| Author |
Message
|
| aronimink |
 Posted: Tue Jan 03, 2006 12:07 pm Post subject: Data container hijacked by another instance? |
 |
|
Apprentice
Joined: 03 Jan 2006
Posts: 26
|
I do not how this happened, and trying to solve the problem.
When the workitem was checked out, the value of all data container changed from another instance. Data mapping is correct, correct values are displyed in correct elements, but the entire data is from different instance. How this possible? Is anyone experience similar problem? |
|
| Back to top |
|
 |
| jmac |
| Posted: Tue Jan 03, 2006 12:25 pm Post subject: |
|
|
Jedi Knight
Joined: 27 Jun 2001
Posts: 3081
Location: EmeriCon, LLC
|
I have never heard of anything like this... and I would doubt that what you say is actually happening. If it is, you should open a PMR
_________________
John McDonald
RETIRED |
|
| Back to top |
|
|
| PisgahMan |
| Posted: Wed Jan 04, 2006 7:08 am Post subject: |
|
|
Voyager
Joined: 27 Jul 2004
Posts: 93
|
| Is it possible that the user and/or user interface changed the data? I have seen that happen and it is very confusing. |
|
| Back to top |
|
|
| aronimink |
| Posted: Wed Jan 04, 2006 9:02 am Post subject: |
|
|
Apprentice
Joined: 03 Jan 2006
Posts: 26
|
From the application, input container is copied to output container
with 2 elements are modified, other than that all remain.
Most of instance are OK, and can not re-produce error to test.
It just happens unexpectedly and get 2 different instances (with different data) end up with same data. |
|
| Back to top |
|
|
| fidelio |
| Posted: Wed Jan 04, 2006 10:52 am Post subject: |
|
|
Apprentice
Joined: 14 Sep 2005
Posts: 45
Location: AttainBPM
|
Sounds like an application problem 
You should use the data default connector on the activity to copy data from the input container to the output container so that your application only has to worry about data elements that have changed (it also insulates your applications from data container structure changes). |
|
| Back to top |
|
|
| aronimink |
| Posted: Wed Jan 04, 2006 11:58 am Post subject: |
|
|
Apprentice
Joined: 03 Jan 2006
Posts: 26
|
Copying input to output (at check-in) by the application is working fine.
The problem happens when checked-out. when the item was checked out, input container values are already swapped from previous values with different data from another instance. |
|
| Back to top |
|
|
| aronimink |
| Posted: Wed Jan 04, 2006 12:03 pm Post subject: |
|
|
Apprentice
Joined: 03 Jan 2006
Posts: 26
|
Like I mentioned, problem is not updating element.
All values are placed at the correct elements in the structure.
Entire container value is from different process instance. |
|
| Back to top |
|
|
| vijaycr |
| Posted: Sat Feb 10, 2007 12:32 pm Post subject: |
|
|
Acolyte
Joined: 11 Oct 2006
Posts: 62
Location: Singapore
|
We faced the same problem too..We raised a PMR but still unable to reproduce the problem consistently...There is another post "Container values changing" which is the same problem too. Any updates will be appreciated. Will post the solution if we find one.
I feel this could be a flaw in the Workitem id , might be I am wrong. |
|
| Back to top |
|
|
| vijaycr |
| Posted: Sat Feb 10, 2007 12:37 pm Post subject: |
|
|
Acolyte
Joined: 11 Oct 2006
Posts: 62
Location: Singapore
|
In our case we are using stuct to struct mapping so its unlikely to be adata mapping, more over the entire container set has a different set of values.
I thinks thats what everyone lese is facing...Weird |
|
| Back to top |
|
|
| jmac |
| Posted: Sat Feb 10, 2007 12:42 pm Post subject: |
|
|
Jedi Knight
Joined: 27 Jun 2001
Posts: 3081
Location: EmeriCon, LLC
|
Like I said initially, I have never seen anything like this, and I assume this is an application issue. You say you suspect it is something with the Workitem id. WHY?
Do you have your own client, or are you using the IBM supplied client?
_________________
John McDonald
RETIRED |
|
| Back to top |
|
|
| vijaycr |
| Posted: Sat Feb 10, 2007 1:20 pm Post subject: |
|
|
Acolyte
Joined: 11 Oct 2006
Posts: 62
Location: Singapore
|
I am not 100% sure...its just a possblity. This is happening once in a while..We are processing 1000's of workitems in a day. It has been reported more than a couple of times in a month. So, first of all it is not consistent and we are unable to comeup with a pattern.
We are using the persistent OID when we check out to check it back in.
We are planning analysing the out put on this front, Might be an unintended work item is getting checked in with the containers of the current workitem.
It is difficult to say,its just my guess. |
|
| Back to top |
|
|
| vijaycr |
| Posted: Sat Feb 10, 2007 1:24 pm Post subject: |
|
|
Acolyte
Joined: 11 Oct 2006
Posts: 62
Location: Singapore
|
| BTW we are using our own client, not the IBM supplied one. But we are more or less sure that it isnt an application problem though we havent ruled it out. We are analysing in this front too |
|
| Back to top |
|
|
| vijaycr |
| Posted: Sat Feb 10, 2007 1:37 pm Post subject: |
|
|
Acolyte
Joined: 11 Oct 2006
Posts: 62
Location: Singapore
|
| I am not sure how this persistentoid is generated. Is it from a database column or some where else? Does any one know how this is done? |
|
| Back to top |
|
|
| jmac |
| Posted: Sat Feb 10, 2007 2:53 pm Post subject: |
|
|
Jedi Knight
Joined: 27 Jun 2001
Posts: 3081
Location: EmeriCon, LLC
|
Persistent OID is not stored in the database, there is a class called OIDGenerator (I THINK) that is not documented that can produce what is referred to as the External OID (this is the Persistent OID) given the database columns representing the Internal OID (i.e the database column). The doc for this class was once found in the Support pac WA0C, I do not know if it still exists there or not.
My guess is that this is an application problem. If you have an open PMR on this, if this was a known problem I believe you would have a solution by now.
And I doubt that a bug like this could go unnoticed for long.
_________________
John McDonald
RETIRED |
|
| Back to top |
|
|
| vijaycr |
| Posted: Mon Feb 12, 2007 11:27 pm Post subject: |
|
|
Acolyte
Joined: 11 Oct 2006
Posts: 62
Location: Singapore
|
| Thanks John. Yes the supoport pac still exists and is supported with both 3.5 and 3.6. |
|
| Back to top |
|
|
|
|
