Hi,
and thanks for your fast response. Actually you answered my question already with your first reply that it's possible to configure the replication in this way. So far I've the idea this will fix my problems. Otherwise I'd like to come back, of course with the necessary information :-)
Regards, Andreas
Am 10.09.20 um 17:15 schrieb Quanah Gibson-Mount:
--On Thursday, September 10, 2020 9:43 AM +0200 Ulrich Windl Ulrich.Windl@rz.uni-regensburg.de wrote:
Obviously you should give some details: Two masters? How many slaves? What is "a lot" of changes? And how long are "some delays"?
I would agree here, there's a significant lack of useful information. What log level is set? You essentially have to run at "stats sync" to be able to understand what's happening at the replication level. There are many different things that could be occurring (for example, the servers weren't loaded initially in the correct fashion for MMR, causing writes to trigger fallback REFRESHes, etc, and generally destroying the integrity of the system).
Also, without any configuration information, it's impossible to know if your environment is even configured correctly at the ground level.
Regards, Quanah
--
Quanah Gibson-Mount Product Architect Symas Corporation Packaged, certified, and supported LDAP solutions powered by OpenLDAP: http://www.symas.com