I used the *iperf* program to check for any issues in data traffic between
the nodes, but no abnormalities were detected.

I also adjusted the Linux *limits* parameters according to the tuning and
performance recommendations from the 389 documentation.

Additionally, I upgraded the *389 Directory Server* from version *2.4.5* to
*2.5.2*.

However, despite these changes, the same errors persist when starting the
replication process.

On Mon, Dec 9, 2024 at 3:47 PM Pierre Rogier <[email protected]> wrote:

> so node2 returns a LDAP_BUSY error which is relatively rare (and usually
> happens just after logging the
> "Retry count exceeded"  error ...) because txn get aborted too often
> because of db locks conflicts
> You can try to tweak the nsslapd-db-deadlock-policy but it is a bit
> puzzling:
>     During a bulk import (i.e importing entries from another supplier),
> only the import is active on the backend, so
>    there should not be db lock  conflict during an import
> And I do not understand why deleting and recreating the agreements could
> solve such issues.
> Especially since agreement toward and from the target replica are disabled
> while the replication is in progress.
>  The fact that they exist or not, should not change anything ...
>  Unless there is a bug somewhere an a db lock is leaking. But that does
> not ring any bells in mind mind ...
>
>
> On Mon, Dec 9, 2024 at 6:37 PM Luiz Quirino via 389-users <
> [email protected]> wrote:
>
>> Hi, Pierre,
>>
>> I appreciate your assertive approach to addressing this issue.
>>
>> I reviewed the logs on node01, which was sending data for replication,
>> and noticed that at the exact moment the error was recorded on node02,
>> node01 logged an LDAP error 51 ("Server is busy").
>>
>> Given that node01 and node02 are virtual machines on the same network
>> segment, I do not believe this issue is related to the network interface.
>>
>> This leads me to focus on the minimum resource requirements or internal
>> parameters of the 389 Directory Server as potential causes.
>>
>> Currently, both nodes are configured with 4 vCPUs and 4 GB of RAM each.
>>
>> I suspect that some internal parameter in the 389 DS configuration might
>> be contributing to this issue.
>>
>> Looking forward to your insights.
>> --
>> _______________________________________________
>> 389-users mailing list -- [email protected]
>> To unsubscribe send an email to [email protected]
>> Fedora Code of Conduct:
>> https://docs.fedoraproject.org/en-US/project/code-of-conduct/
>> List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
>> List Archives:
>> https://lists.fedoraproject.org/archives/list/[email protected]
>> Do not reply to spam, report it:
>> https://pagure.io/fedora-infrastructure/new_issue
>>
>
>
> --
> --
>
> 389 Directory Server Development Team
>
-- 
_______________________________________________
389-users mailing list -- [email protected]
To unsubscribe send an email to [email protected]
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/[email protected]
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue

Reply via email to