Re: [JBoss-dev] Oracle BLOB handling - FIXED?
We've had nothing but headaches using the Oracle JDBC drivers. They don't
handle LOBs the way everyone else does, especially when the size is greater
than 4000. That's a specific cuttoff. Anything below that will behave the
way you'd expect. Anything over that will certainly fail unless you code
specifically to handle it (at least in our testing).
Here are a couple articles that may be useful:
http://otn.oracle.com/doc/java.815/a64685/oraext4.htm#1043351
http://technet.oracle.com/sample_code/tech/java/sqlj_jdbc/files/jdbc20/LOBSa
mple/LOBSample.java.html
Hope this is helpful.
Brian
- Original Message -
From: Dain Sundstrom [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Sent: Tuesday, June 18, 2002 11:04 AM
Subject: Re: [JBoss-dev] Oracle BLOB handling - FIXED?
This is the way the code was before I integrated the Oracle BLOB patch.
The problem is the serialized stream is being truncated to 4000 bytes.
I just got an email from someone willing to work on implementing correct
BLOB handling.
I'd still like someone to try this in Oracle 8.
-dain
Stephen Coy wrote:
I've done some more poking at this, and modified JDBCUtil.setParameter
as follows:
//
// Binary types need to be converted to a byte array and set
//
if(isBinaryJDBCType(jdbcType))
{
byte[] bytes = convertObjectToByteArray(value);
if (bytes.length 2000 jdbcType != Types.BLOB)
{
// it's more efficient to use setBinaryStream for large
// streams, and causes problems if not done on some DBMS
// implementations
ps.setBytes(index, bytes);
} else
{
InputStream in = null;
try
{
in = new ByteArrayInputStream(bytes);
ps.setBinaryStream(index, in, bytes.length);
} finally
{
safeClose(in);
}
}
return;
ie. Forget about the ByteArrayBlob class.
This works fine when writing to the database.
However, reading the data back fails:
2002-06-18 23:59:20,671 DEBUG
[org.jboss.ejb.plugins.cmp.jdbc.JDBCLoadEntityCommand.SimpleEJB]
Executing SQL: SELECT booleanPrimitive, booleanObject, bytePrimitive,
byteObject, shortPrimitive, shortObject, integerPrimitive,
integerObject, longPrimitive, longObject, floatPrimitive, floatObject,
doublePrimitive, doubleObject, stringValue, utilDateValue, sqlDateValue,
timeValue, timestampValue, bigDecimalValue, byteArrayValue, objectValue
FROM SIMPLE WHERE (id=?)
2002-06-18 23:59:20,680 TRACE
[org.jboss.ejb.plugins.cmp.jdbc.bridge.JDBCCMP2xFieldBridge.SimpleEJB.id]
Set parameter: index=1, jdbcType=VARCHAR, value=simple
2002-06-18 23:59:20,707 TRACE
[org.jboss.ejb.plugins.cmp.jdbc.bridge.JDBCCMP2xFieldBridge.SimpleEJB.boolea
nPrimitive]
Get result: index=1, javaType=boolean, Simple, value=true
2002-06-18 23:59:20,709 TRACE
[org.jboss.ejb.plugins.cmp.jdbc.bridge.JDBCCMP2xFieldBridge.SimpleEJB.boolea
nObject]
Get result: index=2, javaType=java.lang.Boolean, Simple, value=false
2002-06-18 23:59:20,713 TRACE
[org.jboss.ejb.plugins.cmp.jdbc.bridge.JDBCCMP2xFieldBridge.SimpleEJB.bytePr
imitive]
Get result: index=3, javaType=byte, Simple, value=11
2002-06-18 23:59:20,715 TRACE
[org.jboss.ejb.plugins.cmp.jdbc.bridge.JDBCCMP2xFieldBridge.SimpleEJB.byteOb
ject]
Get result: index=4, javaType=java.lang.Byte, Simple, value=22
2002-06-18 23:59:20,717 TRACE
[org.jboss.ejb.plugins.cmp.jdbc.bridge.JDBCCMP2xFieldBridge.SimpleEJB.shortP
rimitive]
Get result: index=5, javaType=short, Simple, value=33
2002-06-18 23:59:20,718 TRACE
[org.jboss.ejb.plugins.cmp.jdbc.bridge.JDBCCMP2xFieldBridge.SimpleEJB.shortO
bject]
Get result: index=6, javaType=java.lang.Short, Simple, value=44
2002-06-18 23:59:20,720 TRACE
[org.jboss.ejb.plugins.cmp.jdbc.bridge.JDBCCMP2xFieldBridge.SimpleEJB.intege
rPrimitive]
Get result: index=7, javaType=int, Simple, value=55
2002-06-18 23:59:20,726 TRACE
[org.jboss.ejb.plugins.cmp.jdbc.bridge.JDBCCMP2xFieldBridge.SimpleEJB.intege
rObject]
Get result: index=8, javaType=java.lang.Integer, Simple, value=66
2002-06-18 23:59:20,728 TRACE
[org.jboss.ejb.plugins.cmp.jdbc.bridge.JDBCCMP2xFieldBridge.SimpleEJB.longPr
imitive]
Get result: index=9, javaType=long, Simple, value=77
2002-06-18 23:59:20,730 TRACE
[org.jboss.ejb.plugins.cmp.jdbc.bridge.JDBCCMP2xFieldBridge.SimpleEJB.longOb
ject]
Get result: index=10, javaType=java.lang.Long, Simple, value=88
2002-06-18 23:59:20,731 TRACE
[org.jboss.ejb.plugins.cmp.jdbc.bridge.JDBCCMP2xFieldBridge.SimpleEJB.floatP
rimitive]
Get result: index=11, javaType=float, Simple, value=11.11
2002-06-18 23:59:20,733 TRACE
[org.jboss.ejb.plugins.cmp.jdbc.bridge.JDBCCMP2xFieldBridge.SimpleEJB.floatO
bject]
Get result: index=12,
[JBoss-dev] JBossOne
Great job on JBossOne, Mark et al. Very informative sessions.Pleasure meeting you and best of luck! Brian
Re: [JBoss-dev] RE: [JBoss-user] JBOSS 3.x FINAL
Comments like this jeopardize the investment of a bunch of hard-working JBoss developers and an even larger number of corporate developers that are attempting to server as evangelists for the product. James politely made perfectly valid comments about his experience and offered intelligent suggestions for the product. In the best interest of JBoss, he even offered to help. Marc, I know you and others are working extremely hard to make this a world-class product, and you're doing a great job. But a world-class product obviously has to have customers, and those customers deserve respect. Sincerely, Brian Sondergaard P.S. I'll buy you a pitcher (or two) next week in SF. - Original Message - From: marc fleury [EMAIL PROTECTED] To: Trawick, James [EMAIL PROTECTED]; Jboss-Development@Lists. Sourceforge. Net [EMAIL PROTECTED]; Jboss-User@Lists. Sourceforge. Net [EMAIL PROTECTED] Sent: Thursday, March 14, 2002 10:22 PM Subject: RE: [JBoss-dev] RE: [JBoss-user] JBOSS 3.x FINAL mr trawick james, suck my dick marcf |-Original Message- |From: [EMAIL PROTECTED] |[mailto:[EMAIL PROTECTED]]On Behalf Of |Trawick, James |Sent: Thursday, March 14, 2002 12:33 PM |To: Jboss-Development@Lists. Sourceforge. Net; Jboss-User@Lists. |Sourceforge. Net |Subject: RE: [JBoss-dev] RE: [JBoss-user] JBOSS 3.x FINAL | | |Not unfounded, personally proven. | |In any case there needs to be some sort of kickstart and executive summary |for the new JBoss features. If it means that I'll finally get the answers |I'm looking for without holding for three hours and/or putting up with the |hellfire here, I'll be more than happy to help produce that. | |I was not aware that the docs for sale were actually available. My |experience has been the exact opposite. On three separate occaisons over |the past couple of months I've attempted to purchase the clustering |documentation. Each time I received a message saying that there was a |problem with their side of the transaction and a support rep would be in |contact with me. Despite multiple attempts to open communications about |these cases with Flashline, they have failed to do so. Emails apparently |went to the bit bucket and their phone support reps have proven to be |anywhere from rude to utterly incompetant. So for my purposes, the docs |aren't real because for whatever reason I can't get to them despite my |willingness to pay the mere $10 for quality documentation. At least they |didn't charge me for it. | |chris | |-Original Message- |From: Bill Burke [mailto:[EMAIL PROTECTED]] |Sent: Thursday, March 14, 2002 3:10 PM |To: Trawick, James |Cc: Jboss-Development@Lists. Sourceforge. Net; Jboss-User@Lists. |Sourceforge. Net |Subject: RE: [JBoss-dev] RE: [JBoss-user] JBOSS 3.x FINAL | | |Clustering docs are available for purchase. | |http://www.flashline.com/components/productsbyvendor.jsp?vendorid= |1376affi |liateid=260343 | |I've also written an article on clustering JBoss 2.4.x | |http://www.onjava.com/pub/a/onjava/2001/09/18/jboss.html | |Clustering features in JBoss beta: |HA-JNDI, HA-RMI, EJBs(EB, SLSB, SFSB) Failover and load-balancing | |The code for HTTP Session Failover is in CVS, but is a work in progress. |Ask Sacha and Julian how it's going. I don't work on this stuff. HTTP |Session Failover will probably only be available for Jetty. | |I don't know what you're talking about when you say docs for sale aren't |real, because we've already sold at least a 100 copies of the clustering |docs. Sacha and I have gotten good feedback on these docs so your |complaints are unfounded. | | -Original Message- | From: [EMAIL PROTECTED] | [mailto:[EMAIL PROTECTED]]On Behalf Of | Trawick, James | Sent: Thursday, March 14, 2002 2:28 PM | To: 'David Ward' | Cc: 'marc fleury'; Jboss-Development@Lists. Sourceforge. Net; | Jboss-User@Lists. Sourceforge. Net | Subject: [JBoss-dev] RE: [JBoss-user] JBOSS 3.x FINAL | | | What I'm really interested in is the more corporate questions |(we're still | battling for JBoss over Weblogic, and clustering is a real sore | point). How | does JBoss 3 clustering interact with web container clustering? | If they're | completely separate (which I'm pretty sure is the case, as it was | last time | I walked through the code), how can I put together a full clustered J2EE | stack with JBoss and (for example) Tomcat/Catalina? To what level does | JBoss provide high-availability? What are some failover scenarios? How | does one initiate a graceful failover for server maintenance? How | flexible/controllable is JBoss clustering in reality? | | |Failover scenarios? |- failed network card |- server crash |- application crash |- What else? | |Initiate graceful failover for Server maintainance? |Just gracefully shutdown jboss. The Cluster dynamically figures out |topology. Topology information is piggy-backed over invocation
Re: [JBoss-dev] JBOSS @ JAVAONE TSHIRT CONTEST
For what it's worth, I think this is great! Go Jules! May the Source be with you. - Original Message - From: Jules Gosnell [EMAIL PROTECTED] Cc: 'marc fleury' [EMAIL PROTECTED]; 'Jboss-Development@Lists. Sourceforge. Net' [EMAIL PROTECTED] Sent: Thursday, February 07, 2002 7:20 PM Subject: Re: [JBoss-dev] JBOSS @ JAVAONE TSHIRT CONTEST OK - I know it's corny, but we haven't had any Star Wars references: JBoss - May the Source be with you Someone must be able to come up with a quote about coming over to the resistance, vanquishing the dark side, etc. Jules _ Do You Yahoo!? Get your free @yahoo.com address at http://mail.yahoo.com ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] That flaming fireball in the sky...
Write Once Run Anywhere? - Original Message - From: Bill Burke [EMAIL PROTECTED] To: Jboss-Development@Lists. Sourceforge. Net [EMAIL PROTECTED] Sent: Friday, February 01, 2002 8:36 AM Subject: RE: [JBoss-dev] That flaming fireball in the sky... Here's the full quote. What the hell is WORA? Marinescu: What is Sun's point of view in the debate over whether J2EE licensing restricts open source J2EE products? Shannon: Sun participates in open source because it helps spark innovation, improve software quality, and fosters community. The source code for the J2EE reference implementation is made available publicly as part of the Sun Community Source Licensing program. Sun supports open source developers because their efforts are consistent with Sun's own computing vision which uses open standards and non-proprietary interfaces. Sun's goal is to make our software as open as possible while ensuring portability and WORA. Tegan: At the same time, having a strong brand and compatibility standards are important to the development of a robust market for J2EE platform products, tools, and components. The J2EE Compatible brand has achieved significant momentum over the past two years, and we want to make sure that any open source efforts don't impact the viability of that effort. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of marc fleury Sent: Friday, February 01, 2002 2:16 AM To: Dain Sundstrom; Jboss-Development@Lists. Sourceforge. Net Subject: RE: [JBoss-dev] That flaming fireball in the sky... ignore the noise, focus on the work marcf PS: I agree with your analysis, the J2EE certified brand is weaker for every download we have and we are not certified. Having the most downloaded server not certified takes a lot of credibility out of the brand. I was reading the forum this was posted on and found interesting to hear that people say JBoss has become a defacto-standard of sorts. I believe that statement to be true at least for the development. Look at the download numbers we have 50k month weblogic has 3k for their free version. For the production I am waiting for a report that takes into account free and for pay and counts CPUs installed, I wouldn't be surprised the Apache story would repeat itself. But for now focus on the download numbers. The rest is noise. Noise, bla bla bla. |-Original Message- |From: [EMAIL PROTECTED] |[mailto:[EMAIL PROTECTED]]On Behalf Of Dain |Sundstrom |Sent: Thursday, January 31, 2002 7:01 PM |To: Jboss-Development@Lists. Sourceforge. Net |Subject: [JBoss-dev] That flaming fireball in the sky... | | |I found this on the apache web site. I interpret this statement slightly |different the Jakarta guys. To me it looks like the J2EE licensees are |getting nervous about the high quality of the open source J2EE |implementations (i.e., JBoss), and feel the only way to distinguish them |selves in the marketplace is with some lame certification. | |- |30 January 2002 - That flaming fireball in the sky... | |In a recent article, Karen Tegan, Director of J2EE Compatibility and |Platform Services for Sun Microsystems, had the following to say: | |The J2EE Compatible brand has achieved significant momentum over the past |two years, and we want to make sure that any open source efforts don't |impact the viability of that effort. | |In other words, Sun doesn't give a hoot about whether J2EE licensing |restricts open source J2EE products (in case you missed it, it does). Thus, |the Apache Software Foundation's involvement in the Java Community Process |(JCP) is simply an advertising statement for Sun to claim that they have a |'vision which uses open standards and non-proprietary interfaces'. If you |would like to express your opinions of Sun's licensing terms, feel free to |contact [EMAIL PROTECTED] and let us know what you think. Thanks. | |Followup thread on Jakarta-General: J2EE considered harmful. | |___ |Jboss-development mailing list |[EMAIL PROTECTED] |https://lists.sourceforge.net/lists/listinfo/jboss-development ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] JAAS with MDB
I may have misunderstood, but aren't we talking about run-as in the
security-identity?
- Original Message -
From: Scott M Stark [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Sent: Tuesday, January 29, 2002 6:06 PM
Subject: Re: [JBoss-dev] JAAS with MDB
MDBs need a different SecurityInterceptor that does not attempt
to authenticate the caller as there is none defined. An extension
would allow the caller to be obtained from JMS message properties.
The properties could be specified in the jboss.xml descriptor in
the message-driven bean section.
Scott Stark
Chief Technology Officer
JBoss Group, LLC
- Original Message -
From: Dave Smith [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Sent: Tuesday, January 29, 2002 2:08 PM
Subject: [JBoss-dev] JAAS with MDB
I am trying to get a MDB working with JAAS. Currently when message
delivery is attempted it throws an error Authentication exception,
principal=null. Now in the jms-service.xml I have added
attribute name=PrincipalMappingProperties
UserName=jboss
Password=jboss1
/attribute
but this does not do anything. I can not find any refences to this in
either the free or purchased docs. Can some one point me in the right
direction.
time passes ...
After further investigation it looks like it is not implemented. In the
JMSContainerInvoker it creates the method invocation with principal and
credential = null. Then when the SecurityInterceptor gets a hold of it
barfs. The snip, line 976 of JMSContainerInvoker.java
invoker.invoke(id,
// Object id - where used?
ON_MESSAGE,
// Method to invoke
new Object[]{message},
// argument
tm.getTransaction(),
// Transaction
null,
// Principal
null);
// Cred
How should I go about getting this to work?
___
Jboss-development mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-development
___
Jboss-development mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-development
___
Jboss-development mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] Tomcat 4.0 with JBoss 3.0alpha
I'm sure I'm overlooking the key issues here, but is it currently possible to use JBoss (3.0a) with Tomcat (4.0x)? I got the catalina plug-in down but can't build it. It's apparently associated with the JBoss 2.4, not 3.0. More specifically, it's looking for classes like org.jboss.security.SecurityDomain that I think are 2.4 things. Thanks for any insight. Brian Sondergaard
[JBoss-dev] Please help: ClassLoader Confusion
I'm kind of in a bind, and I'm hoping someone will nudge me in the right direction... I have MQSeries working quite well with JBoss (with MDBs and all), but I'm having one problem. When I use an ObjectMessage, I get an exception on the call to getObject(). The exception is ultimately a ClassNotFoundException coming out of MQ when it attempts to deserialize the contained object. Here's the MQ Series ObjectMessage implementation that throws the ClassNotFoundException: ByteArrayInputStream bytearrayinputstream = new ByteArrayInputStream(messageBytes, dataStart, messageBytes.length - dataStart); ObjectInputStream objectinputstream = new ObjectInputStream(bytearrayinputstream); serializable = (Serializable)objectinputstream.readObject();The exact same code works perfectly outside of JBoss, and it seems clear that I'm running into a problem with classloaders. The object contained within the ObjectMessage is of a type found in my application. This type is used in many other places in the app (including the sending of the message) and works fine. The MQ Series classes are located in lib/ext, and since JBoss gets the connections/sessions started and I'm able to send and receive messages, they also seem to work fine. It appears that the classloader that is being used to attempt to deserialize the contained object (the CL that loaded MQ classes?) cannot see the classes in the app, thus the ClassNotFoundException. However, the MQ classes have to come from lib/ext (or another server-level context), since they're used by various mbeans during startup.I'm probably overlooking something here, but this seems to be a catch 22.This is a build of 2.5 from 7/30. By the way, everything's cool if I stuff a String into the ObjectMessage
Re: [JBoss-dev] Please help: ClassLoader Confusion
If I wrap the contained object in a java.rmi.MarshalledObject (like JBossMQ does in its ObjectMessage implementation) it works. When sending the message: ObjectMessage message = session.createObjectMessage(); SomeObject obj = new SomeObject( "Parms" ); message.setObject( new MarshalledObject(obj) ); When receiving: MarshalledObject mobj = (MarshalledObject)objMessage.getObject(); SomeObject obj = (SomeObject) mobj.get(); This really seems to be an issue with how the ClassLoaders are dealt with in JBoss, as it seems this code would not be completely portable. Is there some better way to work around this problem? - Original Message - From: Brian Sondergaard To: jboss dev Sent: Thursday, August 09, 2001 11:02 AM Subject: [JBoss-dev] Please help: ClassLoader Confusion I'm kind of in a bind, and I'm hoping someone will nudge me in the right direction... I have MQSeries working quite well with JBoss (with MDBs and all), but I'm having one problem. When I use an ObjectMessage, I get an exception on the call to getObject(). The exception is ultimately a ClassNotFoundException coming out of MQ when it attempts to deserialize the contained object. Here's the MQ Series ObjectMessage implementation that throws the ClassNotFoundException: ByteArrayInputStream bytearrayinputstream = new ByteArrayInputStream(messageBytes, dataStart, messageBytes.length - dataStart); ObjectInputStream objectinputstream = new ObjectInputStream(bytearrayinputstream); serializable = (Serializable)objectinputstream.readObject();The exact same code works perfectly outside of JBoss, and it seems clear that I'm running into a problem with classloaders. The object contained within the ObjectMessage is of a type found in my application. This type is used in many other places in the app (including the sending of the message) and works fine. The MQ Series classes are located in lib/ext, and since JBoss gets the connections/sessions started and I'm able to send and receive messages, they also seem to work fine. It appears that the classloader that is being used to attempt to deserialize the contained object (the CL that loaded MQ classes?) cannot see the classes in the app, thus the ClassNotFoundException. However, the MQ classes have to come from lib/ext (or another server-level context), since they're used by various mbeans during startup.I'm probably overlooking something here, but this seems to be a catch 22.This is a build of 2.5 from 7/30. By the way, everything's cool if I stuff a String into the ObjectMessage
[JBoss-dev] How to run the latest jbossmq
I'm trying to run the latest jbossmq. I brought down the code and everything built fine, but when I try to combine this with the latest JBoss, I run into what appears to be configuration troubles. I've studied closely the changes to the JBossMQ config, but I must be overlooking something. When I try to start the server, all goes well until it attempts to deploy any MessageDrivenBeans, at which point the following occurs during initialization: Serious error in init: javax.naming.NameNotFoundException: INVMXAQueueConnectionFactory not bound javax.naming.NameNotFoundException: INVMXAQueueConnectionFactory not bound The latest JBoss runs fine if I don't update JBossMQ. Thanks for your help. Brian
