|
Sorry about that mail - it got delayed by the apache moderators because
I accidently sent it with a different email address than the one I am
subscribed with. It is identical to one you already answered, and the conclusion is that my code expected commitTransaction() to implicit call endTransaction() like in ibatis 1. IMHO this is an easy mistake to make since I associate those two calls with the SQL commands commit and rollback. Actually, altough the documentation says I always have to call endTransaction() it also says "The only way to return a connection to the DataSource (or close the connection) is to call commitTransaction() or endTransaction()" (page 39 in the iBATIS SQL Maps 2 Developer Guide - iBATIS-SqlMaps-4.pdf). This is not true, commitTransaction() does not close the resources, which was my problem. Baldur Clinton Begin wrote: Sorry, I cannot recreate this behaviour. iBATIS certainly DOES close all resources, including the session. I've verified this in response to your report. The code you posted from the 1.x to 2.x conversion is clearly incorrect, so you know what to do with that. Otherwise, I cannot recreate the problem you describe.Clinton On Tue, 28 Dec 2004 17:15:06 +0100, Baldur Norddahl <[EMAIL PROTECTED]> wrote: |
- session leak in ibatis 2.0.8 Baldur Norddahl
- Re: session leak in ibatis 2.0.8 Brandon Goodin
- Re: session leak in ibatis 2.0.8 Baldur Norddahl
- Re: session leak in ibatis 2.0.8 Clinton Begin
- Re: session leak in ibatis 2.0.8 Baldur Norddahl
- Re: session leak in ibatis 2.0.8 Clinton Begin
- Re: session leak in ibatis 2.0.8 Baldur Norddahl
- Re: session leak in ibatis 2.0.8 Clinton Begin
- Re: session leak in ibatis 2.0.8 Baldur Norddahl
- Re: session leak in ibatis 2.0.8 Clinton Begin
- Re: session leak in ibatis 2.0.8 Baldur Norddahl
- Re: session leak in ibatis 2.0.8 Clinton Begin

