verify that it works. Unfortunately it does not. I read quite a lot on the subject, and either all the descriptions are all poorly written and incomplete, or it must be very simple to get it running. However I failed so far. My suspect is that my olcAuthzRegexp does not properly map the certificate's name to the user, or the mapping is not called at all. Can anybody provide a sample configuration for the client user to verify the configuration, and maybe give an example on the server side to get it working. What I have tried so far is having a ~/ldaprc with: TLS_REQCERT demand TLS_CACERT ./User-CA.crt TLS_CERT ./uid=user.crt TLS_KEY ./uid=user.pem LDAPSASL_MECH external And I tried the command "ldapwhoami -H ldap://FQHN -D uid=user,cn=gssapi,cn=auth -Z -v" I tried these olcAuthzRegexp: olcAuthzRegexp: {1} "C=DE,...,O=...,uid=([^,]+)" uid=$1,ou=people,dc=...,dc==de olcAuthzRegexp: {2} "^uid=([^,]+),cn=gssapi,cn=auth$" uid=$1,ou=people,dc=...,dc=de (I left out the details of the certificate and directory contexts) Kind regards, Ulrich Windl

4 3

manual page slapd-mdb.5 and "maxsize"
by Windl, Ulrich 03 Mar '25

03 Mar '25

Migrating to OpenLDAP 2.5 and MDB I hit the default 10MB limit (when trying to import a 6.8MB LDIF). Seeking how to extend the size online, I found: maxsize <bytes><https://manpages.debian.org/testing/slapd/slapd-mdb.5.en.html#maxsize> Specify the maximum size of the database in bytes. A memory map of this size is allocated at startup time and the database will not be allowed to grow beyond this size. The default is 10485760 bytes. This setting may be changed upward if the configured limit needs to be increased. However it's completely non-obvious that the name for cn=config is "olcDbMaxSize" (and not "olcMaxssize"). (The final MDB file size was about 23MB, including the indices) Kind regards, Ulrich Windl

4 6

[authcid=<identity>] [authzid=<identity>]
by Stefan Kania 03 Mar '25

03 Mar '25

hi to all, I like to split the replication for cn=config and the object-db between to different kerberos-principals so I did the following (so far) 1. create to principals and a keytab for both of them. 2. configured k5start to get two different cchache files (one for each user) 3. Now I want to change my syncrepl from: --------------------------- dn: olcDatabase={0}config,cn=config changetype: modify replace: olcSyncRepl olcSyncRepl: rid=001 provider=ldaps://provider01.example.net bindmethod=sasl saslmech=gssapi searchbase="cn=config" type=refreshAndPersist retry="5 5 100 +" timeout=1 tls_reqcert=allow olcSyncRepl: rid=002 provider=ldaps://provider02.example.net bindmethod=sasl saslmech=gssapi searchbase="cn=config" type=refreshAndPersist retry="5 5 100 +" timeout=1 tls_reqcert=allow --------------------------- And ----------------------------- dn: olcDatabase={2}mdb,cn=config changetype: modify replace: olcSyncrepl olcSyncrepl: rid=101 provider=ldaps://provider01.example.net bindmethod=sasl saslmech=gssapi authzid=uid=krepl-main,cn=gssapi,cn=auth timeout=0 network-timeout=0 filter="(objectclass=*)" searchbase="dc=example,dc=net" scope=sub schemachecking=off type=refreshAndPersist retry="60 +" tls_reqcert=allow olcSyncrepl: rid=102 provider=ldaps://provider02.example.net bindmethod=sasl saslmech=gssapi authzid=uid=krepl-main,cn=gssapi,cn=auth timeout=0 network-timeout=0 filter="(objectclass=*)" searchbase="dc=example,dc=net" scope=sub schemachecking=off type=refreshAndPersist retry="60 +" keepalive=240:10:30 tls_reqcert=allow ----------------------------------- to use a different identity. I found the options [authcid=<identity>] [authzid=<identity>] But nothing how to configure these options. When I get the ticket for the user and do a ldapwhoami I'm getting dn:uid=krepl-config,cn=gssapi,cn=auth and dn:uid=krepl-main,cn=gssapi,cn=auth The DNs are: krbPrincipalName=krepl-krbPrincipalName=krepl-config(a)EXAMPLE.NET,cn=EXAMPLE.NET,cn=kerberos,dc=example,dc=net and krbPrincipalName=krepl-main(a)EXAMPLE.NET,cn=EXAMPLE.NET,cn=kerberos,dc=example,dc=net So which option du I have to take and what is the right value for the option? Stefan

3 5

When init new consumers receiving: missing or invalid at 'pwdPolicySubentry'
by erikdewaard＠gmail.com 03 Mar '25

03 Mar '25

Hi, When bootstrapping consumers through replication (clean conf/data dir) receiving msgs about pwdPolicySubentry missing. Its not always the same: sometimes 4 msgs another time 60 etc. I have this on various consumers 2.5.17/2.6.7 I thought this could be related to some ordering, maybe the accounts were created before ppolicy cn=noexpirepasswordpolicy data was replicated? <167>2025-02-27T12:16:20.993758+01:00 consumer22 slapd[1098021]: slapd starting <167>2025-02-27T12:16:32.035015+01:00 consumer22 slapd[1098021]: ppolicy_get: policy subentry cn=noexpirepasswordpolicy,ou=policies,dc=company,dc=com missing or invalid at 'pwdPolicySubentry', no policy will be applied! <167>2025-02-27T12:16:35.580938+01:00 consumer22 slapd[1098021]: ppolicy_get: policy subentry cn=noexpirepasswordpolicy,ou=policies,dc=company,dc=com missing or invalid at 'pwdPolicySubentry', no policy will be applied! best regards

2 1

OpenLDAP Pass-through Authentication
by Dino Edwards 01 Mar '25

01 Mar '25

Hi, Trying to get pass-through authentication working however, I'm running to the following error in OpenLDAP: openldap | 679ceede.3aa31e0a 0x7f2ff617e6c0 conn=1004 op=1 SRCH attr=uid mail displayName openldap | 679ceede.3aa4b816 0x7f2ff617e6c0 conn=1004 op=1 SEARCH RESULT tag=101 err=0 qtime=0.000019 etime=0.000172 nentries=1 text= openldap | 679ceede.3aaafd88 0x7f2ff597d6c0 conn=1005 fd=14 ACCEPT from IP=172.16.32.1:47082 (IP=0.0.0.0:1389) openldap | 679ceede.3aab8f34 0x7f2ff597d6c0 conn=1005 op=0 BIND dn="cn=username(a)domain.tld,ou=users,dc=domain,dc=local <mailto:cn=username@domain.tld,ou=users,dc=domain,dc=local> " method=128 openldap | 679ceede.3aac9267 0x7f2ff597d6c0 SASL [conn=1005] Failure: cannot connect to saslauthd server: No such file or directory openldap | 679ceede.3aad25b9 0x7f2ff597d6c0 conn=1005 op=0 RESULT tag=97 err=49 qtime=0.000005 etime=0.000120 text= openldap | 679ceede.3ab0191f 0x7f2ff617e6c0 conn=1005 fd=14 closed (connection lost) openldap | 679ceede.3ab3d76a 0x7f2ff597d6c0 conn=1004 fd=13 closed (connection lost) I followed the admin docs at: https://www.openldap.org/doc/admin26/guide.html#Pass-Through%20authenticatio n Here's what I did: Created /usr/lib/sasl2/slapd.conf with the following content: mech_list: plain pwcheck_method: saslauthd saslauthd_path: /var/run/sasl2/mux Created /etc/saslauthd.conf with the following content: ldap_servers: ldap://192.168.xxx.xxx.xxx ldap_search_base: OU=Users,DC=domain,DC=tld ldap_filter: (uid=%u) ldap_bind_dn: CN=saslauthd,CN=Users,DC=domain,DC=tld ldap_password: somepassword Added a user that already exists in domain.tld in openldap with the following password: {SASL}username(a)domain.tld <mailto:%7bSASL%7dusername@domain.tld> I would appreciate some help on this. Thanks

5 20

Transparent proxy to existing LDAP but with replacing specific attribute from another source
by Alexey D. Filimonov 28 Feb '25

28 Feb '25

Hi. A have Active Directory domain across 30+ domain controllers. In one single location, I have a server with corporate software that uses AD as normal ordinary LDAP with simple binds and it takes users. I want to add users their high quality photos, but I don't want them across all domain controllers in active directory. Can I build a proxy server with OpenLDAP that will proxy all requests to backend AD servers and return results, but for results it will replace one attribute, jpegPhoto, if it was requested, with some binary contents taken from elsewhere ?

2 1

olcDbMaxSize decrease possible?
by cyril＠stoll.info 27 Feb '25

27 Feb '25

Hi there For some reason I don't remember I set the olcDbMaxSize to 4 GB. Turns out that is way too large for our OpenLDAP servers. Both the admin guide and the slapd-mdb manpage state that "this setting may be changed upward". Unfortunately there is no mentioning of decreasing that size. On our servers the data.mdb and lock.mdb Files together are around 176 MB. I don't really expect a significant growth of our OpenLDAP Servers. So I guess that setting it to 1 GB or even just 500 MB would be more than enough for our needs in the coming years. So my question is, can I also decrease that size? Or should I just let it be at 4 GB even though we will never ever come close to that? Thanks a lot and best regards, Cyril

2 1

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

openldap-technical