dump with stop slave can not restore

[svaroqui]

On MariaDB 2 node cluster when a slave is failed but the replication is setup
The dump is missing --master-data=1 and --gtid on restoring if start slave will auto point to first leader binlog making the restore process failed

Notice that on the user cluster the master was marked as suspect. but we did not really check the replication was still setup , i hope there is no regression in that area