Mailing-List: contact commons-user-help@jakarta.apache.org; run by ezmlm
Precedence: bulk
Reply-To: "Jakarta Commons Users List" <commons-user@jakarta.apache.org>
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Subject: [DBCP] Question to database reconnection
Date: Wed, 11 Aug 2004 17:00:33 +0200
Message-ID: <2DA9A7329E188D4EAA449E2E9BAD30A51EE335@DEIMEXCH1.ndsatcom.com>
Thread-Topic: [DBCP] Question to database reconnection
Thread-Index: AcR/s/Mn1Go/bzXHSOqGwkuj05xLiA==
From: "Leber, Dieter" <DIETER.LEBER@NDSatcom.com>
To: <commons-user@jakarta.apache.org>

I'm using DBCP 1.2 without using any validation of the connections=20
(testOnBorrow =3D false, testOnReturn =3D false, =
timeBetweenEvictionRunsMillis =3D -1).=20

When shutting down the database during running the application, the=20
request using a pooled connection realizes that the database is down.=20
Returning the connection to the pool seems to clean the pool (i.e. the=20
pool has no active nor idle connections any longer).=20

Starting the database again and demanding a connection from the=20
pool a new connection is returned which has never been in the pool =
before.

If this is right no reconnection handling is needed at all because we =
always
have valid connections every time the database was down and started up =
again.

Can anyone please throw any light on this behaviour!? Are there=20
situations where the pool is not cleaned so that validation is actually=20
needed?

Dieter Leber

---------------------------------------------------------------------
To unsubscribe, e-mail: commons-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: commons-user-help@jakarta.apache.org