Senin, 30 Maret 2009

SAP Future Data

In SAP, employee information is partition into more than one InfoType. Each Infotype has its own use. And betwen each Infotype the data could also be redundant. SAP data could be sent to IDM using two ways. PFAL and Change Pointer. PFAL is a manual action where all of the info type will be sent.While Change Pointer, is a scheduled action where only the changed infotype (delta changes) that will be sent.

By implementation time SAP data can be categorized into two, current and future data. Curent Data is the data that the changes will be executed immediately, while current data is the data that effective changes will happened in the future.

We found a bug in our previous implementor code when handling future data, it treat future data the same as current data. To fix this we'll need to put the data from the active synch form to the deffered task. Each object will be stored plain (SAP Plain XML data). When the deferred task is run, then we assign the plain data to a workflow to process the resource. This approach is better than assigning the workflow from the start. Because if something happened with the workflow (bug, changes) the Deffered Task that hasn't been run will Immediately affected. We didn't need to clear the deffered task and put the object all over again.

Selasa, 17 Maret 2009

IDM 8.1

Hi, did you know that IDM 8.1 has been released ?

Other than bug fixes and patch , there are 5 new feature. The one that interest me much is "External Resource Management". This feature would let you provision external resource that didn't connect directly with Identity Manager, like Laptop and Handphone. Interesting, to but i haven't got the time to explore it yet :)

Minggu, 01 Maret 2009

Selective update in bulk action

I find this finding on IdM 6.0sp1, and haven't tested it yet in other IdM version.
Consider this, you want to do bulk action to update bunch of user, but not all the user you want to update share the same attributes. Rather then making separate bulk action to update the attribute, you can 'ignore' update to the attribute by providing an empty value (,,), not empty string(,"",) on your script.

Take this example:
-user01: firstname="locco", lastname="creamer", email="locco.creamer@example.com"
-user02: firstname="nartha", lastname="rover", email="nartha.rover@example.com"

Then we run the followin bulk action script:
"command","user","global.firstname","global.lastname","global.email"
"update","user01","rocco","","rocco.mate@example.com"
"update","user02","martha",,"martha.rover@example.com"

This is the result:
-user01: firstname="rocco", lastname="", email="rocco.mate@example.com"
-user02: firstname="martha", lastname="rover", email="martha.rover@example.com"

Remember, always test your script on your development environment first to avoid any unwanted case.