You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tomee.apache.org by Manu George <ma...@gmail.com> on 2006/12/16 16:26:29 UTC
Bug in MDB container
Hi Dain,
Here is an issue for you to fix in the MDB container :-).
http://issues.apache.org/jira/browse/OPENEJB-409
Some questions also
Will there be an AMQ instance running in the same VM as OEJB? Looking
at the code there seems to be a connector to connect to 51616 port as
per the conf now. Does this mean we need to start a separate
independent AMQ instance at this port. If not how to start the
embedded AMQ in OEJB
Thanks
Manu
Re: Bug in MDB container
Posted by David Blevins <da...@visi.com>.
On Dec 19, 2006, at 9:30 PM, David Blevins wrote:
>
> On Dec 17, 2006, at 1:40 AM, Manu George wrote:
>
>> Hi Mohammad,
>> Yes we can run an independent version of the MDB
>> container. I am able to run an MDB after making some changes to the
>> codebase and running a separate amq instance in that port. There
>> are a
>> number of changes that needed to be made to the codebase to make the
>> MDB work.
>
> Impressive patch. That code was more or less thrown in pre-MDB-
> container hence the odd mix of things going to the stateless
> container -- very tricky to navigate. You did a great job finding
> your way through misleading code. Kudos.
>
> I'm going to see if I can't write a ServerService to boot the
> ActiveMQ broker in the same vm so we can add some MDB tests to the
> test suite.
>
> Pretty exciting!
Ok, I have a basic ActiveMQ BrokerService running now on port 4206.
This runs only in the openejb-standalone assembly, so it doesn't help
much with the embedded testing thing but it will work when unzipping
and running a standalone install.
Going to see if I can't come up with an option to have an in-vm
embedded broker for the rest of the tests.
-David
> -David
>
>> On 12/17/06, Mohammad Nour El-Din <no...@gmail.com> wrote:
>>> Hi Manu...
>>>
>>> Did you try running a separate AMQ instance ?, I think this is
>>> like running
>>> a separate DBMS and connecting to it through its connector using the
>>> suitable configurations, I am interested to know the result of
>>> this :)
>>>
>>>
>>>
>>> On 12/16/06, Manu George <ma...@gmail.com> wrote:
>>>
>>> > Hi Dain,
>>> > Here is an issue for you to fix in the MDB
>>> container :-).
>>> > http://issues.apache.org/jira/browse/OPENEJB-409
>>> > Some questions also
>>> >
>>> > Will there be an AMQ instance running in the same VM as OEJB?
>>> Looking
>>> > at the code there seems to be a connector to connect to 51616
>>> port as
>>> > per the conf now. Does this mean we need to start a separate
>>> > independent AMQ instance at this port. If not how to start the
>>> > embedded AMQ in OEJB
>>> >
>>> > Thanks
>>> > Manu
>>> >
>>>
>>>
>>>
>>> --
>>> Thanks
>>> - Mohammad Nour
>>>
>>>
>>
>
Re: Bug in MDB container
Posted by Manu George <ma...@gmail.com>.
Hi Dain,
Thanks for the info and the spec reference. It saved me the task of
going through the MDB part once again ;). Thanks again.
Regards
Manu
On 12/21/06, Dain Sundstrom <da...@iq80.com> wrote:
> On Dec 20, 2006, at 3:31 AM, Manu George wrote:
>
> > Hi David,
> > Yes pretty exciting ;). Will be looking forward to that in VM
> > service being there. There is another thing that is missing in the
> > code for MDB and that is Transaction Support.Should I open a JIRA for
> > that? Actually most of the code is there but there are some gaps here
> > and there which needs to be plugged. I am trying to add that after
> > investigating how its done for Session beans. Now for an MDB with
> > container managed transaction demarcation the only two supported
> > Transaction policies are Required and NotSupported. What should be the
> > behaviour if the user specifies any other Transaction policy in the
> > deployment descriptor? Maybe I did not look hard enough, but I am
> > unable to find the expected behaviour for this.
> > Should it be defaulted to NotSupported or should there be an error
> > thrown during the deployment process.
>
> Ya, that is part of the EJB spec (3.0 section 5.4.12). The only
> allowed transaction types for a MDB is required, not supported and
> bean managed. For the timeout method, the spec additionally allows
> the requires new transaction attribute.
>
> -dain
>
Re: Bug in MDB container
Posted by Dain Sundstrom <da...@iq80.com>.
On Dec 20, 2006, at 3:31 AM, Manu George wrote:
> Hi David,
> Yes pretty exciting ;). Will be looking forward to that in VM
> service being there. There is another thing that is missing in the
> code for MDB and that is Transaction Support.Should I open a JIRA for
> that? Actually most of the code is there but there are some gaps here
> and there which needs to be plugged. I am trying to add that after
> investigating how its done for Session beans. Now for an MDB with
> container managed transaction demarcation the only two supported
> Transaction policies are Required and NotSupported. What should be the
> behaviour if the user specifies any other Transaction policy in the
> deployment descriptor? Maybe I did not look hard enough, but I am
> unable to find the expected behaviour for this.
> Should it be defaulted to NotSupported or should there be an error
> thrown during the deployment process.
Ya, that is part of the EJB spec (3.0 section 5.4.12). The only
allowed transaction types for a MDB is required, not supported and
bean managed. For the timeout method, the spec additionally allows
the requires new transaction attribute.
-dain
Re: Bug in MDB container
Posted by Manu George <ma...@gmail.com>.
Hi David,
Yes pretty exciting ;). Will be looking forward to that in VM
service being there. There is another thing that is missing in the
code for MDB and that is Transaction Support.Should I open a JIRA for
that? Actually most of the code is there but there are some gaps here
and there which needs to be plugged. I am trying to add that after
investigating how its done for Session beans. Now for an MDB with
container managed transaction demarcation the only two supported
Transaction policies are Required and NotSupported. What should be the
behaviour if the user specifies any other Transaction policy in the
deployment descriptor? Maybe I did not look hard enough, but I am
unable to find the expected behaviour for this.
Should it be defaulted to NotSupported or should there be an error
thrown during the deployment process.
Thanks
Manu
On 12/20/06, David Blevins <da...@visi.com> wrote:
>
> On Dec 17, 2006, at 1:40 AM, Manu George wrote:
>
> > Hi Mohammad,
> > Yes we can run an independent version of the MDB
> > container. I am able to run an MDB after making some changes to the
> > codebase and running a separate amq instance in that port. There are a
> > number of changes that needed to be made to the codebase to make the
> > MDB work.
>
> Impressive patch. That code was more or less thrown in pre-MDB-
> container hence the odd mix of things going to the stateless
> container -- very tricky to navigate. You did a great job finding
> your way through misleading code. Kudos.
>
> I'm going to see if I can't write a ServerService to boot the
> ActiveMQ broker in the same vm so we can add some MDB tests to the
> test suite.
>
> Pretty exciting!
>
> -David
>
> > On 12/17/06, Mohammad Nour El-Din <no...@gmail.com> wrote:
> >> Hi Manu...
> >>
> >> Did you try running a separate AMQ instance ?, I think this is
> >> like running
> >> a separate DBMS and connecting to it through its connector using the
> >> suitable configurations, I am interested to know the result of
> >> this :)
> >>
> >>
> >>
> >> On 12/16/06, Manu George <ma...@gmail.com> wrote:
> >>
> >> > Hi Dain,
> >> > Here is an issue for you to fix in the MDB
> >> container :-).
> >> > http://issues.apache.org/jira/browse/OPENEJB-409
> >> > Some questions also
> >> >
> >> > Will there be an AMQ instance running in the same VM as OEJB?
> >> Looking
> >> > at the code there seems to be a connector to connect to 51616
> >> port as
> >> > per the conf now. Does this mean we need to start a separate
> >> > independent AMQ instance at this port. If not how to start the
> >> > embedded AMQ in OEJB
> >> >
> >> > Thanks
> >> > Manu
> >> >
> >>
> >>
> >>
> >> --
> >> Thanks
> >> - Mohammad Nour
> >>
> >>
> >
>
>
Re: Bug in MDB container
Posted by David Blevins <da...@visi.com>.
On Dec 17, 2006, at 1:40 AM, Manu George wrote:
> Hi Mohammad,
> Yes we can run an independent version of the MDB
> container. I am able to run an MDB after making some changes to the
> codebase and running a separate amq instance in that port. There are a
> number of changes that needed to be made to the codebase to make the
> MDB work.
Impressive patch. That code was more or less thrown in pre-MDB-
container hence the odd mix of things going to the stateless
container -- very tricky to navigate. You did a great job finding
your way through misleading code. Kudos.
I'm going to see if I can't write a ServerService to boot the
ActiveMQ broker in the same vm so we can add some MDB tests to the
test suite.
Pretty exciting!
-David
> On 12/17/06, Mohammad Nour El-Din <no...@gmail.com> wrote:
>> Hi Manu...
>>
>> Did you try running a separate AMQ instance ?, I think this is
>> like running
>> a separate DBMS and connecting to it through its connector using the
>> suitable configurations, I am interested to know the result of
>> this :)
>>
>>
>>
>> On 12/16/06, Manu George <ma...@gmail.com> wrote:
>>
>> > Hi Dain,
>> > Here is an issue for you to fix in the MDB
>> container :-).
>> > http://issues.apache.org/jira/browse/OPENEJB-409
>> > Some questions also
>> >
>> > Will there be an AMQ instance running in the same VM as OEJB?
>> Looking
>> > at the code there seems to be a connector to connect to 51616
>> port as
>> > per the conf now. Does this mean we need to start a separate
>> > independent AMQ instance at this port. If not how to start the
>> > embedded AMQ in OEJB
>> >
>> > Thanks
>> > Manu
>> >
>>
>>
>>
>> --
>> Thanks
>> - Mohammad Nour
>>
>>
>
Re: Bug in MDB container
Posted by Manu George <ma...@gmail.com>.
Hi Mohammad,
Yes we can run an independent version of the MDB
container. I am able to run an MDB after making some changes to the
codebase and running a separate amq instance in that port. There are a
number of changes that needed to be made to the codebase to make the
MDB work. I will verify the changes and attach a patch for review.
Thanks
Manu
On 12/17/06, Mohammad Nour El-Din <no...@gmail.com> wrote:
> Hi Manu...
>
> Did you try running a separate AMQ instance ?, I think this is like running
> a separate DBMS and connecting to it through its connector using the
> suitable configurations, I am interested to know the result of this :)
>
>
>
> On 12/16/06, Manu George <ma...@gmail.com> wrote:
>
> > Hi Dain,
> > Here is an issue for you to fix in the MDB container :-).
> > http://issues.apache.org/jira/browse/OPENEJB-409
> > Some questions also
> >
> > Will there be an AMQ instance running in the same VM as OEJB? Looking
> > at the code there seems to be a connector to connect to 51616 port as
> > per the conf now. Does this mean we need to start a separate
> > independent AMQ instance at this port. If not how to start the
> > embedded AMQ in OEJB
> >
> > Thanks
> > Manu
> >
>
>
>
> --
> Thanks
> - Mohammad Nour
>
>
Re: Bug in MDB container
Posted by Mohammad Nour El-Din <no...@gmail.com>.
Hi Manu...
Did you try running a separate AMQ instance ?, I think this is like running
a separate DBMS and connecting to it through its connector using the
suitable configurations, I am interested to know the result of this :)
On 12/16/06, Manu George <ma...@gmail.com> wrote:
> Hi Dain,
> Here is an issue for you to fix in the MDB container :-).
> http://issues.apache.org/jira/browse/OPENEJB-409
> Some questions also
>
> Will there be an AMQ instance running in the same VM as OEJB? Looking
> at the code there seems to be a connector to connect to 51616 port as
> per the conf now. Does this mean we need to start a separate
> independent AMQ instance at this port. If not how to start the
> embedded AMQ in OEJB
>
> Thanks
> Manu
>
--
Thanks
- Mohammad Nour