You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@avalon.apache.org by "Alexander \"Stiefel\" Müller" <st...@stiefelnet.de> on 2002/01/24 21:25:37 UTC
Bug in Excalibur Datasource package
hi all,
i'm using the excalibur datasource package to access a IBM DB2 database,
which does not support the default <keep-alive> SELECT 1
i did not find any similar in db2 (and for performace reasons) i removed the <keep-alive> config.
now i got errors during startup, which are "bugs" in some of the files.
somebody may commit this to cvs. i used a cvs-checkout a few days ago.
i subscribed to this list today, so i do not know if this is a "known bug" (or a feature??)
file org.apache.avalon.excalibur.datasource.JdbcDataSource
method configure( final Configuration configuration ) (from line 146):
if (oradb)
{
keepAlive = "SELECT 1 FROM DUAL";
if (getLogger().isWarnEnabled())
{
getLogger().warn("The oradb attribute is deprecated, please use the" +
"keep-alive element instead.");
}
}
else
{
// XXX: let the user set this in config !!
//keepAlive = "SELECT 1"; <<<<<<<<<<<< always sets default
}
now keep-alive may be null, but connectionfactory throws exceptions:
file org.apache.avalon.excalibur.datasource.JdbcConnectionFactory
method public JdbcConnectionFactory( final String url,
final String username,
final String password,
final boolean autoCommit,
final String keepAlive,
final String connectionClass) (from line 73)
this.m_password = password;
this.m_autoCommit = autoCommit;
// otherwise new Instance() does not work (NullPointerException during Reflection)
if (keepAlive == null) { <<<<<<<< but i think there is a better way?!
this.m_keepAlive = "";
} else {
this.m_keepAlive = keepAlive;
}
after all a small bug in abstract-connection:
file org.apache.avalon.excalibur.datasource.AbstractJdbcConnection
method public AbstractJdbcConnection( final Connection connection, final String keepAlive ) (from line 63):
// subclasses can override initialize()
this.initialize();
// not &&
if (null == keepAlive || "".equals(keepAlive.trim())) <<<<< wrong bool
{
m_testStatement = null;
m_testException = null;
}
else
{
ok, thanx for now. btw: great work with avalon!
Mit freundlichen Grüßen / Best Regards,
Alexander "Stiefel" Müller
Geschäftsführer
stiefelnet e-Solver GmbH
Friedensplatz 4; D-01309 Dresden
www.stiefelnet.de
fon +49 (0)3 51 - 31 25 41 0
fax +49 (0)3 51 - 31 25 41 7
mobil +49 (0)1 71 - 48 10 96 9
e-mail stiefel@stiefelnet.de
RE: Bug in Excalibur Datasource package
Posted by "Alexander \"Stiefel\" Muller" <st...@stiefelnet.de>.
hello,
works good for me, thanx. i will provide a keep-alive-statement as soon as
i find one for DB2...
Mit freundlichen Gru?en / Best Regards,
Alexander "Stiefel" Muller
Geschaftsfuhrer
stiefelnet e-Solver GmbH
Friedensplatz 4; D-01309 Dresden
www.stiefelnet.de
fon +49 (0)3 51 - 31 25 41 0
fax +49 (0)3 51 - 31 25 41 7
mobil +49 (0)1 71 - 48 10 96 9
e-mail stiefel@stiefelnet.de
> -----Original Message-----
> From: Leif Mortenson [mailto:leif@silveregg.co.jp]
> Sent: Saturday, January 26, 2002 6:01 PM
> To: Avalon Developers List
> Subject: Re: Bug in Excalibur Datasource package
>
>
> Ok, just got this fix checked in. The syntax is as follows:
>
> Default keep-alive of "SELECT 1":
> <pool-controller {attributes}/>
>
> Custom keep-alive query:
> <pool-controller {attributes}>
> <keep-alive>SELECT 2</keep-alive>
> </pool-controller>
>
> Disable keep-alive:
> <pool-controller {attributes}>
> <keep-alive disable="true"/>
> </pool-controller>
> or
> <pool-controller {attributes}>
> <keep-alive disable="true">SELECT 2</keep-alive>
> </pool-controller>
>
> It works for me, but give this a try and make sure that it does for you
> as well.
>
> Cheers,
> Leif
>
> Peter Royal wrote:
>
> > On Friday 25 January 2002 10:02 am, Berin Loritsch wrote:
> >
> >>This is for backwards compatibility. Let's add an attribute to the
> >>keep-alive element to disable it, like this:
> >>
> >><keep-alive disabled=""/>
> >>
> >>The existence of that attribute--wether filled or not will mark it for
> >>being disabled.
> >>
> >
> > That's confusing. How about:
> >
> > <keep-alive disabled="true"/>
> >
> > You can then do a simple
> > configuration.getChild("keep-alive").getAttributeAsBoolean("disabled",
false);
> >
> > to have it default to on.
> > -pete
> >
> >
>
>
>
> --
> To unsubscribe, e-mail:
<ma...@jakarta.apache.org>
> For additional commands, e-mail:
<ma...@jakarta.apache.org>
Re: Bug in Excalibur Datasource package
Posted by Leif Mortenson <le...@silveregg.co.jp>.
Ok, just got this fix checked in. The syntax is as follows:
Default keep-alive of "SELECT 1":
<pool-controller {attributes}/>
Custom keep-alive query:
<pool-controller {attributes}>
<keep-alive>SELECT 2</keep-alive>
</pool-controller>
Disable keep-alive:
<pool-controller {attributes}>
<keep-alive disable="true"/>
</pool-controller>
or
<pool-controller {attributes}>
<keep-alive disable="true">SELECT 2</keep-alive>
</pool-controller>
It works for me, but give this a try and make sure that it does for you
as well.
Cheers,
Leif
Peter Royal wrote:
> On Friday 25 January 2002 10:02 am, Berin Loritsch wrote:
>
>>This is for backwards compatibility. Let's add an attribute to the
>>keep-alive element to disable it, like this:
>>
>><keep-alive disabled=""/>
>>
>>The existence of that attribute--wether filled or not will mark it for
>>being disabled.
>>
>
> That's confusing. How about:
>
> <keep-alive disabled="true"/>
>
> You can then do a simple
> configuration.getChild("keep-alive").getAttributeAsBoolean("disabled", false);
>
> to have it default to on.
> -pete
>
>
--
To unsubscribe, e-mail: <ma...@jakarta.apache.org>
For additional commands, e-mail: <ma...@jakarta.apache.org>
Re: Bug in Excalibur Datasource package
Posted by Peter Royal <pr...@managingpartners.com>.
On Friday 25 January 2002 10:02 am, Berin Loritsch wrote:
> This is for backwards compatibility. Let's add an attribute to the
> keep-alive element to disable it, like this:
>
> <keep-alive disabled=""/>
>
> The existence of that attribute--wether filled or not will mark it for
> being disabled.
That's confusing. How about:
<keep-alive disabled="true"/>
You can then do a simple
configuration.getChild("keep-alive").getAttributeAsBoolean("disabled", false);
to have it default to on.
-pete
--
peter royal -> proyal@managingpartners.com
--
To unsubscribe, e-mail: <ma...@jakarta.apache.org>
For additional commands, e-mail: <ma...@jakarta.apache.org>
Re: Bug in Excalibur Datasource package
Posted by Berin Loritsch <bl...@apache.org>.
Leif Mortenson wrote:
> I was looking into fixing this and have run into a problem. How do we
> want to tell the JdbcDataSouce not to perform keep-alive pings?
>
> Currently, we are defaulting to "SELECT 1" if a <keep-alive> node is not
> specified. The problem is how to specify that we want to disable it.
This is for backwards compatibility. Let's add an attribute to the keep-alive
element to disable it, like this:
<keep-alive disabled=""/>
The existence of that attribute--wether filled or not will mark it for being
disabled.
> The following will all return null for: controller.getChild(
> "keep-alive" ).getValue( null );
> <pool-controller min="5" max="10"
> connection-class="my.overrided.ConnectionClass">
> <keep-alive></keep-alive>
> </pool-controller>
>
> <pool-controller min="5" max="10"
> connection-class="my.overrided.ConnectionClass">
> <keep-alive/>
> </pool-controller>
>
> <pool-controller min="5" max="10"
> connection-class="my.overrided.ConnectionClass"/>
>
> I was thinking of modifying the code so that it looked like this:
> String keepAlive;
> Configuration keepAliveConf = controller.getChild( "keep-alive", false );
> if ( null != keepAliveConf )
> {
> keepAlive = keepAliveConf.getValue( null );
> }
> else
> {
> keepAlive = "SELECT 1";
> }
Did you check to see if that method throws an exception? If it does,
we need to rewrite the test like this:
String keepAlive = null;
try
{
keepAlive = controller.getChild( "keep-alive", false ).getValue( null );
}
catch( Exception e )
{
// element does not exist
keepAlive = "SELECT 1";
}
> Another option would be to add an attribute to the pool-controller to
> disable pinging
> <pool-controller min="5" max="10"
> connection-class="my.overrided.ConnectionClass" keep-alive="no"/>
We want to move all keep-alive pinging to the keep-alive element.
> Any opinions. I think that this last option of adding an attribute
> would be the least likely to cause problems.
>
> Once this is resolved, I can move on to fix the NPE that Alexander was
> encountering.
I like the idea of not placing a string in the keep-alive element, although
the more explicit "disabled" attribute might be more clear to users.
--
"They that give up essential liberty to obtain a little temporary safety
deserve neither liberty nor safety."
- Benjamin Franklin
--
To unsubscribe, e-mail: <ma...@jakarta.apache.org>
For additional commands, e-mail: <ma...@jakarta.apache.org>
RE: Bug in Excalibur Datasource package
Posted by "Alexander \"Stiefel\" Müller" <st...@stiefelnet.de>.
hey,
i implemented this as follows (i needed fast solution), may be
you can use something:
the xml looks like this:
<components>
<jdbc1 logger="system.database">
<pool-controller min="3" max="20"
connection-class="org.apache.avalon.excalibur.datasource.JdbcConnection">
<!--
attributes: perform=true/false: perform keep-alive select specified as value
frequence=time-value: if perform=true, only do if older than xx millis,
valid time-values: Xs, Xm, Xh, Xd or X (millis)
max-age=time-value: check for max-age, values: see frequence, set -1 to disable age-checking
-->
<keep-alive perform="false" frequence="10s" max-age="-1">SELECT bla FROM blub</keep-alive>
</pool-controller>
<driver>COM.ibm.db2.jdbc.net.DB2Driver</driver>
...
now changes to the files:
- everywhere: removed already deprecated constructors
- AbstractJdbcConnection:
new members:
protected long m_maxAge;
protected long m_pingFrq;
extend constructor:
public AbstractJdbcConnection( final Connection connection, final String keepAlive,
final long maxAge, final long pingFrq )
...
m_maxAge = maxAge;
m_pingFrq = pingFrq;
...
extend isClosed():
...
long age = System.currentTimeMillis() - m_lastUsed;
if ( m_maxAge > 0 ) // perform max-age-check?
{
if ( age > m_maxAge ) // over max-age?
...
if (m_testStatement != null && age > m_pingFrq) // over xx seconds ago
...
- JdbcConnection and JdbcConnection:
extend constructor:
public Jdbc3Connection( final Connection connection, final String keepAlive, final long maxAge, final long pingFrq )
{
super( connection, keepAlive, maxAge, pingFrq );
}
- JdbcConnectionFactory:
new members:
private final long m_maxAge;
private final long m_pingFrq;
extend constructor:
public JdbcConnectionFactory( final String url,
final String username,
final String password,
final boolean autoCommit,
final String keepAlive,
final String connectionClass,
final long maxAge,
final long pingFrq )
...
this.m_autoCommit = autoCommit;
// this resolves NullPointerException during Reflection
if ( keepAlive == null )
this.m_keepAlive = "";
else
this.m_keepAlive = keepAlive;
this.m_maxAge = maxAge;
this.m_pingFrq = pingFrq;
Class clazz = null;
...
if ( null != this.m_class )
{
try
{
Class[] paramTypes = new Class[] { Connection.class, String.class, long.class, long.class };
Object[] params = new Object[] { connection, this.m_keepAlive, new Long( this.m_maxAge ), new Long( this.m_pingFrq ) };
Constructor constructor = m_class.getConstructor( paramTypes );
jdbcConnection = (AbstractJdbcConnection) constructor.newInstance( params );
}
catch ( Exception e )
{
/* disable "old" constructors
try
{
Class[] paramTypes = new Class[] { Connection.class, boolean.class };
Object[] params = new Object[] { connection, new Boolean( this.m_keepAlive.equalsIgnoreCase(JdbcConnectionFactory.ORACLE_KEEPALIVE) ) };
Constructor constructor = m_class.getConstructor( paramTypes );
jdbcConnection = (AbstractJdbcConnection) constructor.newInstance( params );
}
catch ( Exception ie )
{
if ( getLogger().isDebugEnabled() )
{
getLogger().debug("Exception in JdbcConnectionFactory.newInstance:", ie);
}*/
throw new NoValidConnectionException(e.getMessage());
//}
...
- JdbcDataSource:
extend configure():
...
final Configuration controller = configuration.getChild( "pool-controller" );
String keepAlive = controller.getChild( "keep-alive" ).getValue(null);
final boolean perform = controller.getChild( "keep-alive" ).getAttributeAsBoolean( "perform", true);
if ( !perform )
keepAlive = null;
String smaxage = controller.getChild( "keep-alive" ).getAttribute( "max-age", "1h");
final long maxage = timeValue( smaxage );
String spingfrq = controller.getChild( "keep-alive" ).getAttribute( "frequence", "5s");
final long pingfrq = timeValue( spingfrq );
final int min = controller.getAttributeAsInteger( "min", 1 );
...
if (oradb)
{
keepAlive = "SELECT 1 FROM DUAL";
if (getLogger().isWarnEnabled())
{
getLogger().warn("The oradb attribute is deprecated, please use the" +
"keep-alive element instead.");
}
}
else
{
//let the user config this
//keepAlive = "SELECT 1";
}
final JdbcConnectionFactory factory =
new JdbcConnectionFactory( dburl, user, passwd, autoCommit,
keepAlive, connectionClass,
maxage, pingfrq );
a new method: (is there a better/existing way??):
private long timeValue( String value )
throws ConfigurationException
{
if ( value == null || value.length() == 0) // not present -> default -1
return -1;
try {
char measure = value.charAt(value.length()-1);
long l1 = -1;
switch ( measure ) {
case 's': // seconds
l1 = Long.parseLong(value.substring(0, value.length()-1)) * 1000;
break;
case 'm': // minutes
l1 = Long.parseLong(value.substring(0, value.length()-1)) * 1000 * 60;
break;
case 'h': // hours
l1 = Long.parseLong(value.substring(0, value.length()-1)) * 1000 * 60 * 60;
break;
case 'd': // days
l1 = Long.parseLong(value.substring(0, value.length()-1)) * 1000 * 60 * 60 * 24;
break;
default: // millis (no measure) or bullshit (handled by catch)
l1 = Long.parseLong(value);
break;
}
return l1;
}
catch ( Exception e ) {
throw new ConfigurationException("Unable to extract time-value: " + e.getMessage(), e);
}
o.k. that's it
Mit freundlichen Grüßen / Best Regards,
Alexander "Stiefel" Müller
Geschäftsführer
stiefelnet e-Solver GmbH
Friedensplatz 4; D-01309 Dresden
www.stiefelnet.de
fon +49 (0)3 51 - 31 25 41 0
fax +49 (0)3 51 - 31 25 41 7
mobil +49 (0)1 71 - 48 10 96 9
e-mail stiefel@stiefelnet.de
> -----Original Message-----
> From: Leif Mortenson [mailto:leif@silveregg.co.jp]
> Sent: Friday, January 25, 2002 3:06 AM
> To: Avalon Developers List
> Subject: Re: Bug in Excalibur Datasource package
>
>
> I was looking into fixing this and have run into a problem. How do we
> want to tell the JdbcDataSouce not to perform keep-alive pings?
>
> Currently, we are defaulting to "SELECT 1" if a <keep-alive> node is not
> specified. The problem is how to specify that we want to disable it.
>
> The following will all return null for: controller.getChild(
> "keep-alive" ).getValue( null );
> <pool-controller min="5" max="10"
> connection-class="my.overrided.ConnectionClass">
> <keep-alive></keep-alive>
> </pool-controller>
>
> <pool-controller min="5" max="10"
> connection-class="my.overrided.ConnectionClass">
> <keep-alive/>
> </pool-controller>
>
> <pool-controller min="5" max="10"
> connection-class="my.overrided.ConnectionClass"/>
>
> I was thinking of modifying the code so that it looked like this:
> String keepAlive;
> Configuration keepAliveConf = controller.getChild( "keep-alive", false );
> if ( null != keepAliveConf )
> {
> keepAlive = keepAliveConf.getValue( null );
> }
> else
> {
> keepAlive = "SELECT 1";
> }
>
> This way, you would get the following results:
> <pool-controller min="5" max="10"
> connection-class="my.overrided.ConnectionClass">
> <keep-alive>CUSTOM QUERY</keep-alive>
> </pool-controller>
> => keepAlive = "CUSTOM QUERY"
>
> <pool-controller min="5" max="10"
> connection-class="my.overrided.ConnectionClass">
> <keep-alive></keep-alive>
> </pool-controller>
> => keepAlive = null
>
> <pool-controller min="5" max="10"
> connection-class="my.overrided.ConnectionClass">
> <keep-alive/>
> </pool-controller>
> => keepAlive = null
>
> <pool-controller min="5" max="10"
> connection-class="my.overrided.ConnectionClass"/>
> => keepAlive = "SELECT 1"
>
> Another option would be to add an attribute to the pool-controller to
> disable pinging
> <pool-controller min="5" max="10"
> connection-class="my.overrided.ConnectionClass" keep-alive="no"/>
>
> Any opinions. I think that this last option of adding an attribute
> would be the least likely to cause problems.
>
> Once this is resolved, I can move on to fix the NPE that Alexander was
> encountering.
>
> Thanks
> Leif
>
> Alexander \"Stiefel\" Müller wrote:
>
> >hello,
> >
> >my update was 2 days ago, for i was setting up eclipse,
> >i did a update right now, and the bug in JdbcDataSource is still there:
> >
> > if (oradb)
> > {
> > keepAlive = "SELECT 1 FROM DUAL";
> >
> > if (getLogger().isWarnEnabled())
> > {
> > getLogger().warn("The oradb attribute is
> deprecated, please use the" +
> > "keep-alive element instead.");
> > }
> > }
> > else
> > {
> > keepAlive = "SELECT 1"; <<<<<<<<<<< you always set this
> > }
> >
> >i see there's a new version of AbstractJdbcConnection, but
> JdbcConnectionFactory still
> >does not allow m_keepAlive to be null:
> >
> > if ( null != this.m_class )
> > {
> > try
> > {
> > Class[] paramTypes = new Class[] { Connection.class,
> String.class };
> > Object[] params = new Object[] { connection,
> this.m_keepAlive };
> >
> > Constructor constructor = m_class.getConstructor(
> paramTypes );
> > jdbcConnection = (AbstractJdbcConnection)
> constructor.newInstance( params ); <<< fails with NullPointerException
> > }
> > catch ( Exception e )
> > {
> > try
> > {
> > Class[] paramTypes = new Class[] {
> Connection.class, boolean.class };
> > Object[] params = new Object[] { connection, new
> Boolean(
> this.m_keepAlive.equalsIgnoreCase(JdbcConnectionFactory.ORACLE_KEEPALI
> VE) ) }; <<< this too
> >
> >
> >
> >Mit freundlichen Grüßen / Best Regards,
> >
> >Alexander "Stiefel" Müller
> >Geschäftsführer
> >
> >stiefelnet e-Solver GmbH
> >Friedensplatz 4; D-01309 Dresden
> >www.stiefelnet.de
> >
> >fon +49 (0)3 51 - 31 25 41 0
> >fax +49 (0)3 51 - 31 25 41 7
> >mobil +49 (0)1 71 - 48 10 96 9
> >e-mail stiefel@stiefelnet.de
> >
> >>-----Original Message-----
> >>From: Berin Loritsch [mailto:bloritsch@apache.org]
> >>Sent: Thursday, January 24, 2002 9:35 PM
> >>To: Avalon Developers List
> >>Subject: Re: Bug in Excalibur Datasource package
> >>
> >>
> >>Alexander \"Stiefel\" Müller wrote:
> >>
> >>>hi all,
> >>>
> >>>i'm using the excalibur datasource package to access a IBM DB2 database,
> >>>which does not support the default <keep-alive> SELECT 1
> >>>i did not find any similar in db2 (and for performace reasons) i
> >>>
> >>removed the <keep-alive> config.
> >>
> >>>now i got errors during startup, which are "bugs" in some of the files.
> >>>somebody may commit this to cvs. i used a cvs-checkout a few days ago.
> >>>i subscribed to this list today, so i do not know if this is a
> >>>
> >>"known bug" (or a feature??)
> >>
> >>
> >>When was the last time you got a snapshot from Avalon Excalibur?
> >>
> >>The current CVS addressed alot of issues including the one you have here.
> >>Let me know if it is still broken.
> >>
> >>
> >>BTW, we are currently renewing effort on releasing Excalibur 4.1 RSN.
> >>
> >>
> >>
> >>--
> >>
> >>"They that give up essential liberty to obtain a little temporary safety
> >> deserve neither liberty nor safety."
> >> - Benjamin Franklin
> >>
> >>
> >>--
> >>To unsubscribe, e-mail: <ma...@jakarta.apache.org>
> >>For additional commands, e-mail: <mailto:avalon-dev-help@jakarta.apache.or
> >>
> >g>
> >
>
>
>
> --
> To unsubscribe, e-mail: <ma...@jakarta.apache.org>
> For additional commands, e-mail: <ma...@jakarta.apache.org>
Re: Bug in Excalibur Datasource package
Posted by Leif Mortenson <le...@silveregg.co.jp>.
I was looking into fixing this and have run into a problem. How do we
want to tell the JdbcDataSouce not to perform keep-alive pings?
Currently, we are defaulting to "SELECT 1" if a <keep-alive> node is not
specified. The problem is how to specify that we want to disable it.
The following will all return null for: controller.getChild(
"keep-alive" ).getValue( null );
<pool-controller min="5" max="10"
connection-class="my.overrided.ConnectionClass">
<keep-alive></keep-alive>
</pool-controller>
<pool-controller min="5" max="10"
connection-class="my.overrided.ConnectionClass">
<keep-alive/>
</pool-controller>
<pool-controller min="5" max="10"
connection-class="my.overrided.ConnectionClass"/>
I was thinking of modifying the code so that it looked like this:
String keepAlive;
Configuration keepAliveConf = controller.getChild( "keep-alive", false );
if ( null != keepAliveConf )
{
keepAlive = keepAliveConf.getValue( null );
}
else
{
keepAlive = "SELECT 1";
}
This way, you would get the following results:
<pool-controller min="5" max="10"
connection-class="my.overrided.ConnectionClass">
<keep-alive>CUSTOM QUERY</keep-alive>
</pool-controller>
=> keepAlive = "CUSTOM QUERY"
<pool-controller min="5" max="10"
connection-class="my.overrided.ConnectionClass">
<keep-alive></keep-alive>
</pool-controller>
=> keepAlive = null
<pool-controller min="5" max="10"
connection-class="my.overrided.ConnectionClass">
<keep-alive/>
</pool-controller>
=> keepAlive = null
<pool-controller min="5" max="10"
connection-class="my.overrided.ConnectionClass"/>
=> keepAlive = "SELECT 1"
Another option would be to add an attribute to the pool-controller to
disable pinging
<pool-controller min="5" max="10"
connection-class="my.overrided.ConnectionClass" keep-alive="no"/>
Any opinions. I think that this last option of adding an attribute
would be the least likely to cause problems.
Once this is resolved, I can move on to fix the NPE that Alexander was
encountering.
Thanks
Leif
Alexander \"Stiefel\" Müller wrote:
>hello,
>
>my update was 2 days ago, for i was setting up eclipse,
>i did a update right now, and the bug in JdbcDataSource is still there:
>
> if (oradb)
> {
> keepAlive = "SELECT 1 FROM DUAL";
>
> if (getLogger().isWarnEnabled())
> {
> getLogger().warn("The oradb attribute is deprecated, please use the" +
> "keep-alive element instead.");
> }
> }
> else
> {
> keepAlive = "SELECT 1"; <<<<<<<<<<< you always set this
> }
>
>i see there's a new version of AbstractJdbcConnection, but JdbcConnectionFactory still
>does not allow m_keepAlive to be null:
>
> if ( null != this.m_class )
> {
> try
> {
> Class[] paramTypes = new Class[] { Connection.class, String.class };
> Object[] params = new Object[] { connection, this.m_keepAlive };
>
> Constructor constructor = m_class.getConstructor( paramTypes );
> jdbcConnection = (AbstractJdbcConnection) constructor.newInstance( params ); <<< fails with NullPointerException
> }
> catch ( Exception e )
> {
> try
> {
> Class[] paramTypes = new Class[] { Connection.class, boolean.class };
> Object[] params = new Object[] { connection, new Boolean( this.m_keepAlive.equalsIgnoreCase(JdbcConnectionFactory.ORACLE_KEEPALIVE) ) }; <<< this too
>
>
>
>Mit freundlichen Grüßen / Best Regards,
>
>Alexander "Stiefel" Müller
>Geschäftsführer
>
>stiefelnet e-Solver GmbH
>Friedensplatz 4; D-01309 Dresden
>www.stiefelnet.de
>
>fon +49 (0)3 51 - 31 25 41 0
>fax +49 (0)3 51 - 31 25 41 7
>mobil +49 (0)1 71 - 48 10 96 9
>e-mail stiefel@stiefelnet.de
>
>>-----Original Message-----
>>From: Berin Loritsch [mailto:bloritsch@apache.org]
>>Sent: Thursday, January 24, 2002 9:35 PM
>>To: Avalon Developers List
>>Subject: Re: Bug in Excalibur Datasource package
>>
>>
>>Alexander \"Stiefel\" Müller wrote:
>>
>>>hi all,
>>>
>>>i'm using the excalibur datasource package to access a IBM DB2 database,
>>>which does not support the default <keep-alive> SELECT 1
>>>i did not find any similar in db2 (and for performace reasons) i
>>>
>>removed the <keep-alive> config.
>>
>>>now i got errors during startup, which are "bugs" in some of the files.
>>>somebody may commit this to cvs. i used a cvs-checkout a few days ago.
>>>i subscribed to this list today, so i do not know if this is a
>>>
>>"known bug" (or a feature??)
>>
>>
>>When was the last time you got a snapshot from Avalon Excalibur?
>>
>>The current CVS addressed alot of issues including the one you have here.
>>Let me know if it is still broken.
>>
>>
>>BTW, we are currently renewing effort on releasing Excalibur 4.1 RSN.
>>
>>
>>
>>--
>>
>>"They that give up essential liberty to obtain a little temporary safety
>> deserve neither liberty nor safety."
>> - Benjamin Franklin
>>
>>
>>--
>>To unsubscribe, e-mail: <ma...@jakarta.apache.org>
>>For additional commands, e-mail: <mailto:avalon-dev-help@jakarta.apache.or
>>
>g>
>
--
To unsubscribe, e-mail: <ma...@jakarta.apache.org>
For additional commands, e-mail: <ma...@jakarta.apache.org>
RE: Bug in Excalibur Datasource package
Posted by "Alexander \"Stiefel\" Müller" <st...@stiefelnet.de>.
hello,
my update was 2 days ago, for i was setting up eclipse,
i did a update right now, and the bug in JdbcDataSource is still there:
if (oradb)
{
keepAlive = "SELECT 1 FROM DUAL";
if (getLogger().isWarnEnabled())
{
getLogger().warn("The oradb attribute is deprecated, please use the" +
"keep-alive element instead.");
}
}
else
{
keepAlive = "SELECT 1"; <<<<<<<<<<< you always set this
}
i see there's a new version of AbstractJdbcConnection, but JdbcConnectionFactory still
does not allow m_keepAlive to be null:
if ( null != this.m_class )
{
try
{
Class[] paramTypes = new Class[] { Connection.class, String.class };
Object[] params = new Object[] { connection, this.m_keepAlive };
Constructor constructor = m_class.getConstructor( paramTypes );
jdbcConnection = (AbstractJdbcConnection) constructor.newInstance( params ); <<< fails with NullPointerException
}
catch ( Exception e )
{
try
{
Class[] paramTypes = new Class[] { Connection.class, boolean.class };
Object[] params = new Object[] { connection, new Boolean( this.m_keepAlive.equalsIgnoreCase(JdbcConnectionFactory.ORACLE_KEEPALIVE) ) }; <<< this too
Mit freundlichen Grüßen / Best Regards,
Alexander "Stiefel" Müller
Geschäftsführer
stiefelnet e-Solver GmbH
Friedensplatz 4; D-01309 Dresden
www.stiefelnet.de
fon +49 (0)3 51 - 31 25 41 0
fax +49 (0)3 51 - 31 25 41 7
mobil +49 (0)1 71 - 48 10 96 9
e-mail stiefel@stiefelnet.de
> -----Original Message-----
> From: Berin Loritsch [mailto:bloritsch@apache.org]
> Sent: Thursday, January 24, 2002 9:35 PM
> To: Avalon Developers List
> Subject: Re: Bug in Excalibur Datasource package
>
>
> Alexander \"Stiefel\" Müller wrote:
>
> > hi all,
> >
> > i'm using the excalibur datasource package to access a IBM DB2 database,
> > which does not support the default <keep-alive> SELECT 1
> > i did not find any similar in db2 (and for performace reasons) i
> removed the <keep-alive> config.
> > now i got errors during startup, which are "bugs" in some of the files.
> > somebody may commit this to cvs. i used a cvs-checkout a few days ago.
> > i subscribed to this list today, so i do not know if this is a
> "known bug" (or a feature??)
>
>
> When was the last time you got a snapshot from Avalon Excalibur?
>
> The current CVS addressed alot of issues including the one you have here.
> Let me know if it is still broken.
>
>
> BTW, we are currently renewing effort on releasing Excalibur 4.1 RSN.
>
>
>
> --
>
> "They that give up essential liberty to obtain a little temporary safety
> deserve neither liberty nor safety."
> - Benjamin Franklin
>
>
> --
> To unsubscribe, e-mail: <ma...@jakarta.apache.org>
> For additional commands, e-mail: <ma...@jakarta.apache.org>
Re: Bug in Excalibur Datasource package
Posted by Berin Loritsch <bl...@apache.org>.
Alexander \"Stiefel\" Müller wrote:
> hi all,
>
> i'm using the excalibur datasource package to access a IBM DB2 database,
> which does not support the default <keep-alive> SELECT 1
> i did not find any similar in db2 (and for performace reasons) i removed the <keep-alive> config.
> now i got errors during startup, which are "bugs" in some of the files.
> somebody may commit this to cvs. i used a cvs-checkout a few days ago.
> i subscribed to this list today, so i do not know if this is a "known bug" (or a feature??)
When was the last time you got a snapshot from Avalon Excalibur?
The current CVS addressed alot of issues including the one you have here.
Let me know if it is still broken.
BTW, we are currently renewing effort on releasing Excalibur 4.1 RSN.
--
"They that give up essential liberty to obtain a little temporary safety
deserve neither liberty nor safety."
- Benjamin Franklin
--
To unsubscribe, e-mail: <ma...@jakarta.apache.org>
For additional commands, e-mail: <ma...@jakarta.apache.org>