On 2/18/19 11:41 AM, Kees Bakker via FreeIPA-users wrote:
On 18-02-19 10:06, Florence Blanc-Renaud wrote:
> On 2/18/19 9:00 AM, Kees Bakker via FreeIPA-users wrote:
>> Hey,
>>
>> Replication isn't working, at least not automatically. If I do
>> a ipa-replica-manage re-initialize then everything is present
>> on the replica.
>>
>> I've looked through all the logs, but I couldn't find anything
>> that hints me what could be wrong.
>>
>> Today I created a new replica. The installation went OK. No error.
>> But also that replica does not receive updates.
>>
>> The IPA master (three at the moment) are running Centos7.
>>
>> [root@rotte ~]# rpm -qa 'ipa*'
>> ipa-server-4.5.4-10.el7.centos.4.4.x86_64
>> ipa-server-dns-4.5.4-10.el7.centos.4.4.noarch
>> ipa-client-common-4.5.4-10.el7.centos.4.4.noarch
>> ipa-server-common-4.5.4-10.el7.centos.4.4.noarch
>> ipa-client-4.5.4-10.el7.centos.4.4.x86_64
>> ipa-server-trust-ad-4.5.4-10.el7.centos.4.4.x86_64
>> ipa-common-4.5.4-10.el7.centos.4.4.noarch
>>
>> [root@rotte ~]# ipa-replica-manage -v list rotte.ghs.nl
>> iparep3.ghs.nl: replica
>> last init status: None
>> last init ended: 1970-01-01 00:00:00+00:00
>> last update status: Error (0) Replica acquired successfully: Incremental
update succeeded
>> last update ended: 2019-02-18 07:50:56+00:00
>> linge.ghs.nl: replica
>> last init status: None
>> last init ended: 1970-01-01 00:00:00+00:00
>> last update status: Error (0) Replica acquired successfully: Incremental
update succeeded
>> last update ended: 2019-02-18 07:50:56+00:00
>>
>> rotte is the main master (doing CA), linge and iparep3 are the replicas.
>>
>> I know that it may be hard to tell me what is wrong, without
>> further information, but I would like to know what information
>> I need to look for.
>>
>> Any help is greatly appreciated.
>>
>
> Hi,
>
> please find more info in the wiki:
https://www.freeipa.org/page/Troubleshooting/Directory_Server
>
> If you add an entry on rotte, does this entry get replicated to the other servers?
and is the reverse true? The "last update status" seems to indicate that
everything is working well.
>
Hi Flo,
Hmm, that's funny. I did not try to create a user on the other two,
because I was trying to do everything on my first master (rotte).
The funny part is, that now a new user on linge is replicated
correctly to the other two. Why haven't I tested this before?
And also a new user on iparep3 is correctly replicated to the
other two. Then I added a new user on rotte, which is now correctly
replicated. All seems to be alright. I'm puzzled.
The logs did not reveal anything suspicious, replication simply
did not work. New users were created on rotte, and also new DNS
entries were created (our DHCP server updates DNS entries). But
nothing was replicated.
Still, there is one added user (test01) on rotte which was not
replicated to linge nor to iparep3. I did a re-initialize on linge and
made user test01 become present on linge. That user is still not
present on iparep3.
Did you also run re-initialize from rotte to iparep3?
If no, the difference may be caused by replication conflicts in iparep3.
The following doc explains how to list them: [1], and how to repair.
Note that re-initialize from rotte to iparep3 would also solve the issue.
BTW. There is a problem on rotte with numSubordinates in
cn=users,cn=accounts,$SUFFIX. The number is one too high.
We have 81 users. Have a look at the output of cipa [2] (which
just looks at numSubordinates I believe).
This issue may be linked to issue 7839 [1], please have a look at this
comment specifically [2]
HTH,
flo
[1]
https://access.redhat.com/documentation/en-us/red_hat_directory_server/10...
[2]
https://pagure.io/freeipa/issue/7839
[3]
https://pagure.io/freeipa/issue/7839#comment-550188
[root@rotte ~]# cipa
+--------------------+-----------+---------+---------+-------+
| FreeIPA servers: | rotte | linge | iparep3 | STATE |
+--------------------+-----------+---------+---------+-------+
| Active Users | 82 | 81 | 80 | FAIL |
How this happened? I think this may have happened when
a user was added on two systems (rotte and linge) when
there was an old IPA master in between, but that server
was switched off. As a result there were errors on rotte
saying it could not delete a tombstone, something like this
[14/Jan/2019:16:29:01.225643460 +0100] - ERR - NSMMReplicationPlugin - _delete_tombstone
- Unable to delete tombstone
nsuniqueid=c0a66e04-125a11e9-bb6698e2-54354ddc,cn=bmot,cn=groups,cn=accounts,$SUFFIX,
uniqueid c0a66e04-125a11e9-bb6698e2-54354ddc: Operations error.
I followed this webpage [1] to delete that manually. A ldapdelete
command failed because of a linked entry. Maybe that caused
a failure to update numSubordinates.
[1]
https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/6/...
[2]
https://github.com/peterpakos/checkipaconsistency
_______________________________________________
FreeIPA-users mailing list -- freeipa-users(a)lists.fedorahosted.org
To unsubscribe send an email to freeipa-users-leave(a)lists.fedorahosted.org
Fedora Code of Conduct:
https://getfedora.org/code-of-conduct.html
List Guidelines:
https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives:
https://lists.fedorahosted.org/archives/list/freeipa-users@lists.fedoraho...