...

1. Master :
   [mysqld]
   datadir=/var/lib/mysql
   socket=/var/lib/mysql/mysql.sock
   # Disabling symbolic-links is recommended to prevent assorted security risks
   symbolic-links=0
   # Settings user and group are ignored when systemd is used.
   # If you need to run mysqld under a different user or group,
   # customize your systemd unit file for mysqld according to the
   # instructions in http://fedoraproject.org/wiki/Systemd
   server-id=1
   default-storage-engine=InnoDB
   character-set-server=utf8
   transaction-isolation=READ-COMMITTED
   log-bin=mysql-bin
   innodb_flush_log_at_trx_commit=1
   sync_binlog=1
   binlog-format=ROW
   #Bin logs cleanup configuration
   expiryexpire_logs_days=10
   max_binlog_size=100M
   
   [mysqld_safe]
   log-error=/var/log/mysqld.log
   pid-file=/var/run/mysqld/mysqld.pid
   
2. Slave
   [mysqld]
   datadir=/var/lib/mysql
   socket=/var/lib/mysql/mysql.sock
   # Disabling symbolic-links is recommended to prevent assorted security risks
   symbolic-links=0
   # Settings user and group are ignored when systemd is used.
   # If you need to run mysqld under a different user or group,
   # customize your systemd unit file for mysqld according to the
   # instructions in http://fedoraproject.org/wiki/Systemd
   server-id=2
   default-storage-engine = InnoDB
   character-set-server = utf8
   transaction-isolation = READ-COMMITTED
   log-bin=mysql-bin
   innodb_flush_log_at_trx_commit=1
   sync_binlog=1
   binlog-format=ROW
   #Parameters to solve split brain problem
   auto_increment_increment=10
   auto_increment_offset=2
   #Bin logs cleanup configuration
   expiryexpire_logs_days=10
   max_binlog_size=100M
   [mysqld_safe]
   log-error=/var/log/mysqld.log
   pid-file=/var/run/mysqld/mysqld.pid
   

...

• Currently Monitoring of Slave by admin is not possible
• Monitoring events are not integrated with MS

FAQs

Q: How fast the fail over will happen if master goes down?

A: It is actually the piing time out which controls this factor. The mysql connector property that controls this is "loadBalancePingTimeout".
     Currently this property is not exposed through db.properties so it uses the default pint time out value which is 0(Mean no ping time out)
     (http://dev.mysql.com/doc/connector-j/en/connector-j-reference-configuration-properties.html) 

Q:  What Mysql DB Versions are supported when DB HA is enabled?

A: The RHEL/CentOS versions supported by the Current ACS versions will get mysql versions of 5.1.X. So by defualt DB HA is also supported for these mysql versions only.

Q: Asynchronous replication V/S Semi-Synchronous Replication which one to use?

A: As Semi-Synchronous replication is supported by 5.5+ on wards so currently we support only Asynchronous replication only.

Q: How much latency is expected in syncing up when DB-HA is enabled?

A: The Synchronization between 2 master nodes is immediate and latency is depending on the actual network latency between 2 master nodes. There is no parameter to control this. The only thing we can get from master server nodes is how many seconds it is behind the other master server.
The following parameter gives us the above value when we so "Show slave status" on any master mysql prompt : "Seconds_Behind_Master"
 (http://dev.mysql.com/doc/refman/5.7/en/show-slave-status.html)

Q: What Network topologies(Same Zone, Multiple Zones) are suggested?

A: .As mentioned in previous question, as long as we make sure the network latency between 2 master nodes is as much less as possible we can use any topology. There are no recommended practices on this in mysql documentation

Appendix

• What if I want to do a Manual Switch over from Master-1 to Master-2?
  • Put a 2 way replication as above and do not enable the db ha in db.properties. I.e put the value false against the following property in db.properties
    db.ha.enabled=false

...