Failed Resource Actions: * ms_mysql_start_0 on centreon-slave 'error'

Hello, @thieryl, 

First, you can check that your user account used for the replication can be accessed from A to B and from B to A

Then, make sure there are no typos in the config files in /etc/centreon-ha/*
You can use the pcs config show command to check that the account configured for the ms_mysql resource is also correct

Once this is done, you'll need to run a sync again, so run :

pcs resource unmanage ms_mysql

Shutdown the MariaDB service on the server that's supposed to be the least up to date : 

mysqladmin -p shutdown  

And then run the sync script (on the most up-to-date server) :

/usr/share/centreon-ha/bin/mysql-sync-bigdb.sh

Then you can check the replication status with the following script who should return a similar result : 

/usr/share/centreon-ha/bin/mysql-check-status.sh
    Connection MASTER Status 'el8-central2-2310' [OK]
    Connection SLAVE Status 'el8-central1-2310' [OK]
    Slave Thread Status [OK]
    Position Status [OK]

Don't forget to remanage the resource : 

pcs resource manage ms_mysql

After that, if it doesn't restart and to make the cluster forget failed operations from history of the resource and re-detect its current state, execute the following commands: 

pcs resource enable ms_mysql
pcs resource clear ms_mysql
pcs resource cleanup ms_mysql

Hope it helps!