Did you also try cloning the cluster or using other cluster for the testing? The metastore down is normally a Hive Metastore issue, should not be impacting here, but you could check for more details on the error on the log4j under Driver logs.