Eugene,
I see... this isn't really a core ODE problem; it is a bug in the
ODE-JBI integration. I take it you are planning a JBI deployment?
-maciej

On 12/26/06, Eugene P. Kostin <[EMAIL PROTECTED]> wrote:
      Maciej - I was talking about this particular problem:

      ODE-81
      "invoking more than one service causes in 90% a NPE"
(http://issues.apache.org/jira/browse/ODE-81)

      Thanks,
       Eugene


      -----Original Message-----
From: Maciej Szefler [mailto:[EMAIL PROTECTED]
Sent: Friday, December 22, 2006 7:26 PM
To: Eugene P. Kostin
Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED]; [EMAIL PROTECTED];
[EMAIL PROTECTED]; [EMAIL PROTECTED]; [EMAIL PROTECTED];
[email protected]
Subject: Re: ODE clustering support

      Eugene,

      Can you please file JIRA issues for the problems you are seeing?

      Thanks,
      Maciej

      On 12/22/06, Eugene P. Kostin <[EMAIL PROTECTED]> wrote:
      >
      >
      >
      >
      > Hi Matthieu,
      >
      >
      >
      > Thanks for quick response! We will evaluate possibility of us
contributing
      > on clustering. We'd appreciate if you could point us to the places
where
      > clustering support would need to be implemented, so that we could
roughly
      > estimate the level of effort required. And also, do you have a
timeline for
      > fixing most severe defects and have ODE to be in official release?
We
      > started running BPEL processes with the current builds and
immediately faced
      > showstopper problems like inability to call multiple services. If we
commit
      > to adding clustering support we'd like to know how it synchronizes
with the
      > official release plan and that we will have a stable version by the
time we
      > need to deploy it.
      >
      >
      >
      >
      >
      > Thanks,
      >
      >   Eugene
      >
      >
      >
      >  ________________________________
      >
      >
      > From: Matthieu Riou [mailto:[EMAIL PROTECTED]
      >  Sent: Thursday, December 21, 2006 8:10 PM
      >  To: Eugene P. Kostin
      >  Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED]; [EMAIL PROTECTED];
      > [EMAIL PROTECTED]; [EMAIL PROTECTED]; [EMAIL PROTECTED]
      >  Subject: Re: ODE clustering support
      >
      >
      >
      >
      > Hi Eugene,
      >
      >  The work that has been done was to make clustering development
possible
      > when necessary but for now it wasn't a really high priority. So
we've added
      > (great work of Maciej) callback interfaces in the process store to
notify
      > the local engine of what's happening, what's being deployed,
undeployed, ...
      > So the interfaces are there, right now the implementations are just
local.
      >
      >  If you're interested in that feature, you're definitely welcome to
      > contribute to the project. Maciej or myself can point you to the
right
      > places in the code. I think it would mostly be making the deployment
      > callback interfaces network aware, so that stores deployed on each
node
      > would notify the others and in turn notify the local engine.
      >
      >  So let us know if that's something you would consider and then we
can go on
      > with the logistic part of it (i.e. submit patches and all that
stuff).
      >
      >  BTW would you mind continuing this discussion on the ode-dev
mailing list?
      > Most committers are already in this list but discussions are usually
done on
      > mailing-lists in the ASF.
      >
      >  Thanks!
      >
      >  Matthieu.
      >
      >
      > On 12/21/06, Eugene P. Kostin <[EMAIL PROTECTED]> wrote:
      >
      >
      >
      > Hello:
      >
      >
      >
      > We are the team trying to select an open source BPEL engine which
would be
      > capable to work in a cluster. We are thinking of using ODE but as we
      > understand clustering support is in development. The timeframe for
our
      > project is about 6 months. Could you please let us know when we
could expect
      > ODE to be released with clustering support?
      >
      >
      >
      > Alternatively, is there an option for our team to contribute? We
have been
      > looking at PXE source code and have some thoughts on how to make it
run in a
      > cluster, but would not really want to invest effort in the engine
which is
      > phased out in favor of ODE.
      >
      >
      >
      > Thank you in advance for any response. We need to be making the
decision
      > about the workflow engine real soon.
      >
      >
      >
      > Regards,
      >
      >   Eugene
      >
      >
      >
      > ----------------
      >  Eugene P. Kostin
      >  Account Manager
      >
      > Software Technologies
      >
      > Tel. +7(8634)315200 x107 /9am-8pm GMT+3/
      >
      > Fax +7(8634)315300
      >
      > [EMAIL PROTECTED]
      >
      > www.softech.ru
      >
      >
      >
      >
      >


Reply via email to