hi,

thanks for replying. Yes, we had seen this already, that is also why I was pointing to the rhel 8 idm server in krb5.conf en sssd.conf

I missed the dns_lookup_kdc = false directive though, in krb5.conf [libdefaults] section, I modified it, tried but still not working.

The rest points to the one host.

#File modified by ipa-client-install

includedir /etc/krb5.conf.d/
[libdefaults]
  default_realm = IDM.DOMAIN.LOCAL
  dns_lookup_realm = true
  rdns = false
  dns_canonicalize_hostname = false

  dns_lookup_kdc = false
  ticket_lifetime = 24h
  forwardable = true
  udp_preference_limit = 0
  default_ccache_name = KEYRING:persistent:%{uid}

[realms]                                                                    
  IDM.DOMAIN.LOCAL = {                                      
    kdc =ds.idm.domain.local:88                               
    master_kdc = ds.idm.domain.local:88                 
    admin_server = ds.idm.domain.local:749            
    kpasswd_server = ds.idm.domain.local:464        
    default_domain = idm.domain.local                     
    pkinit_anchors = FILE:/var/lib/ipa-client/pki/kdc-ca-bundle.pem         
    pkinit_pool = FILE:/var/lib/ipa-client/pki/ca-bundle.pem    
  }


and this:

# grep ipa_server /etc/sssd/sssd.conf
ipa_server = ds.idm.domain.local

When I try ssh from a remote host, I see that sssd tries getting info for a trusted user:

(2024-03-22 16:42:35): [be[idm.domain.local]] [dp_get_account_info_send] (0x0200): Got request for [0x1][BE_REQ_USER][name=user@ad]
(2024-03-22 16:42:35): [be[idm.domain.local]] [dp_attach_req] (0x0400): [RID#8] DP Request [Account #8]: REQ_TRACE: New request. [sssd.nss CID #4] Flags [0x0001].
(2024-03-22 16:42:35): [be[idm.domain.local]] [dp_attach_req] (0x0400): [RID#8] Number of active DP request: 1
(2024-03-22 16:42:35): [be[idm.domain.local]] [sss_domain_get_state] (0x1000): [RID#8] Domain idm.domain.local is Active
(2024-03-22 16:42:35): [be[idm.domain.local]] [sss_domain_get_state] (0x1000): [RID#8] Domain domain.local is Active
(2024-03-22 16:42:35): [be[idm.domain.local]] [sss_domain_get_state] (0x1000): [RID#8] Domain idm.domain.local is Active
(2024-03-22 16:42:35): [be[idm.domain.local]] [sss_domain_get_state] (0x1000): [RID#8] Domain domain.local is Active
(2024-03-22 16:42:35): [be[idm.domain.local]] [sss_domain_get_state] (0x1000): [RID#8] Domain idm.domain.local is Active
(2024-03-22 16:42:35): [be[idm.domain.local]] [sss_domain_get_state] (0x1000): [RID#8] Domain domain.local is Active
(2024-03-22 16:42:35): [be[idm.domain.local]] [sdap_id_op_connect_step] (0x4000): [RID#8] reusing cached connection
(2024-03-22 16:42:35): [be[idm.domain.local]] [sdap_id_conn_data_not_idle] (0x4000): [RID#8] Marking connection as not idle
(2024-03-22 16:42:35): [be[idm.domain.local]] [ipa_s2n_get_acct_info_send] (0x0400): [RID#8] Sending request_type: [REQ_FULL_WITH_MEMBERS] for trust user [user@ad] to IPA server
(2024-03-22 16:42:35): [be[idm.domain.local]] [ipa_s2n_exop_send] (0x0400): [RID#8] Executing extended operation
(2024-03-22 16:42:35): [be[idm.domain.local]] [ipa_s2n_exop_send] (0x2000): [RID#8] ldap_extended_operation sent, msgid = 22
(2024-03-22 16:42:35): [be[idm.domain.local]] [sdap_op_add] (0x2000): [RID#8] New operation 22 timeout 6
(2024-03-22 16:42:35): [be[idm.domain.local]] [sdap_process_result] (0x2000): Trace: sh[0x55bcce14e400], connected[1], ops[0x55bcce119870], ldap[0x55bcce14c1c0]
(2024-03-22 16:42:35): [be[idm.domain.local]] [sdap_process_message] (0x4000): [RID#8] Message type: [LDAP_RES_EXTENDED]
(2024-03-22 16:42:35): [be[idm.domain.local]] [sdap_call_op_callback] (0x20000): [RID#8] Handling LDAP operation [22][server: [xxx:389] IPA EXOP] took [0.752] milliseconds.
(2024-03-22 16:42:35): [be[idm.domain.local]] [ipa_s2n_exop_done] (0x0400): [RID#8] ldap_extended_operation result: No such object(32), (null).


but it finds nothing it seems.

on the idm host everything seems to be running:

# ipactl status
Directory Service: RUNNING
krb5kdc Service: RUNNING
kadmin Service: RUNNING
named Service: RUNNING
httpd Service: RUNNING
ipa-custodia Service: RUNNING
smb Service: RUNNING
winbind Service: RUNNING
ipa-otpd Service: RUNNING
ipa-dnskeysyncd Service: RUNNING
ipa: INFO: The ipactl command was successful


And now I don't know what else to try. It seems that the trust agent is not answering properly, but how do I check this?




On Fri, Mar 22, 2024 at 4:12 PM Alexander Bokovoy <abokovoy@redhat.com> wrote:
On Пят, 22 сак 2024, Natxo Asenjo via FreeIPA-users wrote:
>hi,
>
>at work we are having *interesting* problems with our migration from idm
>servers running rhel 7 to new rhel 8 idm servers.
>
>We have a AD -> idm trust in place, this is working properly.
>
>AD is domain.local
>IDM is idm.domain.local
>
>We add replicas to the set, and this runs properly. No replication errors.
>
>Basically the problem is we cannot log in to newly joined systems running
>rhel 8 as trusted users from AD.
>
>We have a new rhel 8 idm server which has also the trust agent and trust
>controller role installed.
>
>We want to login as a trusted AD user to a rhel 8 host which has this new
>rhel 8 idm server as its ipa host, we have forced it using this sssd.conf:
>
>[domain/idm.domain.local]
>
>id_provider = ipa
>ipa_server = ds.idm.domain.local
>ipa_domain = idm.domain.local
>ipa_hostname = host.idm.domain.local
>auth_provider = ipa
>chpass_provider = ipa
>access_provider = ipa
>cache_credentials = True
>ldap_tls_cacert = /etc/ipa/ca.crt
>dyndns_update = True
>dyndns_iface = ens192
>krb5_store_password_if_offline = True
>[sssd]
>services = nss, pam, ssh, sudo
>
>domains = idm.domain.local
>full_name_format = %1$s
>debug = 9
>[nss]
>homedir_substring = /home
>override_homedir = /home/%u
>
>[pam]
>
>[sudo]
>
>[autofs]
>
>[ssh]
>
>[pac]
>
>[ifp]
>
>[session_recording]
>
>
>We also modified krb5.conf on the client to find the IDM realm only on the
>rhel 8 idm server, not the rhel 7. So we disabled srv dns autodiscovery for
>the IDM.DOMAIN.LOCAL realm:
>
># cat /etc/krb5.conf
>#File modified by ipa-client-install
>
>includedir /etc/krb5.conf.d/
>[libdefaults]
>  default_realm = IDM.DOMAIN.LOCAL
>dns_lookup_realm = true
>  rdns = false
>  dns_canonicalize_hostname = false
>dns_lookup_kdc = true
>  ticket_lifetime = 24h
>  forwardable = true
>  udp_preference_limit = 0
>  default_ccache_name = KEYRING:persistent:%{uid}
>
>
>[realms]
>  IDM.DOMAIN.LOCAL = {
>    kdc = ds.idm.domain.local:88
>    master_kdc = ds.idm.domain.local:88
>    admin_server = ds.idm.domain.local:749
>    kpasswd_server = ds.idm.domain.local:464
>    default_domain = idm.domain.local
>    pkinit_anchors = FILE:/var/lib/ipa-client/pki/kdc-ca-bundle.pem
>    pkinit_pool = FILE:/var/lib/ipa-client/pki/ca-bundle.pem
>
>  }
>
>
>[domain_realm]
>  .idm.domain.local = IDM.DOMAIN.LOCAL
>  idm.domain.local = IDM.DOMAIN.LOCAL
>  host.idm.domain.local = IDM.DOMAIN.LOCAL
>
>On the rhel 8 client, I can kinit with the host keytab, this work, I get a
>ticket with the host principal.
>
>I can also kinit using a trusted AD user, this works, I get a ticket of the
>AD domain.
>
>But as soon as I try logging in from ssh, it does not work. It does not
>recognize the user.
>
>Running id trusteduser@ad does not wok either (no such user)
>
>I have run out of ideas, to be honest. I do not know how to troubleshoot
>this anymore. The rhel8 idm server finds the users, I can log in there
>without any problem too thanks to the rbac rules, but this rhel8 client
>simpley will not see the users.

This all sounds like you see SPAKE vs non-SPAKE behavior by the clients.
https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/8/html/migrating_to_identity_management_on_rhel_8/migrate-7-to-8_migrating
covers this in a yellow warning box:

-------------------------------------
Important

RHEL 8 supports SPAKE and IdP pre-authentication, but RHEL 7 does not.
Having RHEL 8 servers with SPAKE or IdP enabled in a RHEL 7 IdM
deployment may lead to problems such as users not being able to log in.

Red Hat strongly recommends that all servers in an IdM deployment are
migrated as quickly as possible and that older systems should not be
left in operation for extended periods of time.

For more information, see:

     https://access.redhat.com/solutions/7053377
     https://access.redhat.com/solutions/3529911
-------------------------------------

Specifically, https://access.redhat.com/solutions/7053377 should have
enough technical details to confirm it is the case.


--
/ Alexander Bokovoy
Sr. Principal Software Engineer
Security / Identity Management Engineering
Red Hat Limited, Finland



--
--
Groeten,
natxo