Mailing-List: contact derby-dev-help@db.apache.org; run by ezmlm
Precedence: bulk
Reply-To: <derby-dev@db.apache.org>
Received-SPF: neutral (asf.osuosl.org: local policy)
Message-ID: <43E40C30.5060109@apache.org>
Date: Fri, 03 Feb 2006 18:06:40 -0800
From: Daniel John Debrunner <djd@apache.org>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US;
 rv:1.7.3) Gecko/20040910
MIME-Version: 1.0
To: derby-dev@db.apache.org
Subject: Re: Auto-close of result sets on failure?
References: <43E3DEFB.4000606@sun.com>
In-Reply-To: <43E3DEFB.4000606@sun.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

David W. Van Couvering wrote:

> There is a new metadata call, "autoCommitFailureClosesAllResultSets",
> which returns true if a SQLException while autoCommit is true indicates
> that all open ResultSets are closed, even ones that are holdable.
> 
> In my naive scan of the code on the embedded side, it appears as though
> we do *not* close result sets upon exception.  There is a method
> clearResultSets() on EmbedStatement that is only called when the
> statement itself is closed or when executing a new statement (we close
> the old result sets from the previous invocation of the statement).
> 
> Because of the asynchronous nature of the network client protocol, it's
> a little harder to decode, but it appears that we do not close result
> sets upon exception there either.
> 
> Does anyone care to differ?  Otherwise I'll return false for this new
> JDBC4 metadata call.

I think I'll differ. :-)

In embedded I would think that all the ResultSets would be closed on an
exception that causes a rollback in auto-commit mode. You may not see
explict code, but the internal Derby objects the JDBC ResultSet depends
on will be closed. At least I think it's that way, no idea for client.
Seems like a simple test is in order.

Dan.