[389-users] changelog deadlock replication failures with DNA

[thierry bordaz]

On 06/14/2013 01:17 AM, Mahadevan, Venkat wrote:
>
> This means that server was unabled (because of too many retries due to 
> deadlock) to write the update in the changelog.
> This triggers the failure of the operation. If it is on the consumer, 
> that means that the supplier will retry later to send the update and 
> as you have 2 differents CSN in the log I think the updates are also 
> progressing on consumers side. Now they can be late.
> I do not know why it is occurring. Deadlock is quite rare because 
> under default deployment threads are synchronized by a backend lock.
> Is the dse.ldif available somewhere ?
>
> Hi Thierry,
>
> Thanks for the clarification. Actually, the error is only occurring on 
> the master (supplier) side which is receiving LDAP
>
> writes from an LDAP client. It does not seem to occur on any of the 
> consumers. Once this error occurs on the supplier side, it never
>
> attempts to push the change to the consumer. This causes the consumers 
> to go out of sync with the supplier and
>
> they need to be re-initialized. From my understanding, if a changelog 
> entry is not written on the supplier side even
>
> when a change occurs, the error prevents the change from being sent to 
> the consumer? That’s what seems to be happening.
>
> I will mail you a copy of the dse.ldif if that is okay. Thanks again 
> for your advice with this.
>
> cheers,
>
> VM
>
Hi Mahadevan,

If the operation fails to write into the changelog, the operation fails. 
In your case, it means that the ldapclient should receive an error.
So it is like if the operation never happened and is not replicated.

regards
thierry
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.fedoraproject.org/pipermail/389-users/attachments/20130614/9d0abd3d/attachment.html>