How do I configure the number of threads available in Jboss.
I have seen posts on doing this in Jetty on this list, but
not for Jboss itself in terms of available threads for
message driven Beans etc.
Thanks.
David
---
This SF.net email is
Is the JSP and SessionBean within same VM? Then just do new
InitialContext();
make sure that jboss/server/xxx/conf/jndi.properties has the Provider
URL commented out. This will force JBoss to connect locally rather than
going over net.
Maybe you have a jndi.properties file? Or maybe you hav
Thanks Bill, Scott posted this as well. I am using that but here is the
problem
1. Have a box with three virtual IP addresses we will call default,
primary and secondary.
2. Launch JBoss on primary using "--host primary" with nothing running
on the others.
3. Telnet to primary port 1099 reveals
Oh how clever, sorry the correct URL is:
http://tinyurl.com/wpi1
-Neal
On November 26, 2003 04:50 pm, Neal Sanche wrote:
> My patch was successful, I've started an issue on the XDoclet site
> for this:
>
> http://tinyurl.com/czcn
>
> I've tested this, and it works, putting my MDB beans into the
My patch was successful, I've started an issue on the XDoclet site for
this:
http://tinyurl.com/czcn
I've tested this, and it works, putting my MDB beans into the JNDI
tree with nice names.
-Neal
On November 26, 2003 04:25 pm, Neal Sanche wrote:
> I finally got around to checking out why XD
System.getProperty("jboss.bind.address");
Rod Macpherson wrote:
I use this to get the JNDI properties and bind Globals.HOST in a static
class initializer using
java.net.InetAddress.getLocalHost().getHostName(). The problem is that
method returns the hostname of my box NOT the hostname that JB
I finally got around to checking out why XDoclet doesn't honour the
local-jndi-name @ejb.bean attribute when generating MDB beans. In
jboss_xml.xdt the generation section doesn't do
anything with it, so effectively it can't generate the parameter
without a patch. I'm currently writing such a p
I'm not using mod_jk2.
Sacha Labourey wrote:
Rémy posted a messsage recently on tomcat-dev: there seem to be a bug in the
latest mod_jk2 release. Can you try a mod_jk (release 1.xxx) instead?
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of
Mike Linds
We do not use WEB-INF/lib or WEB-INF/classes to avoid this. Instead we share jars
across all tiers and stipulate what jar uses what in the Class-Path for each jar we
assemble. We are guaranteed that there is one and only one image of a class in the
whole system and, the class loader being part o
Here's the gook.
C:\ Telnet sandbox 1099
ÂÏsrâjava.rmi.MarshalledObject|ââÃÏcâ>ââIâhashlocBytestâ[BobjBytesq~âxpÃ?ÂÎurâ[BÂââÂTÎâxp)ÂÏthttp://sandbox:8083/q~q
~uq~ââÂÏsr
org.jnp.server.NamingServer_Stubââxrjava.rmi.server.RemoteStubÎâââÃÃeâxrâjava.rmi.server.RemoteObjectâaâÃ
Rémy posted a messsage recently on tomcat-dev: there seem to be a bug in the
latest mod_jk2 release. Can you try a mod_jk (release 1.xxx) instead?
> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] On Behalf Of
> Mike Lindsey
> Sent: mercredi, 26. novembre 2003 2
Hrm. Not working for me. Ideas? (If the solution is glaringly simple,
forgive me. I've got the flu)
If I stick two jvmRoute statements in the engine block, I get this error:
10:30:41,875 INFO [MainDeployer] Starting deployment of package:
file:/usr/local/jboss-3.2.2/server/test/deploy/jbos
OK, added the code java.bind.address to the JNDI properties and now we
get the IP proper IP address back but still connection refused. Again,
this is only when using the --host switch and the same thing works fine
when multihoming my home PC.
javax.naming.CommunicationException [Root exception is
Show the gook as this is the marshalled rmi stub that the NamingContext
is trying to use, and it contains the address/port the is causing
the connection failure. SANDBOX:1099 is only the bootstrap port used
to obtain the rmi stub. The rmi address should be the same as the
--host value.
--
x
I found the source of my problem. Thankfully I came across Scott's classloading debug
sample in the Jboss 3.2 admin doc. This revealed that I did in fact have another copy
of the class available to my web app - it had been unexpectedly included in a
dependency jar in WEB-INF/lib even though tha
Thanks, I will try that. All the sudden started getting this error on
any machine that I try to multihome at the office. It used to work
AFAIK. Works if no --host is specified, otherwise my session bean
lookups generate this:
10:25:21,281 WARN [NamingContext] Failed to connect to SANDBOX:1099
jav
Title: Message
When
running multihome I am getting this on all of the office machines but not on my
home network: same version of JBoss, same JDK, same application works fine at
home.
[NamingContext] Failed to connect to
SANDBOX:1099
Ironically this used to work on this very machine. W
System.getProperty("jboss.bind.address") == --host value
--
Scott Stark
Chief Technology Officer
JBoss Group, LLC
Rod Macpherson wrote:
I use this to get the JNDI properties and bind Globals.HOST in a static
class initializer using
java.net.InetA
Title: Message
I use this to get
the JNDI properties and bind Globals.HOST in a static class initializer using
java.net.InetAddress.getLocalHost().getHostName(). The problem is
that method returns the hostname of my box NOT the hostname that JBoss was
started with using the --host switch. Gi
Olve Hansen wrote:
On 25.11.2003 03:27 Mike Lindsey wrote:
I've got apache talking to a clustered jboss setup, using mod_jk, but
I cannot get the load balacing to work. I can force connections to a
specific app server, but no matter what I do, I can't get the initial
connection to hit a 'ran
I'll try that today.
Bob Cotton wrote:
Make sure the tomcat config includes a
jvmRoute="node1"
in your
for each of the nods.
-Bob
Mike Lindsey <[EMAIL PROTECTED]> writes:
I've got apache talking to a clustered jboss setup, using mod_jk, but
I cannot get the load balacing to work
On Wed, 2003-11-26 at 16:34, Bill Burke wrote:
> Actually, turning it off in transaction-service.xml has a tiny
> performance improvement as well.
It is a bit more than tiny since the WeakHashMap storing
the debug information is a global point of contention.
Regards,
Adrian
>
> Adrian Brock wr
Actually, turning it off in transaction-service.xml has a tiny
performance improvement as well.
Adrian Brock wrote:
On Sat, 2003-11-15 at 01:22, mark pope wrote:
Hi,
I'm passing a JDBC Connection between two stateless beans. The called bean uses
it and closes its Statement and ResultSet. When
Greetings,
What means the cluster of ejb? Its behavior must differs for each type
of bean, right?
I put below a few more questions about the clustering of ejb:
- If I put EJBs in the farm, they will be deployed in an jboss
instances. Without any jboss cluster descriptors in them, the farm only
Greetings,
I wonder what should I use for executing some business logic and call
stored procedures: ordinary classes or stateless session beans?
For stateless session beans it's possible to cache somewhere the
datasource factory or must I lookup always for the jndi connection pool?
thanks.
P.S.
On Wed, 2003-11-26 at 13:53, Alexey Loubyansky wrote:
> I would think this is absolutely fine if the bean is read-only=false and
> just the method is read-only=true.
> If caching is the requirement then commit option A should be used. What
> do you think, Adrian?
>
I don't see any other way to
On Wed, 2003-11-19 at 15:36, Didi Posselt wrote:
> Hi,
>
> Since the option --host=yourhost causes lots of problems in applications
> that don't specify a host to connect to, I cannot use it for my JMS-problem.
>
> I figured, that everything works fine, if I drop the Attribute BindAddress
> in th
I've found the source of my problem. Thankfully I came across Scott's classloading
debug sample in the Jboss 3.2 admin doc. This revealed that I did in fact have another
copy of the class available to my web app - it had been unexpectedly included in a
dependency jar in WEB-INF/lib even though
You need both.
The client id stops other people trying to access the same
subscription at the same time.
Likely, you've been using anonymous client ids ID:1 or similar
which won't necessarily be the same next time you restart your MDB.
Regards,
Adrian
On Tue, 2003-11-18 at 06:15, Brian Wallis wr
I would think this is absolutely fine if the bean is read-only=false and
just the method is read-only=true.
If caching is the requirement then commit option A should be used. What
do you think, Adrian?
Adrian Brock wrote:
The obvious fix would be to add what is done for no transaction in the
Ent
On Sat, 2003-11-15 at 01:22, mark pope wrote:
> Hi,
>
> I'm passing a JDBC Connection between two stateless beans. The called bean uses
> it and closes its Statement and ResultSet. When the method returns to the
> caller I get a WrappedConnection msg on the console. Both EJBs have local
> interfac
This method says the connection is already enlisted in a transaction
and you are trying to use setAutoCommit() to change that attribute.
Obviously, that is not allowed. My guess is that you already
asked for the NonXA variant already in the transaction.
Regards,
Adrian
On Fri, 2003-11-14 at 17:5
On Wed, 2003-11-26 at 12:47, Alexey Loubyansky wrote:
> Thanks for the test.
> As for me, it is definitely a bug and it should work with any commit option.
> read-only is currently an overloaded notion. read-only in jboss.xml
> should have been better named no-lock or like that to avoid the
> con
At 13:47 26.11.2003, you wrote:
Thanks for the test.
:) no problem.
As for me, it is definitely a bug and it should work with any commit option.
Would you fix this or somebody else? And will be thix fixed in 3.2.3 release?
Regards,
Rafal
---
Thanks for the test.
As for me, it is definitely a bug and it should work with any commit option.
read-only is currently an overloaded notion. read-only in jboss.xml
should have been better named no-lock or like that to avoid the
confusion with real read-only from jbosscmp-jdbc.xml.
Rafal Kedzio
At 11:54 26.11.2003, Alexey Loubyansky wrote:
Is the table updated from another application? There are no UPDATEs in the
log.
I wanted you to try with read-only = false for the methods that have
read-only = true.
o.k. I tryied this and now commit option C is working.
Is this now a small "bug" o
Rafal Kedziorski wrote:
At 11:54 26.11.2003, Alexey Loubyansky wrote:
Is the table updated from another application? There are no UPDATEs in
the log.
The table was updated from JBoss thru direct SQL.
I wanted you to try with read-only = false for the methods that have
read-only = true.
o.k
At 11:54 26.11.2003, Alexey Loubyansky wrote:
Is the table updated from another application? There are no UPDATEs in the
log.
The table was updated from JBoss thru direct SQL.
I wanted you to try with read-only = false for the methods that have
read-only = true.
o.k. I have missunderstand you. I
> app I get a ClassCastException doing PortableRemoteObject.narrow() on
> the remote interface
Sorry, I meant the Home interface.
> -Original Message-
> From: Alastair Rodgers
> Sent: 26 November 2003 11:22
> To: '[EMAIL PROTECTED]'
> Subject: Classloader issues
>
>
> Hi,
>
> I h
Aha! :-)
This would explain why I am getting locking even when I'm not expecting
it.
Thanks guys
On Tue, 2003-11-25 at 17:37, Alexey Loubyansky wrote:
> The read-only is applied to the specific bean/method as defined in the
> DDs. No inheritance rules.
--
Peter Beck BEng (hons) MIEE GIinstIT -
Hi,
I have a legacy EAR which is currently deployed on JBoss 3.0.8. This contains a WAR &
an EJB JAR, and the web app invokes an EJB in the EJB JAR. This has always worked fine
using the UCL and with Jetty's Java2ClassLoadingCompliance set to true.
However, I now need to deploy a new EAR alon
Is the table updated from another application? There are no UPDATEs in
the log.
I wanted you to try with read-only = false for the methods that have
read-only = true.
Rafal Kedziorski wrote:
At 15:39 25.11.2003, Alexey Loubyansky wrote:
Could you try with false?
We don't try this, cause by
At 15:39 25.11.2003, Alexey Loubyansky wrote:
Could you try with false?
We don't try this, cause by default all entity beans are not read-only. But
we try this. We user for all our entity beans this container configuration:
Standard CMP 2.x EntityBean. And we changed commit option (line 227,
sta
Pete Beck wrote:
Yep, we've walked that same path too.
Unfortunately there are few long running transactions which need write
access and these seem to screw things up.
We also have long running transactions, calling potentially external
applications from our workflow app, but we have chosen to re
44 matches
Mail list logo
