Issues with MySQL slave may require a complete rebuild to get replication back in sync. That will eat up a lot of a weekend at this point. Should be able to do it with little production downtime though.