If any of the node from the cluster has been discontinued for use permanently, you should update the active appliances to stop trying to replicate to this inactive node to avoid having replication backlog. Follow below steps on currently active nodes to update status of inactive node.
shell> mys
mysql> update servers set status='Inactive', replication_status='Inactive' where sid=<SID>;
If you are unsure of the SID number assigned to inactive appliance, run below query to find out the SID number and then use that SID number to replace argument in above query.
mysql> select sid, fqdn from servers;
mysql> select * from servers\G
mysql> exit
shell> sudo service glaffishd restart