[ https://issues.jboss.org/browse/TEIID-2558?page=com.atlassian.jira.plugin... ] Steven Hawkins edited comment on TEIID-2558 at 7/13/13 6:31 AM: ---------------------------------------------------------------- The array method above is reasonably performant when a cardinality hint is added, but it is a somewhat limiting approach. The full array must be held in memory (which may be reasonable for even joins with millions of rows that use a single/simple join column) and anything other than a single number/string join would require the creation of an appropriate array/struct type to hold the key data. One possible approach would be to assume that appropriate global temporary tables have already been created possibly by us at deployment type - but given the possible type combinations and possible multiple appearances of dependent sets in a given source query that seems unpalatable. was (Author: shawkins): The array method above is reasonably performant when a cardinality hint is added, but it is a somewhat limiting approach. The full array must be held in memory (which may be reasonable for even joins with millions of rows that use a single/simple join column) and any single number/string join would require the creation of an appropriate array/struct type to hold the key data. One possible approach would be to assume that appropriate global temporary tables have already been created possibly by us at deployment type - but given the possible type combinations and possible multiple appearances of dependent sets in a given source query that seems unpalatable. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira