Mailing-List: contact user-help@commons.apache.org; run by ezmlm
Precedence: bulk
Reply-To: "Commons Users List" <user@commons.apache.org>
Received-SPF: pass (nike.apache.org: local policy)
DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws;
  s=s1024; d=yahoo.com;
  h=Message-ID:X-YMail-OSG:Received:X-Mailer:Date:From:Subject:To:MIME-Version:Content-Type;
  b=mERe0PEKOCO8OYeEUkI2a9Ltl8LH2JYQF6s1kJHOfSolp/bnCmE9Fh9gNRtuxNSf9SsLbtCYn8r1lRqAWV0sA8LCieBqp1ecRPXYIVFPcLDjGov5jNVffJ2FWOabx9L7rz9Ay5bGEgwk/meGzqdRCNlik+eLtqvKCJhGbsY8wGo=;
Message-ID: <104305.65042.qm@web33205.mail.mud.yahoo.com>
Date: Sun, 2 Jan 2011 06:05:12 -0800 (PST)
From: Nishant Ranjan <nishant_kaunve_ranjan@yahoo.com>
Subject: [dbcp] is statement pool working?
To: user@commons.apache.org
MIME-Version: 1.0
Content-Type: multipart/alternative; boundary="0-258906467-1293977112=:65042"

--0-258906467-1293977112=:65042
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable

Hi,=0A=0ATo enable statement pooling using DBCP, one needs to provide a =0A=
KeyedObjectPoolFactory while creating a PoolableConnectionFactory. This its=
elf =0Ais a bit complicated and confusing. (see: =0Ahttp://mail-archives.ap=
ache.org/mod_mbox/commons-user/200503.mbox/%3C003301c52797$9794bf20$0302a8c=
0@vikram%3E).=0A A reference implementation on how to activate connection p=
ooling has been =0Aprovided. =0A=0A=0A=0A// create your connection pool=0AG=
enericObjectPool pool =3D new GenericObjectPool();=0A=0A// and set its prop=
erties=0A// ..=0A=0A// then your connection factory=0AConnectionFactory cf =
=3D new DriverManagerConnectionFactory(.....);=0A=0A// then a factory for y=
our prepared statements=0AKeyedObjectPoolFactory kpf =3D new GenericKeyedOb=
jectPoolFactory(null, 8);=0A=0A// finally, your overall connection factory=
=0APoolableConnectionFactory pcf =3D new PoolableConnectionFactory(cf, pool=
, kpf, =0Anull, false, true);=0ASo far so good.=0A=0AHowever, the problem s=
tarts to happen when one wants to use a statement out of =0Athis statement =
pool. (You would want to use one if you have created one!) The =0Aonly way =
of doing this seems to be by creating a PoolingConnection. This is =0Awhere=
 the problem actually starts.=0A=0AThere is no way of getting a PoolingConn=
ection from the classes which were used =0Ato create the statement pool inc=
luding the PoolableConnectionFactory (where an =0Ainstance of PoolingConnec=
tion is actually created everytime a new connection is =0Acreated if a stat=
ement pool factory was provided earlier - the source code =0Areveals that).=
 However one can get a Connection from the DataSource that was =0Acreated b=
y this PooableConnectionFactory and use this Connection to create a =0APool=
ingConnection.=0A=0A// suppose we already have a DataSource called dataSour=
ce=0A=0AConnection conn =3D dataSource.getConnection();=0APoolingConnection=
 poolConn  =3D new PoolingConnection (conn);=0A=0A When a connection is cre=
ated in this way and then poolConn.prepareStatement(); =0Ais called, it giv=
es the following error:=0A=0Ajava.sql.SQLException: Statement pool is null =
- closed or invalid =0APoolingConnection.=0A    at =0Aorg.apache.commons.db=
cp.PoolingConnection.prepareStatement(PoolingConnection.java:103)=0A=0A=0AH=
owever when I implement a semaphore to keep track of the statement pool (an=
 =0Ainstance of KeyedObjectPool) during creating of a Connection and then p=
rovide =0Athis statement while creating PoolingConnection, there are no err=
ors and =0Aeverything works fine. This also makes sense when I look at the =
source code. =0A=0A=0AThis makes me think that there has to be some way of =
retrieving the statement =0Apool that is created by the PoolableConnectionF=
actory whenever a new Connection =0Ais created. Am I coming to the right co=
nclusions? =0A=0A=0AIf so, then I wonder if in the current implementation o=
f DBCP, is statement =0Apooling at all working? Should someone raise an iss=
ue in JIRA? Can I suggest a =0Asolution?=0A=0AAs this is the 3rd day that I=
 am using DBCP, I might be missing something =0Acompletely. I would be happ=
y to be relieved of a blind spot.=0A=0A Thanks & Regards,=0ANishant Ranjan=
=0A=0A
--0-258906467-1293977112=:65042--