Workflow Parallel Steps

[Trant, David]

Sorry for the delay in responding; I think I posted something similar here =
a few months back and can tell you definitively that the order in which par=
allel tasks executes is completely indeterminate.  I had a situation where =
someone designed a parallel branch, and ended up referencing the same conta=
iner element in both branches.  Each side of the branch had a step that pop=
ulated that container element, followed by a step that used its value.  As =
you can imagine, the workflow behaved fine when the order of execution was =
"Branch A, Step 1" followed by "Branch A, Step 2" followed by "Branch B Ste=
p 1" and finally "Branch B Step 2," but behaved rather badly when the order=
 was "Branch A, Step 1, Branch B Step 1, Branch A Step 2 and lastly Branch =
B Step 2."  Both branches in that case ended using the value found by Branc=
h B, which was not the desired result.
 
 -----Original Message-----
From:   SAP Workflow [mailto:Owner-SAP-WUG at MITVMA.MIT.EDU]=20
Sent:   Tuesday, May 25, 2004 9:05 AM
To:     SAP-WUG at MITVMA.MIT.EDU
Subject:        Re: Workflow Parallel Steps
 
Thanks Mike,
=20
I will probably be doing some testing on this in the next week or so, so I =
will let you know what I find.
 
Regards,
=20
Alon
 
________________________________
 
From: SAP Workflow on behalf of Michael Pokraka
Sent: Tue 25/05/2004 14:53
To: SAP-WUG at MITVMA.MIT.EDU
Subject: Re: Workflow Parallel Steps
 
 
 
Hi Alon,
'Off the top' I've learnt not to rely on things to happen at the time you
expect them to happen. More logically, since a task can be anything from a
container operation or an asynch RFC call - the latter definitely occupies
it's own process, so I'm pretty convinced that it executes in seperate
processes.
Based on pure speculation, I'd say steps are added to a table, and the WF
engine would probably read that and dispatch the tasks off into their own
process - near enough to simultaneous but not enough to absolutely rely on.
Would be interesting to find out if that is the case though.
 
Cheers
Mike
 
Alon Raskin wrote:
> Hi Richard,
>
> I probably will do that but I wanted to know if anyone knew 'off the top'=
 of
> their heads. The scenario is a little more complex then you describe but =
its
> not worth going into....
>
> Thanks for your help.
>
> Alon
>
>
>
> ________________________________
>
>> From: SAP Workflow on behalf of Richard Marut
> Sent: Tue 25/05/2004 14:32
> To: SAP-WUG at MITVMA.MIT.EDU
> Subject: Re: Workflow Parallel Steps
>
>
>
> Alon,
>
> How about if you turn on the technical trace and look at the execution ti=
me
> of the results? Or, get basis involved to watch what's happening.
>
> I'm assuming that each branch will contain a task that results in a work
> item being sent to someone which ultimately put those steps into a wait
> state and the odds of two different people processing them at the same ti=
me
> would be minimal.
>
> Richard...
>
> -----Original Message-----
>> From: SAP Workflow [mailto:Owner-SAP-WUG at MITVMA.MIT.EDU] On Behalf Of Alo=
n
> Raskin
> Sent: Tuesday, May 25, 2004 8:12 AM
> To: SAP-WUG at MITVMA.MIT.EDU
> Subject: Workflow Parallel Steps
>
> Hi Guys,
>
> Just some questions about Workflow Parallel steps:
>
> 1. Does the Workflow engine guarantee that two parallel branches are
> executed in parallel? What if there are not enough system resources to do
> this? Does it wait for enough to become available?
>
> 2. If two branches are executed in parallel (via the use of a fork), does
> each branch utilise a seperate batch/dialog process?
>
> Regards,
>
> Alon
>
>
 
---------------------------------------------------------------------------=
---------------------
This message is for the designated recipient only and may
contain privileged, proprietary, or otherwise private information. =20
If you have received it in error, please notify the sender
immediately and delete the original.  Any unauthorized use of
this email is prohibited.
---------------------------------------------------------------------------=
---------------------
[mf2]