It has been identified that the database responsible for the GPII wiki should be highly available during increased usage periods (like the APCP Pilot).
The MySQL server responsible for the GPII wiki database is a single-node server. It needs to be enhanced with at least a failover node.
Historically, MySQL master/master or master/slave replication has been weak and prone to failures. It was initially attempted with the GPII wiki database server but quickly came out of sync and required constant maintenance.
That being said, a simple arrangement with a load balancer (HAProxy, nginx, etc) in front of a read/write instance that forwards requests to a secondary MySQL instance to be kept as failover/standby should be enough.
This ticket is concerned with deploying a master/standby set of MySQL servers, with a load balancer in front of them working as explained above. Unless, it is determined that some other robust MySQL replication solution has become available and is worth further consideration.