Update the JDBC Migration Queue to allow for an "ERROR_REASON" column that would contain the stack trace of why that particular node failed. A client with a big Solr portal requested this because they had a ton of documents that failed, and it was hard to track down individually why each document failed from the logs