MXS-1980 Extend test program

Now the test program will

1) Write to each node in a Galera cluster and verify that the data
   ends up in the slave.
2) At the end of 1) execute STOP SLAVE and START SLAVE to check that
   replication can be stopped and started again (won't work unless
   each node has the same server_id and value for @@log_bin_basename).
3) Block the node BLR is replicating from and expect it to connect
   to the next configured master and that replication continues to
   work. Do that for all nodes.
4) Stop MaxScale and restart it and expect 3) to work. That checks
   that BLR saves all necessary information in master.ini and is
   capable of reading it.

Documentation/Routers/Binlogrouter.md

@@ -381,7 +381,9 @@ From MaxScale 2.3 onwards it is possible to specify secondary masters that
 the binlog router can use in case the connection to the default master fails.
 
 **Note:** This is _only_ supported for gtid based replication in conjunction
-with a Galera cluster.
+with a Galera cluster and provided the following holds:
+* `@@log_slave_updates` is enabled on all servers, and
+* all nodes in the Galera cluster have the *same* `server_id`.
 
 The initial setup is performed exactly like when there is but one default master.
 ```