[ https://issues.jboss.org/browse/TEIID-4858?page=com.atlassian.jira.plugin... ] Steven Hawkins commented on TEIID-4858: --------------------------------------- Sorry for the delay in following up. Yes the source sql is in there - the difference it that the slow log has an "order by" clause. There does appear to be references to the order by operation not being well optimized on hive, so the simple solution would be to disable order by support on the hive translator with an override: {code} <translator name="hive-restricted" type="hive" /> <property name="supportsOrderBy" value="false"/> </translator> {code} And then reference the hive-restricted as the translator for your source. It would be good to know if there are some other considerations in play here as I wouldn't expect such a difference in performance. Do you see all sorts performing poorly, or perhaps is it related to the data type, or some other factor? -- This message was sent by Atlassian JIRA (v7.2.3#72005)