[ https://forge.continuent.org/jira/browse/SEQUOIA-1002?page=comments#action_14189 ]
Tommi Takkunen-Lauri commented on SEQUOIA-1002: ----------------------------------------------- This property has been documented and will be published in the next release. > Add defaultFetchSize property to JDBC driver to make it easier to handle very > large result sets in the controller > ----------------------------------------------------------------------------------------------------------------- > > Key: SEQUOIA-1002 > URL: https://forge.continuent.org/jira/browse/SEQUOIA-1002 > Project: Sequoia > Type: New Feature > Versions: Sequoia 2.9 > Environment: All > Reporter: Robert Hodges > Assignee: Tommi Takkunen-Lauri > Priority: Critical > Fix For: sequoia 2.10.10 > > > Controllers run out of heap space for very large requests that are typical of > reporting or large batch operations. It is possible to alleviate this by > using set fetch size hints (Statement.setFetchSize() or > ResultSet.setFetchSize()). However, these hints may be impractical to put in > many applications after the fact or the application itself may be non-Java > and connecting through Myosotis. > To help manage fetch size hints more easily the Sequoia JDBC driver will be > extended to add a defaultFetchSize property. This property value replaces > the default 0 fetch size hint and allows users to specify connections that > use result set fetching without having to change application code. The > property behavior is as follows: > 1.) If the defaultFetchSize property is 0, it is ignored. This is the > default behavior. > 2.) If the client request has a fetch size other than 0, the request value > takes precendence and the defaultFetchSize property is ignored. > 3.) If the client request has a 0 fetch size *and* the defaultFetchSize is > non-zero, the request will be assigned the defaultFetchSize property value. > Users whose applications generate very large result sets or who have > encountered Java heap out of memory errors should define the defaultFetchSize > property. For reporting or other applications it is best to define a > separate connection, as using fetch size slows down result set processing and > generates load on databases for longer. -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: https://forge.continuent.org/jira/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira _______________________________________________ Sequoia mailing list [email protected] https://forge.continuent.org/mailman/listinfo/sequoia
