[jira] Commented: (DERBY-1142) Metadata calls leak memory

[ http://issues.apache.org/jira/browse/DERBY-1142?page=comments#action_12429260 ] Mayuresh Nirhali commented on DERBY-1142: - If the singleUseActivation is closed in the finalize method of EmdebResultSet then the memory leak is not

[jira] Commented: (DERBY-1142) Metadata calls leak memory

[ http://issues.apache.org/jira/browse/DERBY-1142?page=comments#action_12420461 ] Daniel John Debrunner commented on DERBY-1142: -- Two patches applied (420656 421015) that stop the memory leak shown in the repro. A leak still exists if the

[jira] Commented: (DERBY-1142) Metadata calls leak memory

[ http://issues.apache.org/jira/browse/DERBY-1142?page=comments#action_12420085 ] Daniel John Debrunner commented on DERBY-1142: -- An addition to the subset of changes of DERBY-827 fixed the issue, increasing the number of iterations from

[jira] Commented: (DERBY-1142) Metadata calls leak memory

[ http://issues.apache.org/jira/browse/DERBY-1142?page=comments#action_12419953 ] Daniel John Debrunner commented on DERBY-1142: -- The subset of changes from the draft patch in DERBY-827 (not 872 as I mis-typed previously) do not address this

[jira] Commented: (DERBY-1142) Metadata calls leak memory

[ http://issues.apache.org/jira/browse/DERBY-1142?page=comments#action_12371700 ] Knut Anders Hatlen commented on DERBY-1142: --- My proposed solution actually fixed the memory leak, but broke derbyall. These tests failed:

[jira] Commented: (DERBY-1142) Metadata calls leak memory

[ http://issues.apache.org/jira/browse/DERBY-1142?page=comments#action_12371561 ] Anders Morken commented on DERBY-1142: -- The direct cause of the memory usage is in the org.apache.derby.impl.sql.conn.GenericLanguageConnectionContext class, more

[jira] Commented: (DERBY-1142) Metadata calls leak memory

[ http://issues.apache.org/jira/browse/DERBY-1142?page=comments#action_12371579 ] Knut Anders Hatlen commented on DERBY-1142: --- Thanks for looking into this, Anders! The activation is removed from acts when Activation.close() is called. Normally,

Re: [jira] Commented: (DERBY-1142) Metadata calls leak memory

Knut Anders Hatlen (JIRA) derby-dev@db.apache.org writes: Could we solve the problem by invoking owningStmt.close() when forMetaData is true and the activation count is not greater than one? Sorry, that should have said when forMetaData is true only. The activation count is not relevant in

[jira] Commented: (DERBY-1142) Metadata calls leak memory

[ http://issues.apache.org/jira/browse/DERBY-1142?page=comments#action_12371598 ] Daniel John Debrunner commented on DERBY-1142: -- May be unrelated but part of my example patch for DERBY-872 I had a feeling was a more generic issue about