|
Home > Archive > BizTalk Server Orchestration > April 2004 > [BT2004] One message to more than one Receive Shape?
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
[BT2004] One message to more than one Receive Shape?
|
|
|
| I am doing some analysis for a new orchestration design and was wondering if
anyone could tell me if it is possible to have two different receive shapes
in an orchestration both subscribe to the same message (same message type,
correlation set and inbound port)? Will both shapes "see" the incoming
message?
Thanks in advance,
Dan
| |
| Christof 2004-04-18, 10:53 am |
| Makes not a lot of sense to me... Can you please elaborate a little more on
your scenario? As soon as one receive shape has the message, it is
available to whole your orchestration (or the scope the message was defined
in).
Best Regards
Christof
"DL" <dan@dont.spam.me.com> wrote in message
news:eHxom%23%23IEHA.556@TK2MSFTNGP10.phx.gbl...
> I am doing some analysis for a new orchestration design and was wondering
if
> anyone could tell me if it is possible to have two different receive
shapes
> in an orchestration both subscribe to the same message (same message type,
> correlation set and inbound port)? Will both shapes "see" the incoming
> message?
>
> Thanks in advance,
> Dan
>
>
| |
|
| In this case, I have two parallel paths that each wait for a different
document type to arrive. After each path receives the document, it needs to
go on to an analyisis step, but only if an "OK to Proceed with analysis"
message has arrived at the orchestration. I would like to put a recieve
shape in each path to wait for this message so that analysis in a branch may
proceed immediately if it has already arrived.
Imagine a bunch of parallel paths with "gates". All the gates need to be
opened at the same time, even if the flow in any one path has not yet
reached that gate. I was hoping to avoid polling in a loop to test for a
"gate open" condition.
Dan
"Christof" <bts2004communityfeedback (at) hotmail (dot) com> wrote in
message news:O40vD2$IEHA.3200@TK2MSFTNGP11.phx.gbl...
> Makes not a lot of sense to me... Can you please elaborate a little more
on
> your scenario? As soon as one receive shape has the message, it is
> available to whole your orchestration (or the scope the message was
defined
> in).
>
> Best Regards
> Christof
>
>
> "DL" <dan@dont.spam.me.com> wrote in message
> news:eHxom%23%23IEHA.556@TK2MSFTNGP10.phx.gbl...
wondering[vbcol=seagreen]
> if
> shapes
type,[vbcol=seagreen]
>
>
>
| |
| Christof 2004-04-19, 12:36 pm |
| I'm still not sure if I understand you, I'm sorry. A drawing would help I
guess.
Anyway, in my opinion the branching in BizTalk Server could have been
better. I think that is actually what you are reffering to. Communication
between branches actually can only be established using 2 mechanisms...
- exceptions
- a message box hop
A beautiful example of this is actually the HWS template orchestration,
which does exactly this. When one branch needs to notify another, it may
send a message to the mbox (using "direct mbox binding"), which is then
picked up by the other branch again.
The "exception" mechanism can be used as well. If you throw a message in
one branch, that may cause the exception handler to trigger, outside of the
parallel. This causes the engine to skip executing the other branches.
Hope this makes sense to your scenario. Please come back if things are
still not clear.
Kind regards,
Christof
"DL" <dan@dont.spam.me.com> wrote in message
news:OcmuLzhJEHA.2680@TK2MSFTNGP11.phx.gbl...
> In this case, I have two parallel paths that each wait for a different
> document type to arrive. After each path receives the document, it needs
to
> go on to an analyisis step, but only if an "OK to Proceed with analysis"
> message has arrived at the orchestration. I would like to put a recieve
> shape in each path to wait for this message so that analysis in a branch
may
> proceed immediately if it has already arrived.
>
> Imagine a bunch of parallel paths with "gates". All the gates need to be
> opened at the same time, even if the flow in any one path has not yet
> reached that gate. I was hoping to avoid polling in a loop to test for a
> "gate open" condition.
>
> Dan
>
>
> "Christof" <bts2004communityfeedback (at) hotmail (dot) com> wrote in
> message news:O40vD2$IEHA.3200@TK2MSFTNGP11.phx.gbl...
> on
> defined
> wondering
> type,
incoming[vbcol=seagreen]
>
>
|
|
|
|
|