Hello all, last night our icinga2 master server went down very quietly because of a network problem, so today I decided it is really about time to have at least two masters.
I have few questions though, any comments/advices appreciated. >From the documentation: "Keep in mind that only one node acts as configuration master having the configuration files in the zones.d directory." So, the currect master has all the config in zones.d, others, including the second master, get the configuration synced, saved to /var/lib/icinga2/api/zones/. If the current master fail, the second master takes over and everything keep going, no problem. The second master is now the current master, so either I can move the configs from /var/lib/icinga2/api/zones/ to zones.d, or I can provision the configs on the master node by chef, or I can fix the failed master… Is it so? Another question is about database, this is the main confusion for me right now. There is only one database server, it's no like each master within the zone got its one database instance, right? >From the doc, I understand that only one instance at a time writes to the database. So, the database server better be located on different machine than the icinga master is running, right? Also, in our case, we will have one Icinga2 master located in CZ and the second master in US. My guess is that it would be good to also have a database server in each location and setup some kind of master-master replication maybe (I will need a lot of help from our db guy here). But I don't know how to instruct each icinga master to use the database server in its nearest geolocation. Thanks for help, guys.
_______________________________________________ icinga-users mailing list icinga-users@lists.icinga.org https://lists.icinga.org/mailman/listinfo/icinga-users
