Hi all,
Does commit option A currently have
any meaning in the context of a clustered JBoss instance (since there's
no distributed locking and presumably the preload cache is not distributed)?
Dave
Hi,
Is it possible to have multiple WARs in an EAR which share classes? I have
tried this on a project we're on, with what seemed like all of the correct
module declarations in the application.xml and also w/ a Class-Path entry
in the manifests of the WARs. I've tried several permutations of
That's interesting.
The web testsuite in JBoss seems to
have an example of a test which demonstrates using the Class-Path manifest
entry, but it's in a jar inside the war file, not the war file's manifest.
I wonder if that's the secret?
Also, I am using Class.forName(), I
will investigate that
Hi all,
I'm getting an EJB spec violation warning
on deployment:
2003-06-09 19:23:44,734 WARN [org.jboss.ejb.EJBDeployer.verifier]
EJB spec violation:
Bean : MyBean
Method : public abstract Collection
findDynamically(String, Object;) throws FinderException
Section: 10.5.6
Warning: Every
Entity Beans can have both remote and
local interfaces, that's not a problem.
Finders and select methods are two totally
different things. Secion 10.5.7 of the spec states that the bean
class itself is where select methods reside, not the home interface.
If you just want a
Never mind, I found the answer in Dain's
excellent CMP docs. I will put some dumb basic query in the ejb-jar.xml.
Dave
[EMAIL PROTECTED]
Sent by: [EMAIL PROTECTED]
06/10/2003 10:19 AM
Please respond to jboss-user
To:
[EMAIL PROTECTED]
cc:
Subject:
[JBoss-user]
OK, I take it back, this problem is
not solved.
Dain's CMP docs say that an empty ejb-ql
element is all that's required for the ejb-jar.xml, and that the query
in jboss-jdbc.xml will override that.
Well, I've got an empty ejb-ql
element in the query element for the bean in ejb-jar.xml, I've
got
Brian,
The XDoclet utilobject pattern caches
Remote and (Local Homes the latter is of course only valid w/in one JVM),
I've seen this mentioned on theserverside.com and elsewhere as a best practice,
and after digging through the spec to be sure, apparently both Remote Homes
and Remote interfaces
No offense meant, but this group is
for technical questions about JBoss right? I would hope the Core Developers
Network folks from JBG will still post here and answer questions, but aside
from that, I have a ton of lists to read and a job to do and if I want
gossip and intrigue, I'll stay home
It's not allowed as a finder
(if you mean a method named findXXX in the bean Home class), but it's valid
EJB-QL if its in a select method (private method -- see note
below -- called ejbSelectXXX in the bean class, that's called from a business
method -- i.e. getSomeFieldBasedOnQuery()).
Finders,
Whoops, I totally misunderstood your
question (obviously). The spec says that EJB-QL operates on abstract
schema types. I would imagine that, for DVCs (which is what we're
talking about here, right?) the abstract schema type is just Object.
I don't see anything in the spec which contradicts it,
The findByAge(Inteter) one works fine
by itself? I.e., you verfied that the syntax is valid for Integers but
not Objects? What's the SQL-type/JDBC-type your CMP POJO class is mapped
as?
Dave
Rupp,Heiko [EMAIL PROTECTED]
Sent by: [EMAIL PROTECTED]
06/05/2003 12:00 PM
Please respond to
Handles are serializeable, refs aren't.
But for Entity beans anyway, refs are supposed to survive a restart. From
the spec:
Each entity object has an identity
which, in general, survives a crash and restart of the container in
which the entity object has been created.
The object identity is
ObjectNotFoundException is a subclass
of FinderException.
NoSuchObjectExsists is what gets throw
if you reference an actual bean interface method on an Entity that's beed
removed.
Dave
Jon Haugsand [EMAIL PROTECTED]
Sent by: [EMAIL PROTECTED]
06/06/2003 10:41 AM
Please respond to
Finder methods always return the type
in whose home interface they are defined, according to the spec: The
return
type of a finder method on the local
home interface must be the entity beans local interface, or a type
representing a collection of objects
that implement the entity beans
Are LIMIT OFFSET supported in
JBossQL in 3.2.x? Or just in 4.x? I can't find where in the code it's supported
(haven't tried to use it yet).
Dave
Well, that's a bummer, but it would
explain why I can't find any source which mentions either of them.
:-(
Dave
Jonathan.O'[EMAIL PROTECTED]
Sent by: [EMAIL PROTECTED]
06/04/2003 11:23 AM
Please respond to jboss-user
To:
[EMAIL PROTECTED]
cc:
Subject:
Re:
If you want them to be nullable, they
have to be Objects (primitives get initialized to their default value).
Dave
Brian McSweeney [EMAIL PROTECTED]
Sent by: [EMAIL PROTECTED]
05/29/2003 05:35 AM
Please respond to jboss-user
To:
[EMAIL PROTECTED]
cc:
Subject:
Not only that, but w/ commit option B it seems that there's info left in
the context that may/may not be enough to keep some of the relevant data
(where to start reading again, etc) asl well. I'm looking at the code to
try to see how it all hangs together.
In a couple of days I'll be a
19 matches
Mail list logo
